·No text on this page.

·B FOR
FOR
MEDICAL
WRITING
A USEFUL GUIDE TO TRI\CIPLES AD
PRACTICE OF EFFECTIVE SCIE\TIFIC
WRITI\GAID ILLUSTRATION
BY EDWIN P JORDAN, MD
AND WILLARD C SHEPARD
Quelquiel Pruccimos Esto Bretus
WB SAUNDERS COMPANY
PIILIDELHII LO\DOT
VCILII

·COPYRIGIT 1952, DY W D SAUNDERS COMPANY COPYRIGHT UNDER THE INTERNATIOVAL COPYRIGHT UNION All rights reserved This book is proteclad by coprright no pari Il nxay be duplicated or reproduced in ani niannet without MTINICE peruission from the publisher fade in the United States - America of the Press of B Saunders Company Phulladelphia

·True else in anting comes from art, not chance,
As those pose asiest who have land to dance.
ALEXANDER POPE

·No text on this page.

·PREFACE Experienced and successful mers are agreed that writing is to a pft with which somt ure naturally endored, but that proficient composition is a laboriously acquired skill Good wriung is one part inspiration and mine parts pietupinion Many medxal men and women seem to find the preparation of articles for pabiscation in urduous tash. Consequently the pur pose of this book is to try to bdp then say what they mean and arord-or climinate some of the faults to the final manuscrip what may present the finished product from doing justice to the *orl on which it is based. There are many ccellent books on English composition framour and style which thus manual malos no pretence o duptaaaling. If the supportons presented best proic of some proe ucal help to med kal there in the preparation of their papers 111 be fulfilling its obyczne It will be understood that of the underrigned the first b se sponsible for the chapters on medikal wung and statistie, in the koond for the chapter on Mustrations EMTN P JORDAN WILLARD C SHEPARD Julvon 1957

·No text on this page.

·CONTENTS 3 IAFTER 1 THE PRELIMINARIES Chad of Subcat Radrip Recording Lotos and Retarnets Part the Story Entrs the Audio Odot 3 4 + 5 $ 7 CHAPTER 11 THE FIRST DRAFT The Sant Projet Langth and Safe Saan East Doors 7 7 9 CHAPTER 111 THE FIRST REVISION Bulma mi trotoo Order andError Orma 9 10 10 11 CHAPTER IS THE SECOND REVISION Snik O Wintot-Tur Jarpon Sun 1 mentors 11 14 14 (HAPTER THE THIRD REVISTOS 17 Scrits od korth Hack Akulam Am 17 17 19 (HAFTRI THE BEGING AND THE END The One acces 2 ·Contents CHAPTER VII SPECIAL PROBLEMS Footnotes Protocols Bibliography Cus Reports Editing and the Editor Form of Copy And Proofreading CHAPTER VIII MAKING THE INDEX How to Set About It Special Points How to Alphabetite Typing the Index CHAPTER IX ILLUSTRATIONS The Prepantion of Illustrations in General Identification or Illustrations Preparation of Protographs Photomicrographs Roentgenogram Label The Writme of Legends Permission to Publish Copyrights Miscellaneous Information Photoongraving Line or Zinc Engravings Halftone Engravings Illustrations in Color CHAPTER X STATISTICS EMPLOYED IN MEDICAL PAPERS Aun in Using Statistics Collection of Data Analysis of Data Presentation of Stabian Darn APPENDICES References Style for Listing References Common Abbrevptions in Medical Manuscripts Proofreaders Marks Table of International Atomic Waghus Useful Equivalents Table of x' Probability Sale INDEX ·CHAPTER 1 THE PRELIMINARIES Aceescal papers CUST ID *nle, 26.d the final paper is more Mey u te apie ben the sole project has here carefully planned inahang For this reasona discusion of media wnung an fromah bepa ih kome of those preliminary factors about belp Lo determurshether the incle will appear in priot, ard of ndoa ton regist Chovce of Subiect Pusschny take so much of modern medical educanson michonochabcs of Fig mousty acumulated ſariwal mforma UCL min fhverirs find it hard to select a subat to micsupate and her. But thetil this is dkrae there om be bo progress toard the medral anxie. 110* to Get IL There art man ns of obtaining deas for a suable lopte 41-50 en el meeting or conference will to cal points 21 bact Merinosed door u deficitat, shonckell the queron Zornothy and a possible sabe nadio dodatxon comes watery TUTTL At oue sople hoor burg dical conferencelopor Mual dono e subjects bort mopht kod thensches to such thus to an truck The correct wedraal literatut try be larly cloned as a source of xas, discouns erminou the more he surtabe for Kotr und write by Onenation of a market or o posing of parents almost 10 SETELS Bowtimp ce imperfecthy underwood nd therefore aan de studied. This Laboratory us as lentile and fy the name and the procesinst the fill of it the cathes of fruits something to stuch lot to be (17) 23 a varit of suhete mattet ur doctors besta traite connes au comerulon leads woim lo fer of Idilnon mas of the froile la 1rows ays 0.48 min sekeet the memoried to the used tattoos 11table. A 3 ·2 Medical Wnting common failing is to be too ámbitious and to start an investigation which will require more pauents, more laboratory equipment or more time on the part of the worker than is likely to be available. Such factors as these as well as the skill and training of the writer must be weighed before making the final selection Significance The medical literature is already overburdened with papers which are merely rehashes of information previously published Hence it is important that the subject should be one which stands a reasonable chance of making a contribution to the advance of knowledge. In other words, the study and subsequent report should be significant. It may be difficult to anticipate the full significance of an investigative problem but the question should always be asked of oneself in advance What will the results of this study mean?" For example, a report based on the observation of three or four patients if articles describing several hundred similar cases arc already in prot does not stand much chance of being of enough significance to justify the effort of writing it! Some physicians are too critical of this factor and seemn to feel that they should not write at all unless they can contribute something which would revolutionize accepted thought. But this is the work of genrus, and most contributors to the literature must walk the tightrope between these two extremos Tineliness The subject chosen should be evaluated in advance for ume liness. A study which at one period is worthwhile and should be published, may at another time constitute a waste of valuable time. One example will serve to illustrate this point Soon after penicillin first became available many physicians started wing it on their patients. The results, which were often spectacular were submitted in large numbers to medical journals in the form of reports of from one to a half dozen casos For the first few months --since these studies were the only source of information on the clinical actions of penicillin-most of them were accepted then come studies of larger groups under more controlled conditions of observation The result was that soon the authors of case reports just as good as those which had been accepted by the editors short time before found their manuscripts promptly returned This factor of choosing the right time cannot be safcly ignored if the ·The Pralimmani 3 uibor does not with to run the rol of baring bus work come to maught so far as publication is concerned Reading Up Not long ago I rad en excellent thesis prepared by a capable young physiciansbo was just completing hus training in a medical calty The work had been carefully done and was worthy of publication test for Ox defect someone else had published an almost Identkal article about the years artier' The young man had spent bundreds of hours on his paper and it was my unhappy duty 10 tel brm of the estetxe of the published on the This expTITKE, hacks by no means uncommon, illustrates the necessity of reading not only while the study in progress, but also before the final sekection of 10px has been made. Textboolis occupy a distinctly secondary place for this purpose. They should the used merely as a gunde to those onginal reports which the aathor of the leubook found Important. Access to the literature on a partacular subject is best obtained in three ways the war icir Cumulate Inder Virdinus published by the American Aledical Assocmion the Index Catalogue of the Surront General's Office and the bibliographies of the most recent and comprehensive ankles on the top under study At the time of uniting these words the suis neither quanerly not cumulatie and is doiresungly far behind her ertheless it offers the samost com plete irse of references vailable up to the preceding two years or so Ecty physician sbould become famlarized with it and should be able to track down any medical subject though it may be accasary to consult seneral subject headmgs in order to do so The Inder Catalogue is less useful for current material but it is nonenheless salvable erecally in finding references of holorscal monte Published articles on the subject scherted for study are of mense valde as a source of bibliography The author of suct marers, homene bare made their onn sckctions and their own cilation and intensietauons so that it is always well for the prospectnc*DICT lo read from original sources himself and to add 0 referencts, Mauricularly those of most recent onkin The list of references should be personally coronaled. The Taco habit of exting 3 stelary or cien a modal libraran ti prepare a In of references and make abstracts for use in the manen dangeToTM No one but the author chahearted to make a proper selection of 10 ks, and a morte stufung the respony mlity in this manner nearly this leads to KTIOUS error of fait ·4 Medical Wnimg or interpretation which cannot fail to reflect on the reputation of the writer There are prolific writers who differ strongly on this point, claiming that their time is so valuable that they cannot look up the literature themselves this is an honest difference of opinion but, in my judgment, the quality of their publications suffers from the lack of personal attention to this important phase of their work Recording Notes and Referencos Notes of reading, clinical or laboratory observations, and stray thoughts are lost or are likely to be inaccurate unless they are recorded in writing at the time. For this purpose some flexible system of note keeping is invaluabic I like the method of using plain or line 5 by 8 inch Cards (the 3 by 5 inch size is too small) which can be carried in the pocket. Excerpts or abstracts of refer ences consulted can be made easily on these cards but the source should be accurately and uniformly placed on each card even when more than one natation is made from a single source. It is convenient to leave a wide space at the top so that the subject, og etiology pathology treatment can be entered later and perhaps changed once or twice Any data accumulated in this way can be reviewed from time to time, and subject headings may be entered shuffled and rearranged so that when writing begins the cards will be in the exact order to be followed in the composition Planning the Study An unplanned study almost invanably results in a disjointed final manuscript. One of the carliest steps in the preparation of a medical paper therefore, is to ask oneself two quesuions What point or points can be clarified by the proposed study” and What evidence must be gathered to convince the reader? This is not the place to analyze what constitutes scientific evidence. The most common failing, however consisu of inadequate controls or none at all, and this is more true of clinical papers ihan of those pre senting laboratory data. Rors, writing in the Journal of the Americon Medical Association reviewed 100 articles from five leading medi cal joumals published during the first six months of 1950 and found 45 in which controls were not used 18 inadequately con trolled 27 well controlled and 10 in which control was impossible This scarcely constitutes a record of which we can be proud In addition to failure to use controls in many types of investi gations where they could have been employed there are other frequent defects in planning the project For example anyone who ·5 The Prendre has ulempiad to compue a KTID of din cal observations from hosptul chans acumulated are a period of years and recorded by a number of diferent physkans an testify that the nformation mostlestad is often lacing or is too afue to be of value. Further more, different obxectivos do por ahiyo micrpret the same climcal orco Laboratory data m the same way Hence, when et possible, charts which are lo he wed 19 supply dimxal data on a series of parents should provide space for the desired mfortation und some sort of standard should be ke up so that different obseries will record the observations in the same way Obrowh the cannot be done for the patients have passed from observation. The cmena baxcta sbould apply to clinical studies bear with at least qe force bea esperimental axences to be presented. Controls out also be used, and phenomena obosenied must be as readily apparent to one trained in cuptor B to another All thus of fundamental importance to Lbs medical mion- Knife because it vitally affects the quality of the cod prodot and soodentally the basic work has beeo adequately completed at the utte antung starts, it mals the compoupon itself Doch Caper Evaluating the audience Everyone has lissented to talks of speeches which fel LaL DOT because they did not contam DLETONg und worthwhile informa- 1xom tut because they ere not adopied to the parikular audience or octauon This is kys common in published papers than in Sacche isomally because editors sand in the way Honnet the mer bo does not pre thousit to the kind of ruders he an to resch and the youma huch a best fried to do it nsis rection of what othera ke may be a finc The nter must asl bunk before preparing an outlet bether the infor malion he has to prevent is best started to the grocral Pracutioner the iperoleh ke research other of sotae orher froup Asringe ment and organizalot of material are partxularly dilcult if the authay naming at med groups Cantanh both the hope and Come ti the cineserint should be influenced by the ancipated vuit Outing There is a preness about an er helming majority of nounal article aashich indicato that most entry follow the structors outline and perhaps lack the literary court to solc out for thetoselies, Inundacion history strology pathology ·6 Medical Wnting pathogenesis clinical observations experimental observations, des cussion summary and conclusions follow one another with monotonous regulanty Not only does this similanty become bor ing but it leads to considerable difficulty in avoiding duplication and repetition The writer is often left in a quandary about whether to place some significant reference under history ctiology pathogenesis, or discussion If the paper includes somc pathologic observations, he is forced to decide whether to have fHd sections on that subject one involving the work of others and the other his own. If he decides on two, how is he going to compare his results with those of others without repeating? I do not know how this curous ngidity has arisen but suspect that the current craze to examine Medical students and doctors on every possible occasion has led to the establishment of a roughly similar formula which satisfies the examiner Whatever the cause I should like to plead for the discard of this set system But it is not safe to omit the outline before starting to write since this would lead to complete disorganization If there should be no absolute formula (and I for one believe there should not) how then can an outline be prepared? First the writer can ask himself what is the most important point that he wants to bring out in the paper With this question answered and jotted down in wnung, an out line can be built which will carry the evidence in a convincing manner For example the author may omit entirely sections on history and etiology and include all that he wants to say on these subjects as he recounts hus on observations or in the discussion. It is comparatively easy to juggle the outline headings until they satisfy the writer that everything he wants to say will be included Effort expended in this direction will prevent difficulty later when a decision must be made as to where in the paper particular observations or references are to go Too many papers are submitted completely without headings Consequently it would be well if the outline headings were used as section headings in the completed work or at least submitted separately as a guide to the editor ·HAPTER JI. THE FIRST DRAFT The Start .. Most of us find that the hardest part of writmg us getting Harted Excuses Tot deby multiply tile rabbits and lack of time copy all the resul For the busy physkran or surgeon this is no minot hurdle. Some meet the problem by hiding away from their parents and the telephone for an hour or to each day Olberstas tonn, rent a hard room for several days, or burn the midnight OH, Hon 1er accomplished, ute must be found, for without the opportunity to conceptrate buch umc alone pron da pood woms is unattain able. Hith tome set aside the dread moment comes en entwally when the water inds himself sealed pencil in hand, facing reams of Intan raret hos bands scaun miklly and his mind a blank, Is there anyone at this stage who does not so that he had not commartiod himseln The first few words are like a plungt imo an ice cold pool. It 1991 so bad after the start has beco made Each uniter has his own methods I uke the rapidly antici first dran ketting the faulis creep in where they will and loving their correction uniillatet The has the advantage of Ketling the writer arm up without constant Interruption to the form of thougbi and is less likely to lead to abruta changes in pact of style. Every first draft bonnet cartiet onc parikular danger for the medalsmer because be bas worked haru in the composition, he all too ollca considets cach word or phrase of the onginal draft as his own brain child and struggla again later tanges or deletions as though they constituted inſaniaide) Proper length and Scope Many ruhams. en before they start ning are concerned atast there long that are should be this question should be act and uninem, I think, dict the per has been fitten and They had e There no standard some repets of three pages are inding thets of twenty may be too short for the subzett matter 7 ·8 Medical Writing to be presented. A sare ruic of thumb is that a manuscript should be as long ts is necessary to tell the author s story but not one word longer If there are definite space or word restrictions im posed by an editor or publisher- and this is not too frequent--the trimming of the paper or additions to it can be best accomplished during one of the revisions. Some Early Donts There are some important don t cient in the first draft Reports from the clinical or pathologiu laboratories, or from the department of roentgenology are commonly attached to patients charts They are not written for publication and to incorporate them verbatim in the paper is wholly undesirable. When such information is to be included the original report should always be rewritten and woven into the fabric of the paper Occasionally I have encountered an author who assembled a mass of notes sorted them in rough order and tumed them over to a medically untrained 'writer to whip up into a paper True, some find it extraordinanly difficult to write and may need help even in preparing the first dralt, but the practice menuioned musi be condemned as avoiding the responsibilities of authorship A medical paper should be the composition of the author and not that of someone else. With a detailed outline prepared and references, notes_and data arranged on cards according to the outline it should not be too difficult to write freely and rapidly once the start has been made. The blank pages fill rapidly and before long the first draft 1$ completed. Then comes the slow and difficult task of revising If Robert Louis Stevenson William Osier and a host of others could spend as much time and cffort on their revisions as the records show those of us ho do not aspire so high certainly should not resent the task. And the difference between a first class manu script and a second class one comes from the revisions. What makes easy reading was hard composing ·KAMILR 4. THE FIRST REVISION There are u MANY *235 of DUS us there are wnions. Come quently the jugeHions which follow are rosdem ibe ful talin. tion that somentet i 1st to make more mapas ihan those proposed some lenet and that different combinations often more more convenient of horier santed to the panxular occasion Regardless of bar method followed, I would ware that the manuscript he set asxde for Ance of tho before drafting the final copy in order to en the subconscious mind to attack the subm xol and all on the paper to be oned with fresh utention zod umferese. Those to do this wie cease to be astonsbed by the CTTOTS they un cont and the mattonements they can alloch Fourressions of ik sosedical paper are sugeested as a mer totiu, tant in each of them the author will need to min more than Onc aspect of his planuscript The first of those revisions may be directed at the paper as a whole. Is it adapted to the audience for hich it rotended and the publication to bxh it is to be sent These questions must be kept in the forefront of the mund while muling the first animal setesding. At the winterlox, sacral other foneral features of the manuscript an be reviewed. Boroncand Propornon A bordseye its of the first draft should readily real hether the material textied as roperty tabanced. Most first drans con un too much of Sortie falures and not enovrh of others this is the umt to storien of errand the sections which need it. Some mical ropers, for cumpic, contan lengthy historsal revious al pagos Ik with onherrt of Inovase reports or other ipinat coninburon such a peper hour balanct and perds yamaton In other words, the motions of the paper should haar lexal relationshup to each other and to their relathie im Tant to the whole Erramon of thone portion of the paper which need it does Dey usually Moon 27 STIK dificulte, but shortcang sections alry entien often casos perling problem. Onc useful mi uta may male the poh RT is to braon paragraphs or ?

·8 Medical Wnimg to be presented A safe rule of thumb is that a manuscript should be as long as is necessary to tell the author s story but not one Hord longer If there are definite space or word restrictions im posed by an editor or publisher--and this is not too frequent--the trimming of the paper or additions to it can be best accomplished dunng one of the revisions. Some Early Don's There are some important don ts cven in the first draft Reports from the clinical or pathologic laboratones or from the department of roentgenology are commonly attached to patients charts. They are not written for publication and to incorporate them verbatim in the paper is wholly undesirable. When such information is to be included the orginal report should always be rewritten and woven into the fabric of the paper Occasionally I have encountered an author who assembled a mass of notes, sorted them in rough order and tumed them over to a medically untrained 'Writer' to whip up into a paper True some find it extraordinanly difficult to write and may need help even in preparing the first draft but the practice mentioned must be condemned as avoiding the responsibilities of authorship A medical paper should be the composition of the author and not that of someone else. With a detailed outline prepared and references, notes and data arranged on cards according to the outline, it should not be too difficult to write freely and rapidly once the start has been made. The blank pages fill rapidly and before long the first draft is completed. Then comes the slow and difficult tash of revzıng. If Robert Louis Stevenson William Osler and a host of others could spend as much time and effort on their revisions as the records show those of us who do not aspire to high certainly should not resent the tash. And the difference between a first class manu script and a second class one comes from the revisions What malos easy reading hos hard composing ·CHAPTER THE SECOND REVISION Style 1o the second reason the gyle of the part may be pundered Bus LAIB ML There are 25 many different detinatos as there are promtho bave me on the soljet For purposes of medical ting, at leas. I doubt that I should be considered 25 Ke Safar moattainable object the acquisition of xbxtos en only to the culty endowed. I like Canby s definition oks the roundur baxs has lo he punaan tety Con Vattben Antold put much the same Pertet dar l an tear thora al Bhat moth. Have amb la Dda Abdart you. That is the mert ottic If such adikt os sound and I would he bold to quand with - dort in this mocy of the manoseapt can be directed mward malog cria lhat ach thought Is erfressed cheart and lecs euah tar the mer means. Evdenth many matical *nenter hard to do Otraute Waling-Puddery It fofinets from the You Cant Win departner of the en Sohn Het te metala ponerareke.them whom type E 21- Donger that s drurity the prosent et sermohoi te kon itch oney Manche dal parecs dc no atre RL by the Litote mes koromby creating *HTMrs mothes from north auto dalsstra) got Touth Damot sitter mn mal be hot ned the alap ested to put Kereti.munistyndelman ·10 Medical Writing sentences which might be omitted and then to look over the section as a whole to see if the deletions have helped If the bracketed por tions can be omitted without ham they can be crossed out Another method of eliminating unnecessary material and banging thc paper into balance is to study the review of literature, which is so often longer than necessary When someone else has recently published a comprehensive review of the subject it is often sufficient to cite this source, with only those additions which have appeared moro recently or which the writer particularly desires to empha size Balance cannot be defined in absolute terms and the author shows his skill in achieving it in his writing very much as he might in his dancing or clinical work Order and Emphosis Some of my views on the order of presenting the material in a medical paper have been expressed under the section on Out lining (page 5) But a final look on whether the matemal is arranged to best advantage may be taken during this first revision This too is intimately connected with the nature of the report, the audience to which the paper is addressed and the point or points which the writer desires to emphasize On many occasions what the writer really wants to stress is buried deep in the paper where it is almost impossible to dig out It is well to ask oneself the question What do I most want the reader to get from this paper and is this so placed and so cmphatically worded that there is no chance of missing it?' Emphasis can be obtained in several different ways by repetition by presenting the evidence in such sequence that the reader anxiously awaits the conclusion and by the arrangement of the paragraphs sentences and words Within the sentence emphasis can often be obtained by placing the key word at the beginning or end of the sentence rather than in the middle Placing for emphasis applies also to the paragraph and indeed to the arrangement of the paper as a whole Omissions Finally 10 the first revision the author might pay attention to whether he has left out anything which should have been included Here is the chance to make a final study of the literature, to check the clinical or laboratory observations, and to ask oneself whether the readers likely to want to know about something left out which might just as well have been included. A transposition of identives is helpful if the writer can imagine himself the reader he may spot several things onginally omitted but worth including ·13 The Strand Revision ork-Jibble thought into pohn ilahtc immis. Resist the capulse to say that Abd schrhungu machine made and me instead hearty 11 alions on the industry lend themselo to performance by macture, and an grados of meas dotlung sold in significant quanunt insolve 2107) substantu) amount of machine work Put the obvious in Ibe terms of the ununtainabit. If the thought you wish to conting that cacharxlt semt 10 the cat is handled separatets you an word thus shea something like this "Withm the cleaning plant proper the business of the mdustry imohes several well-defined processes. bext, from the economic point of Kay be characterized simply by saying that most of ibem require separate handling of each indrul pment of PIECE OT mutant to be cleaned" Announct what you are not to say hefore you say it, or ar what you have ad alter you say n. Fmally Hillson 61%, you must defend your style as sentific Sner at dear cumple Eorplish as popular of joumalistke. "Explain your failure to put more megtal meat into your ning he ym. on the ground that the social i tots ho want to be sepufic hdrone that we ao base stie des or of human bebavos and trustworthy productions in the scientific sease only as we build adequate lawonomic systems for observable phenomena and sym hol System for the musirubtion of deal and abstract cubes. It is all well and food 10 quote borrible euamples of what SOTITEX disc has fition, but the medical author ofira faced itboing nd of smilur stuff in hrs on a first draft- entirely dillereol proposition. There are mass of maling that cusut One 15 to ky s le dit mietiepe betworn the use of the first dran and the res is on this olien permits finding and COTTecting poddery hxc law fooc quite undone at first Another 3 10 mne the paper to a cotle for to read and closure of this b done in food fasthed the readers comments are meted without resentmcat the ruhs are invaluable. Ako theniet homself at this ren Bion caa suunit has a fet ualements that art not clar und for Ons in the mi or for more hicho duplicated by charts my us The lanter foran orspeisen eurem common ament modxal niets. Its concessary the unitet should desde herb the data he has to pan can be grasped more easily on the roager in the form of this chans, photographs of running descriptXtam the L Oort this has been derde it is pot neces un LAN oming thKG, Onerio 2 table and again in the iait ·12 he says Medical Wnimg Sir Ernest Gowers, whose father was so well known in neurology in hus delightful book Plain Words, calls this kind of obscure writing puddery 'The basic fault of present-day writing is a tendency to say what one has to say in as complicated a way as possible. If the patient who came to exitus or progressed to fatal termination had just died some of our forests would have been preserved! “Don t say Quiller-Couch pointed out, Hc was conveyed to his place of residence in an intovcated condi tion say He was carried home drunk Puddery is apparently a failing of bureaucrats and sociologists as well as medical writers. Essentially puddery seems indistinguishable from padding and repetitiousness both of which reflect an uncertainty about what the author wants to say Gowers gives a wonderful example of saying at length what is so obvious that It need not have been said at all The planning of the Government Evacuation Scheme falls in o two parts (a) the anangements for the removal of persons from the dreas to be evacuated, and (b) the arrangement for their reception in the areas to which they are transferred. Those arrangements will not be of uniform concem to all local authorities, the arrangements for the removal being a matter for the authoritia for the cacuation areas and the arrangements for reception for the authorities for the receiving urats. Our profession has plenty of company Williamson has discussed the same subject in an article entitled How to Write Like a Social Scientist He says it has become a habit for social scientists to clothe their thoughts in smothering vertal garments As an cxample he gives this In the long rum, developments in transportation, housing optunum size of plant, etc, might tend to Induce an industrial and demographic pattern similar to the one that consciousness of vul nerability would dictate. Such a tendency might be advanced by public parnuosion and government inducement, and more effectively of the causes of urbanization had been carefully studied Indeed there is no better statement on this subject than the sit rules laid down by Williamson Never use a short word, he said when you can think of a long one. Always say currently rather than now sufficient instead of enough and 'lermination instead of end Never use one word when you can use tho or more If you can say "available evidence would tend to indicate that it is not unreasonable to suppose," use this instead of probably Put ·13 The Stond Revision one-syllable thought into polysyllabuc terms. Roost the impribe to way that man s clothing is machine made and white instead "Narly all operations in the mdustry lend themselves to performance by machine, and as mdes of mens clothing sold in apniticant quantity lavolve very substantial amount of machine work Put the obvious in the terms of the unmielligibk. If the thought you wish to conveys that auch article sent to the center is handled sepututcly you can word this idea something like this "Within the deanmg plant proper the business of the industry involves several well-defined processes, which, from the economic point of VICW may be characterized straply by waying that most of them require separate handling of auch nodividual garment or piece of material to be channed." Announce what you are gomng to say before you say it, or say what you have ad arter you hy it. Finally Williamson Ayr, you must defend your style & scientific Soar ut dar simple Eoylish us popular or journalista. "Explain your falure to put more mental sweat into your writing," he says "on the ground that the socal scientists who want to be screntisc belxeve that we can have scientific description of human behavior and restworthy predictions in the settific keuse only as we build adequate taxonomic systems for observable phenomena and symtbole systems for the manipulation of deal and abstract entities, Ito all well and good to quote bomible cramples of what someone else be written, but the medical author is often faced with petung nd of similar stoff in bas oorn fing draft- an entirely diferent proposition. There are ways of making this auct One is to be a few days latervene between the writing of the first draft and the revision this often permits finding and correcting guddsty wbuch bad gode quite unouced at first Another n to me the paper to a colleame to read and criticise of thirs done in good fath and the readers comments are accepted without resentipent the bus ure imaluable. Abo the water bumsel at this revision an krutime has paper for gatement that ut pot clear and for repeutions in the text for susterial ticho duplicated by chans or tablet. The banter form of repetition #stressly common umont medical titetu. It is unnecessary the writer should decache #bether the dats he has to present can be grasped more austy by ibe reader in the form of tables, charts, photographs of runnmg description is the tal Ooct this has been decided it tot nece sary to say merything to, once in a table and again in the tel 1 ·Medical Waling Jargon and Slang There is a subtle but ral difference between slang and Jargon which some medical enters fail to appreciate Slang applies to words or expressions which are coined comparatively recently because they contnbute some color or shade of meaning which cannot be so well expressed in any other way many English words and phrases which were originally slang are now in good usage and have greatly enriched the language Jargon refers to technical expressions used by a profession or cult which by no stretch of the imagination can be considered good English and which are often confusing, not only to those outside the fold but often also to those within it. The case went on a spree is an example of both spree was ongipally slang but is now clear and acceptable but case as used is jargon and the sentence is unacceptable. Mediciac, being a dynamic field of activity is constantly receiving new words some of which may be considered slang. These, if they clearly express a meaning which cannot be sold as well in any other way and if this meaning will be clear to the reader may be left as written Occupational jargon however, which is so favored by current methods of writing patient histories and reporting hospital conversations must be ruthlessly deleted from the written manuscript Such thlogs as using case instead of patient (or man woman, boy or girl) should never be allowed in pript. To speak of gastroscoping, appendectomizing. or cystoscoping a patient is inexcusable. This kind of jargon is all too often carried over from careless speech into a written paper A few other careless expressions of the bedside or the laboratory which all too often creep into the manuscript are operate for operate on acute abdomen for acute condition within the abdomen appende for appendicitis, diabetic for pauent with diabetes and cardiac nephritis, etc) chronic car for chronic outus media, specific for syphilitis, dues for syphuis, gland for node normal solution for physiologic solution of sodium chlor ide popular for commonly used and negative for absence of abnormalities or negative result of test. Fine Writing Although not now as common as it used to be, medical waters occasionally feel inspired to insert flights of literary fancy into the manuscripts An autopsy report which says thc teeth sparkling like gems in the sunken countenance Is hardly ·The Second Rorrosion 15 appropriate to the subject. If a writer feels Impelled to include phrases of this sort kt huma by ul mean put them and uale ibcto out at the frst opponunity A urollar fauh that sometimes gets into the first draft the insertion of words or piraso in some dead or foreign language This too has karpety died a natural death, perhaps because of changed educational empass But this lots Deed not be moumed. Rarety is the use of foreign ctpressions more than an affectation. If soch are found in the first draft they are better remondd, except for those lon hoch bave no English equivalent.

·No text on this page.

·CHAPTER 1 THE THIRD REVISION In the third Insion One can get down to the details and study the boxe building blocks of the paper Paragraphs Apparently the tric function of the paragraph b not well ondergood. Aparagraph should ordinarily COOLAJA tenaKTtonces connected by a common thought or there. Madial papers are replete with parmphs of almost unending length in which senal different thoughts are expressed, often witbout lopical order it is not uncommon to read an article in which a angle paragraph runs for two or three printed peopes. This works o bardship on the reader who may be forgonen for turnmg eben bere. Contrario, somentets seem to forget that the pangraph should be actually more than I engle sentence utkess the sentence is purposely kt sport for emplugs. Scagle potence paragraphs keep the reader constantly happing about trying to follow the sequence of thought, and redug them bs tiresome. la thus revision, therefore, paragraphs ao be constructed or reconstructed to that they will set to the function of developing thought out of acral sentences and forming them so that the reader who find then casy to understand and follor with the crew Sentences Medal THOTI Seeta to get more upplad up in their tentenca 20 any other part of composition Part of the trouble sectos to be the result of failure to realize that a uople sentence should express only one thought. The common results that sentences are dien too long and so confusing us to be scarcely intelligible. The follow my B a serra de cumple-its conuined in a paper cialty subtried to a medical Journal Apno othen dooded that park of the tomat of bon coilerpadualog of the woter of depret, focus of the VD had state of the funds of years be professor of medico o arpaty the Loremost medical alce of the United Sutes the matta olen betondobou physican or house ·18 Medical Writing surgeon for the largest and most completely equipped hospltal in this country of the number of years he had been mored in an extensive, reputable, ethical private practice of the number of recommenda tions for excellence of character und ability ghen him by reputable aumining boards of other than basic stence states, and regardless of the emergency of his present stuation, u li health or the Ilnets of a member of hos family accessitatag rettioval to the more favorable climate or other conditions of a basic stence stats, such a thoroughly educated, qualified, but sorely distressed doctor must appear before a Basic Science Board, made up mostly of laymen, who know nothing about disease or the practice of medicine, und pass cremonations in Anatomy Histology Pathology Bacteriology Physiology and Chemistry-(Some hata omat one or another of such subjects and substitute Hygiene or Public Health) with a general Average of 75 per cent of correct answers, for which privilege he must pay a fee of $15 to $25 before he may appear before the medical examining board and apply for licensure through reciprocal endorsement of ho examinations in the same identical tubjects by his college of grad utron and by the state board that sued his license to another state, for which reciprocal lleensure he must pay another fee or $20 to $100 in order to te allowed to car an honest liimg in the Basic Science State But shorter samples can be almost as bad This one, quoted by Gowers will surely strike a chord in the minds of all physicians who are also taxpayers A deduction of tax may be claimed in respect of any person whom the individual mamtains at hos own cyperse, and who b (1) a relative of hu, or of his wife, and incapacitated by old age or in finity from maintaining himself or her cit or (2) his or his wife widowed mother whether incapacitated or not, or (3) his daughter who resident with hum and upon whose services he b compelled to depend by reason of old age or infirmuty In this third revision arch sentence should be examined individually to make sure that it accurately expresses the authors meaning. In most awkward sentences the correction lies in short ening, since short sentences are CASKT to compose than long ones Rearrangement of words and phrases is often necessary All this is casier said than done and writers often have a bad time trying 10 revise a sentence which they know is bad Everyone, I think encounters this difficulty Sometimes it can be solved by giving up trying to revise the offending sentence and restating the thought on a fresh sheet of paper in an entirely different form This frequcotly ends the problem often the restatement does not con employ the same hords In other instances I have found it helpful to break the sentence into two or more new ones with some rear ·19 The Third Region Topment of words of punsch Uswlly the lime of attack pertuits some de mond which onbet suplify the final forat. Faced with in a ward sentence which diffon 10 mand, the author thould remember that be b pot obliged to reword å in a smelc senience metely because it was written that way anpinally Words, Words, Words The Coundation of langues spolen or written, is the indu dual word and its positioning in the sentence The rules of grant par are supposed to come all this and make it easy but many of us ne forpotten what premmar it may have known too grammar alone does not sohe all problems of img Rading and practice prove the ability to select words and put them together but choosing the ngtut ord and ung it in the right płet BD EASY I beliec mopossible to teach hou w select the right word It can only be learned. Yet there are methods of easing the job and also common faults to avond, a non lede of which may be useful to the medical wract The first practple, pertungs is to choose a Corano nord rather than sa common ose, provided the TOTTINET will she Bal. Many authors, particularly those lo Locbaical Delds like malkine best to feel that their colleagues Il comsades them more highly they dosplay their learning by USPT long and compared words. Quote the serve nart, archak, of everloog words sboukl be Arorded unless itey erpress an tua meaning och cannot be obtained to any other way For the selection of the right word a medical dictionary is indispensable. A good several'dictobary B also used to buic at omt : elbow For inding the shades of meaning of closely related Words I basc found it abges Datlonort of Simons temcy helpful. That book, or a thesaurus, can be used profitably to look up tynonyms Placing the word in the proper place in the sentence that o. the ordering the xosds correctly-os quite as important Strangt top Indeed, an occur iſ attention on paid to the art ment in bach the words appass as its what Goneti oburned (con Brussbeekly publication Tbt odio talento the mettere Gamo pobonaith pots out the House of Contents Sono problems in the use of ronds desene special attention because our rotasional moling show a peculiar duluby ith Uhom ·20 Medical Writing Plurals Singular nouas with plural modifiers or plural verbs are not uncommon in medical papers The most common error of this nature is the use of the plural data as though it were singular "This'data is convincing is so frequently seen that it passes almost unnoticed It should of course, be These data arc Discord ances with phenomenon and phenomena are almost equally com mon. But these are not all Other plural subjects may be gren singular verbs It is necessary therefore, to watch for such errors and correct them wherever found Pronoums Many medical writers appear allergic to the pronoun ! apparently considering it as immodest and resort to almost any device to avoid it True, the reader might balk at too many first person singulars, but when an author is obviously referring to his own work or ideas and to none other he might as well say so When a writer uses we or one to refer obviously to himself alonc he is indulging in an affectation of modesty which scarcely adds to the lucidity or honesty of his presentation Let us by all means have more Is when that is what is meant. The we in contrast to the I tends to be used too much and often without any clear antecedent The 'we believe when written by a single author leaves the reader guessing as to what persons and how many hold that conviction We should be used only when it refers to two or more people who can be readily identified by the reader It offers another problem So convenient and short a word is too handy It often appears after two or three antecedents and the reader is left to guess to which one it refers “It is obvious that it can no longer be necessary to employ this treatment and that other methods are preferable to it is a sample of too much it using Adjectires and Adverbs The use of unnecessary adjectives and adverbs is a problem to practically all writers I do not know why this should be but it is perhaps an unconscious reflection of the authors desire to make his case as strong as possible in almost any medical manuscript one is astonished by the number of times marked very com paratively relatively many and other words of this sort appear ·The Third Revivon 21 Note bou ofica such words an be crossed out murtły with an utual trengthening of the sutikti. Adyenes and adietti cao be used spammety u log in bedral papers, but one put in ure olm romboned arth profit in 8 un Uccptomally weat word and can almost ezys be andeo marid Ba't much better and can wally be deleted or some stronger substitute found. louts as Adjectives The use of Dours B adjectivas cus the purests of language a lot of destross. In scenufc nung serta that common sense is ibe only rule to ſobor Slice usage bort made some Downs employed as adjectno more farmlar and they sound loss sinned than the correct adjective. She cancti for exemple, sounds better to more familor than dermatologic cancer heart disease 13 more often used and is probably as correct today u cardar disease. On the other hand, weter dilatatos Tward us well 13 incorttct and welttal bould be used. An interesung series of letters on this subyert appeared in ST sanctal youts ep The discussuon centered largely around whether auch a len us horse serum is good English. Frank C. Carins, it seems to me, but the mil on the heard ben he wrote Mo Ope bleted with borxe could alitume mange 200t bo did the please trout. borse-lugh. And a warm od turned frod a horse » equipe, ont obamed Iron a donary could op te bron It would be mostalc, bonge to conclude that it tuales so deference bethet a noun is used to modify DOUT OF DOL Probe ably los persible to be SONOS 85 modifiers only ben sunable a Iyectos de Bavailak orben usare has established the noun 3) firmly that the proper adjecthe sounds queer Simused and imod lords Albuth Fishben and Henat bele all compiled lists of words hxcha wedical niers commonly use sonety of woo often. Some are porusuhart, objectionable Among these is the use of fact ben the wtbor mans contort of theory Arorbet is symt alt o Temalt bey deny to mea or xoten mber than imah Case for paurat. Dort for negative, 200 popular for videty used ut others bich leap quickly to mind. Indeed, that are many others which slip by usess the adthor is on the Toolout and diminates theul ·20 Medical Writing Plurals Singular nouns with plural modifiers or plural verbs arc not uncommon in medical papers The most common error of this nature is the use of the plural data as though it here singular *This data is convincing is so frequently seen that it passes almost unnoticed It should, of course, be These data are Discord ances with phenomenon and phenomena are almost equally com mon But these are not all Other plural subjects may be uien singular verbs It is necessary therefore, to watch for such crrors and correct them wherever found Pronouns Many medical writers appear allergic to the pronoun apparently considering it as immodest and resort to almost any device to avoid it True, the reader might ball at too many first person singulars, but when an author is obviously referring to his own work or ideas and to none other he might as well say so When a writer uses 'we or one to refer obviously to himself alone he is indulging in an affectation of modesty which scarcely adds to the lucidity or honesty of his presentauon Let us by all means have more I's when that is what is meant. The ive in contrast to the I tends to be used too much and often without any clear antecedent The 'we believe when witten by a single author leaves the reader guessing as 10 what persons and how many hold that conviction We should be used only when it refers to two or more people who can be readily identified by the reader It offers another problem So convenient and short a word is too handy It ofien appears after two or three antecedents and the reader is left to guess to which one it refers. It is obvious that it can no longer be necessary to employ this treatment and that other methods are preferable to it is a sampic of 100 much si Using Adjectires and Adrerhs The use of unnecessary adjectises and adverbs is a problem to practically all writers I do not know why this should be but it is perhaps an unconscious reflection of the authors desire to male his ase as strong as possible in almost any medral manuscript one is astonished by the number of times marked CT) pantively relatively many and other words of thus son appear com ·The Third Rorldon 23 AB medical noxucluse is a dynamk and changing thing li cannot and should not be ſtoren, because it is inescapabły tmed to adiants la the state of medicine, netheless there is a real dvantage when aldical men 19 One country an understand bat those in mother are talking about. There is, I thank, to principles which should be observed *s closely as possible first, that the technical terms employed in scientific papers be those which are understood by all workers in the parruar field as referrior to the same thing in other words there should be bo possibility of one person meaning one thing ind one another Second, in those fields so hxh attention has been arren to standardizing the technical Domenclaturs, the stand and should be followed cien at the expense of personat napremenu Examples of standardization are the BA Anatomical Tomerdature The Standard look iclare of Direcke and The Internal List of Courses of Drosh Eportims The use of personal famts as Labels for diseases, syndrome ceses, or symptoms is COTTON medical ting. Murty advances have come through mdadual effort and it is therefore not sorria print that the mome of those who have made original contribubons should be attached to the discontries. There are many objections, bone to the use of popytus in medical peper. The wane form of hyperthyroidism variously called Grimas disease, Parry disease 109 Bosedon a disease, and Flajan's disease depending on national prejudice. There are a Potts fructure and Pori desease. There are thosAmds of Pacts distance, cach called allet a differem man sith the same last name. Eponymus are confusing and onihforming about the nature of the process or lose to hact they are attached. Their hustoncu vahe may be learned balio in other ways and it would be wise to cumulate bene from published articles etsept so those for manets forbica a desenptne lym doos met just Abbreviatora Many pedical Danuscripts #hen they are submitted to the ednot contaro abbtes tons which are hard to deciphes Nether dito DOC teado sbould have to be a coptographer to find out What the water stalling about yet that sometimes happens. Fen readers would quxth Neoty pochi abbirallons 25 MT PPD VED of Pukes they are using them tbora- ·22 Medical Wning It is almost equally difficult to get rid of stoch words which tend to be repeated over and over again that is certainly one word which scientific writers commonly use too often He stated that he told his physician that he had never had that trouble before is typical of the situation which can be encountered unless care is used with the thats Split Infinitives As I sec it the rule against splitting infinitives is an arbitrary one which has been fought for so successfully by grammarans that the author departs from it at his peril It is a soſe rule, ond to avoid criticism it is easier to obey than to defy But perhaps it has received more attention than it deserves there are few literary craftsmen in whose writings onc cannot find examples in which the rule has been broken For most of us, however it is better to conform Circunilocution Medical authors are prone to the use of circumlocutory expressions which I believe make them feel as though they had not committed themselves quite so much as the direct statement. In the case of 'with regard to It goes without saying (then why Say it?), in the final analysis, it remains to comment on and similar opening expressions can usually be omitted or some single word employed to advantage It an be pointed out it must be emphasized that, and it is noteworthy that appear with amazing frequency in medical papers may their inbc decrease. In the majority of instances instead of most, and in some cases instead of the simple some have become part of our way of thinking. The author must make a conscious effort to free his wning of such circumlocutions Scientific Terminologi The medical literature of all countries is plagued by confusion in technkal terminology Latin Greek French English German Spanish Arabic and coined names are used by medical wrilets and frequently mired up with each other so much that it is hard to know what the author is talking about Typical examples, such as the many terms used to describe clinical vancties of tuberculosis or arthritis are not far to seck Using lues in place of syphilis series no purpose other than the possible confusion of a felt potients and not many of that ·CHAPTER 12. THE BEGINNING AND THE END Esery medical waiter should try to recall bow he selects his own readmore. Since the son of medical literature is not so rast, no one anted everything. Most of us book first at the ink, then the fint les sentences and finally at the samimary before deding whaber the abole peper bs worth our um If the water wishes to entice Avote rader, therefore, he can devote his final reynsion to these ste portions of the manuscript. Titling The inle of a medical article ought to present the subject in a POKUTILE and duracine a form sposable without attempting to make or it an abstract. Sometisies uites are short enough but maningless. One such appeared in a landmg medical Journal smoply as "Clinical Experience. This leaves the rader completely mibe darkod the indexet vil bare to read the urticle to kom bere to classify Il. A ulle such as Headaches would be nearly us bod for & penodal urucle slacc obryforshy the hole subject can not be convered in a few pages. More often the opposite Tault is present od the inless too long and cumbersome Examples appar on every band the follow mes hot aucti sore than many others HERNERY SYNDROVE (PROGERLA OF THE ADULT) AND ROTH VUNDU SYNDROME TWO TYPES OF CLOSELY RELATED HEREDOFAMILIAL TROPHIC DERMATOSES WITH JUVE YUL CATARACTS AND ENDOCRINE FEATURES A CRITICAL STUDY WITH FYE NEW CASES From the standpoint of the reader the min titk should be short enough to be printed in one or two lons in the journal to hat the trucks seal Sometimes the author us able to daim that he cannor pe un adequate description of the subject disa cussol (and this all that 1 uile need attempt) in such a short 4race. this the case ko say use 1 subhead undemath the 75 ·24 Medical Wnimg selves It is even more difficult to translate MTR Into Meinicic flocculation reaction or PcB into near point of convergence Real errors, in fact, an result from such use of abbreviations For examplc E.coli can be interpreted either as Endamocba coli or as Escherichia coli U C. can refer to urca clearance or urethral catheterization and quite likely other things Not long ago I picked up a medical history which said simply, "SOB climbing stairs." It is well that physicians recognize shortness of breath when they see it but the patient might find some cause to question his doctors fnendship! Abbreviations lend themselves to dangerous mistakes and are unfair to the reader or manuscript editor in medical papers their use should be kept down and they are excusable only when their meaning has been explained the first time they appear so that there can be no possible cause for confusion and they are justified by the space saved. Punctuation The stops of writing are for the purpose of making the reading casier They deserve more thought and study than can be gien here, but there are many sources for this information In general medical writers do not overdo the dash or the parenthesis but they tend to use the comma too much They often forget the value of the semicolon and colon both of which may serve at times better than the penod or comma In choosing and placing stops there 18 no substitute for reading aloud what has been written the stops should fit with the rhythm of respiration as well as with the sense of what is said Spelling In these days when secretaries and typists are taught to spell and doctors are not it is by no means rare to find manuscripts which contain a considerable number of misspelled words When a physician writing in some field with which he is intimately familiar submits hus paper to a medical editor with words in hus own specialty misspelled the editor is inclined to wonder whether the other parts of the paper can be relied on of the writer is so ardless in this respect Some argue that spelling is not important, but the fact remains that on borderlinc papers the editor may reach a decision to accept or reject because of such an unimportant point. Poor spelling in a fineshed manuscript is a sign of carelessness and 13 a responsibility which the author cannot avoid ·CHAPTER UT THE BEGINNING AND THE END Every medical Writer should try to recall how the sclerus huons readine. Since the form of medical litenfurt row so rast, no one anted at thing. Most of us look for at the titte, then the first la sentences, and finally at the swanary before deciding whether the boke paper B korth our ture. If ibe TIUT ashes to entice more rodes, therefors, be an den ote bus fimltton to these ihree portions of the manuscript Thing The title of a medical aucle ought to present the subject in as acurate and attracte a form as possible without attempting to make of u in absintet. Sometimes tiks are short couch but maanlokkess. One such appeared in a leading medical Journal umply a Clinical Experience. This lease the reader completely in the darked the indeter i han to read the article to know bere to classify IL A DK such as Headaches would be nearly as med for a periodical auch hier otrously the whole subject cantot be coered in a few pages More often the opposite fauh present and the title is 100 long and coobersome Examples appear on try hand the follon ME 13 Dol Ducborke than many others HERNERS SYNDROVE FROGERIA OF THE ADULT) AND ROTHUUDS SYNDROME TO TYPES OF CLOSELY RELATED HEREDOFAMILIAL ATROPHIC DERMATOSES TTH JUNE VILE CATARACIS AND ENDOCRINE FEATURES A CRITICAL STUDY WITH FIVE NEH CASES From the Handpoint of the reader the mam tle should be sbort etworph 10 be printed in one or two lines in the journal 10 bist the and sent Sometimes the author is able to claim that be cannot fac un acquait description of the subira do wa (and this is all that a trike sed utemp) in such a short pract If this is the case he may use a subhead underneath the 25 ·24 Medical Writmg selves It is even more difficult to translate MTR into Meinicke flocculation reaction or PCB into ncar point of convergence, Real crrors in fact can result from such use of abbreviations For example, E.coli can be interpreted either as Endamocba coli or as Escherichia coli UC can refer to urca clearance or urethral cathetenzation and quite likely other things Not long ago 1 packed up a medical history which said simply SOB climbing stairs. It is well that physicians recognize shortness of breath when they see it but the patient might find some cause to question his doctors friendship Abbreviations lend themselves to dangerous mistahes and are unfair to the reader or manuscript editor in medical papers their use should be kept down and they are excusable only when their meaning has been explained the first time they appear so that there can be no possible cause for confusion and they are justified by the space saved. Punctuation The stops of writing are for the purpose of making the reading casic They deserve more thought and study than can be given here, but there are many sources for this information In general medical waters do not overdo the dash or the parenthesis, but they tend to use the comma too much They often forget the value of the semicolon and colon both of which may serve at times better than the penod or comma In choosing and placing stops there is no substitute for reading aloud what has been wntten the stops should fit with the rhythm of respirauon as well as with the sense of what is said Spolling In these days when secretanes and typusts are taught to spell and doctors are not it is by no means rare to find manuscripts which contain a considerable number of misspelled words When a physician writing in some field with which he is intimately familiar submits his paper to a medical editor with words in his own specialty misspelled, the editor is inclined to wonder whether the other parts of the paper can be relied on if the writer is 50 careless in this respect Some argue that spelling is not important but the fact remains that on borderline papers the editor may Teach a decision to accept or reject because of such an unimportant point Poor spelling in a finished manuscript is A sign of arelessness and is a responsibility which the author annot aloid ·CILATTER 17. THE BEGINNING AND THE END Every medical inter should in to malt how he selects het onn reading. Since the flow of medical bietsmurt is so ist, no ope cand trythin. Most of us look first at the title iben the first seatences, and finally at the summary before deciding bether the whole rapes is north out toma If the wnies wibos 1o Cole mort trackers, therefore, ht can do ott his final reision to these three portions of the manuscrip Titting The title of a medical tricke ought to present the subject in us Baruri edd attractne z form as possible without attempting to male of it an abstract Somaihi uitles are short coueb bu mezongless. One such appeared in a leading medical journal Honly as Climxal Experience. This lesses the reader completely on the dark-and be indevet will have to read the articks to Inom here to sassify SL A utle spet us Hesdaches would be reasty as tad for a periodical anxk une obr sously the bolt subject cannot be covered in a forum More often the opposite lault is present and the ulle is 100 long and cumbersome Examples appear on every hand the follox ing a got much worse than any others HERVERS SYNDROJE (PROGERIA OF THE ADULT) AYD ROTKVUNDY SYNDROVE TWO TYPES OF CLOSELY RELATED HELEDOFAMILIAL ATROTHIC DER VATOSES ATTH JUNE NLE CATARACTS AND EXDOCRINE TEATURES A CRITICAL STUDY WITH FUE VER CASES From the mandpoint of the reader (hemm title should be shon mough to be misted in one hs to lines in the mountal to Arch tbe anck u kent. Soexmo be author u able to claim Ibat he cannot me an ataque description of the subject dose cussed and thn is all that take seed Himpt) in such a short space If this is the case he may use a subhead undkreath the 25 ·26 Medical Wahing main title which can be set in smaller type and will not interferc with the attractiveness of the paper Three examples of satisfactory uitles follow RESUSCITATION OF THE NEWBORN DIAGNOSIS OF OBSCURE HYPERPARATHYROIDISM PHYSIOLOGICAL EFFECTS OF VAGOTOMY A STUDY OF THREE HUNDRED AND THIRTY-ONE PATTENTS The Opening The first few sentences of an article should intrigue the reader and lead him on. If they do not attract or if they are obscure and confusing the reader will pass to some other article which may not contain as much real worth Timc devoted to the start of the paper is thus well spent An opening sentence like this does not do justice to the writers subject Clinical evidence of pulmonary Involvement in typhoid and the paratyphord revets is well recognized by most textbooks of medicine and the rather common occurrence of signs and symptom of bronchial irritation which is usually referred to as bronchitis, us well u the possibility of complication by actual pneumonio ero pointed out From this confusing statement it is hard to tell what the paper is about and probably at least half of those who would otherwise read it will shy away after this approach A more experienced medical writer recently started out The important and difficult approval of the role of sympathec tomy in the tratment of hypertension and hypertensive heart disease requires all the follow up information that is available. This is a real come-on for the reader as is the following Frigidity is one of the most common problems in pynecology Most writers take time to warm to their subject and consc quently the opening sentences can rarely be left safely as they were originally written They will usually have to be completely rewrit ten Sometimes it seems almost impossible to revise the first pamagraph into an attractive form Try starting with the second or third paragraph it is astonishing how often the onginal beginning can be removed entirely to the bencfit of the paper ·27 The Dopinions and the End Summary ond Conclusions Most medal articles with the exception of those which are murety rates of the literature or case reports, close with a sum mary and sometimes with a section beaded Conclusion A sur man sarcely seens worth lite for the review and if the titling has been done properly it barely desirable for a cake report. For other articles, homoty a well prepared summary greatly belps the busy roader a decading what uidos he visher to peruse in full, ma keeping up with developments outside his immedlatc fields of laterest, or m searching the literature while bulding a bibliography for a paper of bus on Unfortunately many sontaanies are madequate for these purposes, diretly because the autbor has tot preot enough time and effort in condensing the material in the paper Il Bergeraling 20 KCC summary which war out "The cunal data on forty cases of peptic ulty hait been reicned. Thus tells the reader nothing about tbc result of the soney which the only thing he really wants to know What the summary really should present art the most apportant results of the study shorn or blagt und yet adequately describes it tare important reyntre results sbould be included along with the pourtie To wstte a good summary is no easy task and time should be gren to its preparation so thut it mitt truly serve the purpose for which it is intended If the paper closes with section hended Conclusions this should contain ibe author s deductions from the sidesce prestitod 4 the paper only It seems to me that many medical and med #mtets bre loo relactant to draw cochuions from their work. True, they may get borted down later on by some colleague who finds flaws in be deductions of the evidenice, but in the long run progress would be faster if more yestigators, clinical as well as Laboratory showed greater willingness to commit themselvet.

·No text on this page.

·CHAPTER VIL SPECIAL PROBLEMS A medical manscript can be an ongmal stody 1 017 a book, or a case repost. Each of these presents special problems of the cott straction n to be such as sz] mpeal to the reader and espress xarately the thomhts and work of the author Ahhough the uudence to hincte the manuscript addressed most be constantly forordost on the inter s mind, there are frequently features of these 11110th gapers fucta hase a common detsommator and bich may be sobed os foiles TARET Footnolor The question of the desirability of footnotes, and if they are meded the method of ctoployzes them, not asily sohod Foot notes to ofc tuing art extely nostruments for the inclusion mtherapet of multarch the author fools should be available to the reader tun had of placed in the man body of the mant Impoh s not part and parcel of the principal stream of aspiment Of coupht. lodged by such antena i sahrays wist to custa thetical con all footnotes morder to determibe bethes bear inclusyon o necessary 1 2. f necessary then perhaps thty should te restored to an appropriate spot the main body of the paper Certamly nothmes more distracting to the reader than to come to footnote which he wants to cousah but to do to must drop hus go on the page (or turt Keral pages), thus losing the thread of argument. This one excuse for the footport that I can see for the maiusxon of mainal hch ninety nine out of one hundred ftudeni 1 slip but hics of paramount importance to some fa Abdests of the article. At or the use of footdotes should te spanne and a harusorded. Protocols The reader has a neht to Loom Fractly how the ante condoctel bis eypaments Of Obsted hos patients. The description of the unter's procedares and observations should therefore, bo iduded in the protocols of his oratistips. The troubit comes a 29 ·30 Modical Wnimg deciding how detailed the protocols should be There seems to be no arbitrary rule, and the discretion of the author must be cm ployed The procedures and results of experiments need to be described carefully enough so that other investigators' aan duplicate exactly the conditions and can judge accurately whether the results of repeating the studies are comparable. Assuming the validity of this general principic, each protocol of an experiment must be considered individually it should be no longer or more detailed than is necessary to fulfill the requirements. Protocols of clinical observations may be similarly evaluated They should con tain cnough information to enable other observers to know exactly the kind of clinical subject studied, how This was done, whatlesis were made, and the criteria by which the results were measured A great fault in clinical protocols is the use of such terms of improve ment as marked moderate or slight without defining their meaning in unmistakable terms. Perhaps this has held back the progress of clinical investigation by decades. Bibliography It should be assumed that the author of a paper for publication docs extensive reading on the subject of his report but it is not alivays necessary for him to list everything he has read in the bibliography Nevertheless this is commonly donc apparently in the mistaken belief that some professional colleague will jump to the conclusion that the writer is unfamiliar with the literature unless he has listed crerithing which has been written with even the most remote bearing on the subject This current policy is wasteful of time and paper with the exception of comprehensive reviews it is rarely necessary in many iſ not most medical manu scripts it is sufficient to record for publication those references only which are strictly applicable to the development of the waters evidence and ideas If an adequate review of the literature has been published within a few years the bibliography attached to that paper can be cited Then it is necessary to add only those references which have appeared in the interim By following such a policy our medical joumals will become less crowded with long lists of reſ erences many of which are unnecessary duplications and of dubious value to the reader Too many present day bibliographies include references which have not been personally viewed by the writer This undesirable pracuce fails to take account of the fact that various readers obtain vaned meanings from the same source It also leads to misunder ·31 Special Problemu standing and to the perpetuation of errors, some of which aan be Curaordinary The most unusual cumple of this type nu described by Dobail in 1938, and editoriaured in the Journal of the American fedical Association lo lant in para as obsbed by the late De Jaroslav Has the Coop har deze tournal of the Crech Phy XTOS) recording the decay of amebus in ibe Hoot and be tekee! ce of pobres Ruderbug from cotay He also reported suce la dyrenteryo at by the intatal tooculation of tyrement roots. De Hawts professor of pathologic anatomy af Prop Hluk his papar Al written bolly in the Czech Lan park *entiiled o uplaka, Predhitet doktel mantay On Dracatery Precarinary Communition, the origloal boom toonn to post workers through a bremi Carmen det bar the late De Stephanos kartos of Akuadrla. By some extraordinary tule the wtbor i ore (Hans) was entry online and the title the partoprenin in pha Upland, Kartaly In fact relared to the author personally a Uplina and my neptions correspondence with some of this perenaine la the coortents of the Cerrubado hich He paper a aburicted the author manchmaus Uplantel, O Lotne subject dette pune pas Hose Up but in the bor hodet "s the only chtry Uplnd kanot shed sithout any ctplanation. The repilt bus been the one the briremos yan the netborstep of this paper hits hero biped vanoudy o Hova lipsid of Hama), ko Hart, Uplavo to though Wingut H* sforthame)od to D. Hara 10 LALAKI though the ro are synodymo in [act mattont paper both HD Uplavid ut mentioned in though to the sent rhontis bed studied dysentery locatam turly day. feally on the des Cadw of fremd Set Zoology Na is pcr first deed to the root hostblo buto uttnbuted, hoc u Splitu, ko 'Upink, Owo this one i forned by the funker information Of (in part broke) As announced by Dobel, Dr O Uptavia lastly dond The metkulow consulent of the literature an qurte euy pract op less remarkable creoples of misspelled 'nanes, misquoted erthes, and Croneous page Isrags shuts bave been transmited rror onc pathot to atsorbes Setely because of fulutt 10 consult the ongnal if the orloma dettoce is not avavable Otwas not consulted it should be opted or attributed_atatation to the source from which is obtained The method of Une references in a sexstific paper 2000- ant problem. If the shor locus in dinoe hat medial Arnal is point to publish bis paper he should conform_hes takliography to the style of that periodical This merely requires contutung a copy of the popital and following its style of listing ·32 Modical Writing references or obtaining from the periodical in question its style book The listings in the bibliography should be uniform If the author does not submit them in this form, someone else will have to make them so I have frequently been asked about quotations When excepts arc copied word for word from the writings of some other author they should be placed in quotation marks and the exact source acknowledged Articles and books usually have been copynghted by the publisher so that permission to quote-or to reproduce Illustrations-must be obtained by waiting to the publisher and to the author Such permission is generally granted for any legitimate purpose. In many cases instead of direct quotation rewşting or paraphrasing what the other writer has said will serve as well as direct quotation in this event permission need not be obtained for use within limits prescribed by copyright law but the obligation to mentiorrthe source remains. Cosa Reports The case report occupies a peculiar and deserved role among medical articles. Many worthwhile clinical observations do not lend themselves to controlled or scientific study but deserve record ing and can contribute to medical advance. There are two types of case reports. In one the entire paper is built around observations on a single patient. In this type of report the best procedure is to start with a bref review of the literature citing only the most recent and important papers on the subject There is no standard length. What should be included depends on the circumstances Many such reports, however are too long and include a great deal of unnecessary descriptive material which does not add to the presentation Case reports therefore, in most instances should be short and pointed. In the second type several similar cases are presented as part of a longer paper for the purpose of illustration Lengthy descriptions of results of examination should be avoided Only features of the case which are truly significant need to be recorded for publica tion and a simple statement to the effect that certain aspects of the physical examination or laboratory tests revealed nothing abnormal should suffice. The use of the patients real initials or of the hospital case number on the published case report is undesirable. If included these serve no useful purpose other than to raise the possibility that the patient could be identified and therefore subjected to ·33 Speed Problemas ndicak and a filed against the Ruthor Of course, the author must keep a record, but it s bath unaccessary and unwise to include denursing saposts to the patient sideouts Editing and the Editor Medical writers like all obers, are not infrequently annoyed by stanges made in their papers by the editors of the periodical in baxt the article to be published. Certainly the editor should represent the reader taler than the writer and as the readers representalne his first obligatog us to actept only those papers huchi shon rol work and thought and which are presented in an understandable under Nor this in chat a rather rart type of judgment. An editor of a medical petrodal annot and should not reject material umply because it is not If he does, be il be standing in the path of medical adine. On the other hand if he secepe papers presentint nte concepts whxc art sot based on wdequate midence, bewūl be subjecur the readers of the periodxcal 10 barrage of material with simpty will not sand the out of Lume and day actually damnare ibe standing of the periodral and the cause of malade. But this not the only duty of the editor of bich the writer should be ner. The editor of the Journal of Ctrical Imossigen TUDT on tho pobyect, sud to complete Edith the principle that li ha to fonction dictate the myke that a mattor food along to the presentation of the studio, Honna AB our function to take cor Auct or the thica che da tre protect the cut with hich the detection in dened, the audity of the reference, as well nother Guctoritat desert the airportality of a moupe Tyetar of the depot of otrein ny th which door ore remory Bito cum manusipy, but Attitude toward the duni deductom heated by the uppearance of the moon One od te na WCA Rent Hoop coo:fronted Horth might not help booket bate la tas o actorite orla boxe toata at the results of circa etpert mentatoo Once & manipt has beco scepted, what there are the pikers and dutxes and respect to miston on the part of the edhor end of the apiborThis is duicult matter indeed and custom sing indety Most of the popolar manos Otxt tho have purchased an authors artick, feel completely free to alter or moment in any manner the odhton te fit. At the opposite orureme are its editors of many medical and scientuk journals who may ·34 Medical Wnting publish an article, one accepted, word for word as submitted in the obvious expectation that all faults and merits will be assigned to the author rather than to the journal This wide difference of practice indicates that there are no definite rules of the game, in my opinion the editors of medical periodicals should take an intermediate position if a paper is believed by the editors to have ment but to require extensic changes. Before publication the manuscript should be resumed to the author with suggestions of the editor given as specifically as possible, leaving any serious rewriting to the author if this is done the author should not be resentful, as is too often the case, but should realize that the editor is trying to assist him to prepare a more readable and better article. Even though the author may not agree with all the suggestions of the editor he is usually wise to follow them since the editor by reason of his profession is more likely to be familiar with the reaction of readers With manuscripts which are acceptable and which need only minor changes rather than extensive revisions it is probably best for the editor or his assistants to make the necessary corrections, returning to the author not only the proof but the onginal manuscript in many of these cases, sentences or paragraphs which are not clear have been altered Authors justifiably reseni il when such editing changes their meaning. Therefore, the conscientious editor or manuscript editor would do well to call the attention of the author to any rewordings introduced which might possibly change the thought the author wished to convey Finally the medical wnter should not ordinanly object to the minor editorial revisions which are made by the editors of any periodical or book publishing house in order to bring the manuscript into conformity with the particular style adopted by the publisher True the author may not agree with the style adopted but in submitting his manuscript he has tacitly consented to abide by the rules of the periodical or publisher in question Form of Copy and Proofreading The mechanical details connected with the submission of a manuscript for publication and the handling of proofs are not unimportant Failure to know or observe the simple rules required works a hardship on the editor publisher and their stalls and is frequently detrimental to the interests of the aunhor All manuscripts submitted should be in final form and not preliminary drafts. They should be typewritten double space and ·35 Sprool Problerts the oneral not the carbon copy stit in. The ontmg should be on one side of the pope only on ordinary letter head cred paper with ample margins on the sides, top and bottom, Paramaphos should be indented and the pace should be numbered consect thely It may seem strange to have to mention these rules but may editor has seen all of theme broken and other fuilings like dirty finger marks and illepible penciled corrections us well. Since typists are not infallibk, the author should read the final paper after the typing is done, no matter how sck be bs of it by that the. Corrections of the final draft (if the page is not recopred) sbould be entered with the typortiter rather than by Lorehand. Lengthy additions and there should not be many when the paper is this far alone) an be typed on separate sheets and instructions en en us to where they are to be inscried. Rightly or wrongly typesetters and other members of the printing trades are more highly paid than typost Consequently the cost of raisons once the type has been set is grtal More and more medical journals and publishers are therefore passing on at Last part of the cost of revisions made on proofs to the author thr an be quite costly us well as wasteful in time and hard on the doposition Proofs ben tekened should be read cardully and promptly Freures, especially those for donaces, should be checked for Hoxuncy Boetiers bave been toonn to et Changes in the proor should be made of Decessary and of counte, corrections indicated, but not more than required. The author should follow the marling men in the Appendix (p 96) Most type o now ket by machine and an entire linc SCT A ark time. Cosisequently m correcting proof iſ the author can add about as many letters in a boe u he takes out or delete a maty as he puts it. It may scan that only one or two hpa will have to be reset instead of the entire part of the paragraph following the correction Whepca feasible in the addition of material one should try to make such InseTLOOS at the ends of paragnphs rather than within them. There are minor points, perhaps, but ther obsetiace a base a good dal of time and apertise The progress of the book with be cupedited if certam mechan Kal procedures are follona (1) write corrections clearly in pencil in the margin of Ally OstT the tyre itself (2) 101d unnec cun comments or explanations in making cungos (3) amer a questes on the proof itself and do not dangt orignal copy 10 copluma lo picy corrections (1) adopt quities by crotong

·36
Medical Writing
through the question mark only and reject qucrics by CTOSSING
through the whole query
It should be remembered that if the author has seen the
proofs he is responsible for any errors they may contain regardless
of where the crrors may have occurred It is incumbent upon him,
therefore, to give all proofs his close attention

·LAPTER VIIL MAKING THE INDEX Practically all medical tools require in sockex to enable the reader to find nprity and sturately those portions of the book which he whes to consult. Many publishers, including the publisher of this Tofume, prefer to have the indexes or boobs prepared by their own experienced person on though a charge apaint the authors royatxa most be wide. Some uthors, homener prefer to prepart, of at least super se, they own indexes and these notes are made for their ass.CARE or to od than in checking an indet which has been prepared clean bere Good indung is an art. The chief characteristics of the good todevet ure (I) common sense, (2) the ability to do in the needs of those who will use the lader, and (3) mastery of technic. Of these toquements, the mont essentul a good sense Udng it, one 1 sl od such references as these made by an upskilled indeer Chat wont colma tead of t.da Cothis) Types of therapy in malena Stke dolct of, for Yamato brodruct of nokt TEET Obslousty het no point an indering under these alcoorde A safe ruke is to put oneself in the plot of the tiset were LISTOE the book, under what banding would you post to find the Deeded mormaron How to Set About I Get some cards about by 4 inches of the citos are to be handa STILETL tyre me k to be used perſonaled sheets art Concert, or reglas tyranter fire paper may be employed m wuchs minder that enable itay be cut apan in voform H2C for attabeat rearrangement 2 Obune book mu smilar subject and with an inder built opon a plan you have la mund for your book Thrs will help you by out your scheme in adiance 27

·36
Modical Wnimg
through the question mark only and reject queries by crossing
through the whole query
It should be remembered that if the author has seen the
proofs he is responsible for any crrors they may contain regardless
of where the errors may have occurred. It is incumbent upon him
therefore to give all proofs his close attention

·Moking the Index 39 Ascita broocha nd audi, andation tone, 14 bronchial uschors la pulmonary haandness, 199 Cadac asthma, 114, 186 ditavatation from tronctotal asthma, pathogona 113 tramcol. 207 Rather fit doaded to treat Asthma, bronchoat and Asthma, cardiac as the main entries, the Richmonds Brosul asthma and "Cardus asthma should carry only the cross detences, See Agoo, bronchar and See Asthma, cordia respectes Adjectnes ordinanly need sot be axed as catchords unless they have a special connotation which sught be belpful to the Oset of the modes. For exampk, Acute nepbms 200 Hyper Trophix arthinus should be died only wider the second words, but Glossopharyngeal neuralgia and Gonorrheal arthrits should be mdered under both, Anglobal adjectes sboukl generally be changed to their parent boros c... Cerebral tumor should be indeed as Brain, tumor Repal anomalies us Nidot inoma bes, sod Pulmonary abscess as Lung, abscess. All of thest, of course, should ako be entered under the second words Tumor of brain, etc. A cross reference may be made from the aderen forma fits moportance warrants it 5 Enincs generally should consist of nouns (av onding use of other parts of speech) s Bancta, entfauon 3 Soon runon m Buttons nemofying as forn, bon 10 camere 371 6 As chinos are then the bacal amants sbould be tm lated pring a fude post to the matemal on nery termin which might be sought by the reader is Apparators, cotta vonon n. 101 ind Cotra vehom napona pecto, 101 7 Special destes, socht is 1 rence Symbol abbreuison Of u al ore face with appropriate explanatory notes under ·38 Medical Wnimg 3 Take the page proofs of the book to be indexed having made certain that they correspond with the final pagination of the book contain any significant additions or deletions made in the text and show the transposition of any matena) that might have been moved then underscore the topics to be indexed with a colored pencil Such preliminary operation permits the formulation of over all policies before any entries are put on paper 4 Use a separate slip or separate space on regular typenter paper for each item Select a heading that would be likely to occur to thouser and under this heading put any subordinate entry that may be indicated, together with the page number In entenng page numbers, accuracy is rital Keep on using fresh slips, no matter how often the same headings recur Do not waste time in trying to find a previous entry so as to add another folio number The rearrangement of the entries will take care of itself when the slips are alphabetized and edited 5 Include the number of entries necessary to give adequate coverage of the material in the text-no more, no less. 6 Omit the articles i and the when writing the cntnes Capitalize only the first letter and proper names. Special Points 1 In small indexes, cross references should be avoided as much as possible but in a large work in which much information is grouped under well chosen heads it is well to have crossrefer chees from subordinate subjects or synonymous terms Thus Arthritis hypertrophic. See Osteoarthritis, is an acceptable entry when the subentries are grouped under Osteoarthritis 2 Consistency in terminology is of paramount importance If synonyms for medical terms are used in the tert all should be entered in the inder, but pege numbers and subentries should be concentrated under one, with the others cross referenced to this one, Common synonyms not actually mentioned in the text should be entered and referred to the term employed there. 3 Special caution should be crcrcised not to include some subentres in one place and different subentries in another (which seems to be a pitfali common to many authors) or in any way giving in one place part information which may be interpreted as the whole. For example, an inder should not be filled with such entre as ·39 Mokong the index Asthand bronchial ndardec, 67 arola po tme, it Arendanl them to potrosary heart disexxc, 199 Condoc actora, 114, 186 draatation from trooctral tth. Matback 113 trener, 20%. YT Rather is decided to treat Aghmy, bronchial and Asthma. cardmc u the main emines, the catchwords Bronchial asthma zod Carda ushma should carry only the cross refereixes, See Aghma, broochat and See Asthma, cardiac respect cty 4 Adjectnes ordmarily pred pot be used us caichmonds unless they base a special consolabon buch morght be helpful to the We of the under for comple, Arvit pepbrites and Hyper trophic arthus should be indexed on imder the second words bere Glossopharyngeal neunigua and Gonorrheal anhruu should be undered under both. Anatomical adjecinos should geocally be duoted to their part Dows cerebral tump should be ondered u Brim, tumor 'Renal moralies as hdpey adoma Liss, eod Pulmonary abscess as Loop. abscos All of these, of course, should also be entered under the second words et Taxor of braio AC A CTOSS deresct may be made from the adyat al loro of its importance warrants i 3 Entos generally should conset of nouns (nording tase of other parts of speech) as Dana, mtaat, Fenatron, cu senty125 Footy, how to rum, 778 6 Archers are mien the logical samants sbould be formu. biad, puna pude poslo the material on ency ter which mught be soup by the reader as Arya Atcem, cotra para 101 and Com era maters, 101 Spazzal deixes, such as a datase stobol, aber von of a real pre face. With appropriate suplinitorz noces wider ·40 Medical Wing the heading of the index, may be used to denote pages which bear material to which special attention is called, or main discussions of the entry term as Pneumonia, 305-336 areas of consolidation I, 311 mortality in, 325 radiologie findmp in, 318* with an explanation such as Boldlace type indicatos pages bearing special discussion of the topic an asterkak indicates a page with illustration the better followiny Colto indicates tabular material How to Alphabohzo 1 Take the slips or cards and sort them into the twenty 51% letters of the alphabet, each in a separate pile. Now take the letter A and alphabetize the second letter of cach word, placing Aa in one pile, Ab in another And so on The process is continued until the alphabetical order is as accurate as a dictionary 2. Single words constituting an entry with their subentries, should procede longer terms which begin with the same word, and both should precede longer words which begin with the same letters eg. Tooth deciduous, 356 eLamination 350 atraction 166 permanent, 360 Tooth pastes, 380 Toothache, 390 3 Prepositions and conjunctions should be ignored in the alphabetical rearrangement of subsidiary material and the entna alphabetized according to the principal words, as Leukoplakla, 1611 of bladder 2343 differential dugnosis, 3211 syphilitic, 3257 of vulya, 2597 4 When the index material has been carded and alphabetized the next step is to go through the cards or slips from A to Z to make sure of the alphabetical order to climinate duplicates and to combine entries common to various folios. Thus, suppose there are three slips, as follows

·Making the Index
Acrobacter stron, 19
Anotacza wropeor, 41
Anotacto Soto 45
the last two should be anceled and thar foisos tntered on the first
slip
Acrobacter oortoes, 19 4 46
The material may be sent to the publisher at this stage. Typing,
howner is desirable, for a typed manuscript by more converxent
for the printer and itb a arbon copy of the typed sbects the
danger of loss or destruction of the only available copy of the
inder is alminated.
Typing the Index
la typus entries the auctinord should be osoitted on com
Kuine sobenines, and the proper indentions sbould be follomod.
Indentions should be radicaled wluch may be done as follows
Tooth, deatos 3%6
Curation 39
Cucuog, 166
Copromdialom, 167
bemorrhape fotos g. 16
Sachs typed manuscript should be carefully rebecked agains
the marked page proofs co eliminate any possibility of error m
cuines or tag numbers

·No text on this page.

·HYAPTER X ILLUSTRATIONS Joseravons are expensuse to reproduct and therefore #bether photographe prais, drawing or starts the author should make bu sefectons spannely with due separd for tbest teaching potential and probabilny of ſwthful reproduction Good lusinations mbance the value of a book or of an article and no watien irui can tiet be satisfactory subsutute. The praat aims of human an toray dependent largely on alles Iratrons for exposition, and the aphx conitiburons in the Belds of surgery and pathology are Anunk tramples. To be of the greatest value, Ulustrations should be designed with a specific audience in miod. Wustrations intended for the use of physicians and medical sudents are frequently unsuitable for Other spool and we associated groups Emphasis is an important aspect of ioanations. In many Nounatuons there are areas or potats'of ratne first importance which should be stressed (Fir 12 The originals or ustrations are thays prefered for publica con and the author thould not submit photographx coples dhe has the ongmals or can obum their. Top den photographic comes (Fiet 2. 3) and printed reproductops of at the simpler Ainds of carts and diagrams are faulty from the term of all (actory engraving results. The Preparation of Hustrations in General Authort und Muststort may astume that publishers will be Happy to cooperate with themIf information Ananted concerning any aspect of the reproduction and printing of illustrauons, the publisher will gladly supply it In the preparation of illustrations for publication the first consderations are the space available (Fig. 4) 10 krommodate the reproductions and the type of paper which will be used in printing. Thm per stock ind unooited papers Support the best of lint drainp so far as feasible ince beintores (hich reqbirt coala goch) may complicate prinung and binding Line technics

·44
Medical Wnting
Adhorint intustine
Fig 1 Jlustration by Max Brodel showong emphasis of the area of
Interest and relative summpler treatment of the anatomic relationships. (Cullen
Surgery Gynecology and Obstetnc.)
PER CANT or POTI POMOU (HON MORE QUI AT LACH NTERVAL
AM
PM
(44 CU
MR CONT MOHIM
LI:
10-20
ALL NTUL
7 HOUR LTERWLS
Fig. 2 Reproduction from photograph of the original chart FBing in
and spreading used to some atent by intermediate photography are a
dent. (From Neeſen Medical Clinics of North America)

·45 Ilustrations 11 Letra طلا Fr 3 L Locomparing from the original musthoe Riz Reprodaction from a rithet pood photoshopuc ar libero. 7be difference to quality of the lo copiar la colly descended. (Pty permission of Dr Borry I Amon) are streaper to reproduce than tope methods Tone methods kend therosches more readily to realistic representations than line Lechouc. The greater proficiency of the illustrator in the handling of one method over mother may be the factor wikta sbould decide the chance of medical The wze of the orlethal Illustrations is not bughly important chce that in 1 related serves a can connsterscy os desirable for formony of reduction. The sece of the ktienes in all labels honet should be carefully considered from the view of uze het reproduced. Lach of omiformity of printed labels a partir brty noticeable and disturbuon Cardal poupany (Fig. Hof Hasiravons and the use of insens (Fig. 6) le capilme potse and sometimes will utlize pace which others se would be wasted. Photographic material b especraft subke for grouping since frequkotly several PTIAS Oy te adunsted by enterior (cropping) to fit definite spaces about tarm to their value os iustrations (Fig. 7 8) Drawing (which cannot be cropped) ben drawn u spartial figures, are less adaptable to grouping by the publnber than photographs. To use the space on Pe duastageously drawinp should be purpered » pous by the Nagrator f the sothor and the publisher agree that it is advisable to in an overlay of color in order to dutetemuate the areas of op illustration or pabam the soutters shorn, the overlays should be prepared separately in black. The portions of the illustration to

·44
Medical Writing
Adherent ntestina
Fig. 1 Illustration by Max Brodel showing emphasis of the area of
Interest and relative simpler treatment of the anatomic relatronships. (Cullen
Surgery Gynecology and Obstetrics.)
PCA CENT OF MOSTMI VIOLONG (14 OR MONCLU) AT LACH INTORAL
0.0
AM
PM
CUCU
7
PER CENT POST
!:
11:
D-20-M
ALL NTUVAL
HOUT TIALS
FiL 2 Reproduction from photograph of the original chart. Fuhing in
and spreading aused to some extent by intermediate photography are i
dent. (From Ncelem Medical Clinics of North America)

·45 httrafront FILme tips me from the original Postmon. Muu Raydactic from itthor pod photographic cop of the droot. The oitierence in quality of the two copin mops i euly dexternet ID fermenton of DT Larry J Amon.) art dere to reproduce than lone method. Tope methods terud betrische more readily to rulisu seposentations thas lie lothons. The prester proficiency of the illusrator in the handling of on asethod one another may be the factor bxch should decide the choice of medium. The mee of the original Illustratons es pot fertily important CLETA tharm a rebied te certain consistency os desarable for uniformony of reductiot. The site of the lettering in all label, bowelet should be carefully considered from the aim of size ben reproduced the or omformity of panic Labeba u party Carly poikable and discusbus Careful pootus (Fig. 5) of illustrations and the use of users (Frp 6) ut cogni tipetise and sometimes i ue space bachi Octaneould be wasted. Photographxc material is espe- Dy swubke for groupong since frequently sencial prints may teadused by truring (opping) to bi definit spaces thout tarm to ar value as lugn015 (Fig. 7 8. Draung (which annot be cropped) when draxn as separated figures, ure less adorable to grouping by the publisher than photographs. To use the space on a more advantageously dranna should be prepared 35 3 Froup by the distritos If the author and the polishes agree that n nadisable to xe ap oncly of color m order to differentiate the areas of an Illustration or perhaps the uructures show the oietlayı should le pared sparatet, un black The potions of the lastution 10 ·46 Medical Writing Diagonal ling (A through B) for a full posts Vertical line for twice LZC enlargement (reduction 'h) 1 ! B Diogonol jine (A through C) for o half pope Vertical linc for y cnorgement (44 reduction) * 1 Fig. 4 How to scale the size of illustrations to the knowTi proportions of 4 type post or type column (shaded area) A reverse procedure el determine the linear proportions of an illustrapon when it reduced receive a color should be traced on a substantial transparent material or else inked on engraver s proof which has been printed in light blue The latter method requires engraving of the illustra tion first. In making the overlays, the printed tone may be varied by utilizing line technics, mechanical (Fig. 9) and handmade, instead of unbroken black alone. The basic picture must be kept rather light in tone value of the color is to be effective when printed Correlation of illustrations and written text is most desirabic and consequently the author should furnish a copy of his manu script to the illustrator whenever possible Identification of Illustrations To save possible confusion and correspondence, the author should write on the back of each illustration bus name, the figure ·kadratos 3 11, uns CYCATO Och 3, c. Jaceer tot ir better. Ythry non emundtar tra Od rewed ton Anretruz grand por vo will ing dindad me 3 Corre F#. Sarpal procedert dastrados forared froup. by Tom ne menom tum corony to the box offpace. The orpoal of per 10 by 10 nabos, 35 dra fox one to reduction, a tomat urt wit the shop here la tik toston to scoreplay kapoda Apt the other topoal Hyattestorn ** near woot the surgical por described the block the ble under the bencial dos (1) ob utrd the boonother stutto that the actual income (bul Hysterectomy Sopot Pokos, Ith EL, by famon of Ethuron Suturt Laborator, 10) numbet nod tbe subject of the sur le should design the destroy Localsoo in relation locuind refer to each stlunration at the proper point in the tal. Preparaton of Photographs The preparation of photographs for medical publicadas the mon mat the dokal photographer lodeterent pnats too often ·46 - Medical Writing bio Diagonal line (A through B) for a full post Vertical line tor twice LZC enlargoment (reeluction') 1x Diosonol line (A through C) for a half pope 1 1 ? Verticol line for % enlargement (34 reduction) Fic. 4 How to scule the size of Illustrations to the known proportions of a type page or type column (shaded arca) A rescrie procedure wi77 determine the linear proportions of an Olustration when it is reduced. receive a color should be traced on a substantial transparent material or else inked on engravet s proof which has been printed in light blue. The latter method requires engraving of the illustra tion first in making the overlays, the printed tone may be vaned by utilizing line technics mechanical (Fig. 9) and handmade, instead of unbroken black alone The basic picture must be kept rather light in tone value of the color is to be cllective when printed Correlation of illustrations and written text is most desirable and consequently the author should furnish a copy of his monu script to the illustrator whenever possible. Idontification of Illustrations To have possible confusion and correspondence the author should write on the back of each illustration his name, the figure

·49
Ilustrohon
то е
26 PICAS CEMENT FLNIH
Fit 7 Four directly related photographs tried and arranged by the
editor for coppe og oor plan. The refertne konto the Indbu
ructures have to placed and instruction to the probleted
If for the proposes of orientauon, H is ntorssary to ndade
nonostials, the listuing sbould be adrmed in order to cups
the essential futuros.
Enough digance between cimeti rod sobat s mornir
When they are too close tofather the resultant forestalenie IT
arrear as discorion
\pare
Spou must be elminta oribe purus
at
touched. The manied areas should be communa
raint if white on the priot ill provide bere

·48 Medical Writing -Drain po 1'LD Fig. 6 The use of an insert to show 1 secondary step of a surgical procedure Drawing by Leon Schlossberg. (Wharton Gynecology) reflect inexperence with the camera and a lack of appreciation of the purposes and limitations of photography For those with little cxperience the following suggestions are offered The background and the subject should be in sufficient sharp contrast to emphasize the desired features Flat unbroken back grounds are really best. Strong, peripheral shadows on the subject are likely to merge with a black background with a consequent loss of contour of the subject With a brilliant morginal light on the subject against a white background the contour of the subject will be weakened because of a lack of contrast Extrinsic objects and appuntenanas divert attention from the point of interest So far as possible the picture should be limited to the arca of interest

·49
Hustrations
TOP
b
24 PICAS CENT* FLUXH
Fry 7 fou directly related tomopartis artrooped and arranged by the
for kompone pare the reference letters to the men ideal
Mtunt bonne heu pard and wstructions to the mana muca
If for the purposes of onentation, it is necessary to code
noncent wall, the lighting should be adjusted in order to rmphasize
be all latutes
Emuch descance between camera and subject is important
When they are too close together the resultant foreshortmung may
2774 a drton kn.
Trous
Sports must be dimmnain or the priats Mill have to be re
touches The vonanted areas should be covered with orque
paint shute on the il prossde a berler contrast

·50
Medical Writing
1
زمره
C
d
Fig. 8 Same group as Fig. 7 The final cropping and white outlines were
donc by the engraver
h
с
Fig. 94, Two of the many printed patterns of shading film which
may be utilized for purposes of differentiation. The two patterns have been
overlapped, producing a third pattern.

·iustratrants 53 Protts Prots sbs 7 actes lacluior roentgenograms) ure preferred by come publishers swer the sere approximates that of the type mage of the ordinary Journal or book, and the reduction will be slight to accommodate the picture to the column. Glossy surface mpor is dearable. A print should not be mouled It is likely to be combined with other prots or other in work. The author should nodicate nth ut peycol (casily trased) where stranehi (CTOSS and Upand down) crotags feasibte. Sequetxe betiers, babets and any other added designation waris should be inserted a strong contrast to the immedote sur rounding ueblack in light arean white dark areas (Fig. 82 Do cot use colored laks, espeeally blue, which become touch ughter when photographed. Parker clips and plas injure the surface of a print and should be an orded oner curs to be reproduced. Adhesives a removal A neocen) may remone lbe wurface of the preDL A Pen DT soft peal would be used for Writing Donations on the back of a paint waris made urth pressure will raise the surface of the front Photorolcrographo When the publication of large poromicTorraphs is desired the author should drade on the aseatul field m relation to us accommodauon on the page, and albct crop the point accordingly or to face the edgor mark with a var pencil ibe Timus of crop stab Tbc lacter procedure is usually preferabk unless ibe ambor ha pored to a party the stressen reduction Small photomutograph art unikely to reqmre much reduc Hon of hy Berd rone 11 all the permitting the mention of the orginal maguficauen Small prints are used effecuvety in com hinawon ** Large potomctographs or other illustrative mater from the same forcent of water-thal , superonposed on an unimportam vrea Fe 10) The cost of optavine s reduced in thrs 25 The mapfication of botom tographs should not be doug nated on the mists but indicato read in the accompanyie kepenats that may be cometed all the sun'u have been repro ducu ·52 Medical Wnting CM Photomicrograph Gross specimen Fie 10 Ahore Small gross specimen combined with its MKTOSCOPK stru turt. Below The combination scheme and emphasis reversed Roontgenograms Rocntgenograms submitted for publication should be postine prints. The opaquc masses ought to appear as nearly white, and the less dense structures considerably darker (Figs II 12) of all photographic prints, those from roentgenograms orc the most likely to suffer in reproduction Differentiations cuident enough in the films may really be too subtle to sun ise reduction through the engraving process If considerabic reduction of rocnigenogtams is required for page accommodation the greater part of the reduction should be ·53 Austrations Ir 11 The use of tons to butoate the margins ofan va pot defined the roen poopin (Call und Hald) done by the photographer when he makes the prints. More detail slilch, to be retained than when the entire reductions incrdent to mietanę. Also the author a better able to judge the probable value of a print it approuties the size of 143 reproduction The value of roentgenopuns may be increased under some Curcumstances by the addition of labels, aToHS (Fig. 11) or other donations emphasize the structure or arta bich the nuthor deuros to demonstrate Line daftans, used Isette, are useful adjuncts (Fig 121 The retouching of photographs, tray punts and photomuto graphs to clanfy details as considered permissible creept in instances Ibuch Any modification would destroy the sercatific value of the Tuction

·54
Medical Wrong
LESION
R
Fig. 12. Grouping arrangement of roentgenogram, alth amplifyme da
gram, and photomicrostaph from the same case. (Surpical Clocs of North
Απκτη.)
Labels
Just as neat, legible labels add to the attractiveness of illustra
tions carelessly drown and placed labels detract from what may

·55 status B с Fx it la ruel Vocate the area decrease don there to the At the trucat can be con lenttype tot of the Lid l. Thaalbar no&edorind sat 13 alth an et male Buon the longer 10 c thermotector mera narande dnt Shanno e admiraht euant of an noch Labeh should be at the dece 23 Nistradun bene pt. Prepart maina.cnhini ast famanEmA nd halinr art 01:a 100 extes et unemot F think tha sem " In out the latest 3 10 Wibe poenal design. Thus Reyecah troebe em Ubat at goed 10 peber ud aa aatne page theran and tested for labe's (ading ahat) Deated the time ins are frequenth red dif centros The entirch tsent, bonnet the terms call attend them the mommy te Suh mas realizado un punto Lemon Athard dimorei afto the co- Teo Cah mens the contenhle at ease of anne The the nahlala kius arve of snaked on what bruto nostet sbena possit This tipe of deuputon a home and its the resident 21 ecure a the beaut amicurato necusumo

·54
Medical Wnimg
LESION
R
Fig. 12 Grouping urangement of roentgenogram, with amplifymg da
gram, and photom.Topraph from the same ase (Surgical Clinics of North
Americ..)
Labels
Just as ncat legble labels add to the attractiveness of illustra
tions corelessly drug and placed labels detract from what may

·55 Ilustrations -her width TOTE KA A B с 13. Label placernent. More often than not be adth of the rup or column the akerxon of Darry reducuon and, therefore, the size of the pictat be reduced mary te direct ected by the placement of the label (bad) Tbc on ell bocar mboga (doned outlinos) of Amd Bare lot al, although the actual picture (shaded mra) Bb consudenkly the langer Lo C the picture occupies the colic vtth of the type patent a Might montat ok oxber be be adminble cumples of art ork. Labds should be part of the design of an illustration bere et pomible Properly spaced and disenbuted, such factors as symmetry and balance are muuntamed, and sometimes an attained. For the reason some Dustralors by out the labels so # to and the general design. This especzally true when drarn Ulustrations are prouped together und an attracene page the cod in 1 A sandard terminology for labels (including abbre, nations) A preferred once the same illustrations are frequently used in dif ferent publications. To be entirely consistent, bonnet the terms used sbould correspond to those in the stompanying lett. Such Toatiets u CONSIRODI spelling capitalization and punctuation would be reina. A misspelled word, discovered after the entprong is muck, wenlly means the considerable etta eiperise of une plate. The ox of symbol of kry letters in place of spelled out labels should be loded whenever possible. This type of dentations tombersons and requires the reader to depend almost entirely on the legend and other content for oriaution and identification of grudur.

·54
Medical Writing
sery
LESION
R
L
2
Fle 12. Grouping arrangement of roentgenogram, with amplifymy do
ruth, and photomicrograph from the same ase (Suraxal Clians of North
America.)
Labels
Just as neat, leuble labels add to the animactiveness of illustra
tions, carelessly drawn and placed labels detract from what may

·55 Hierdoa 3562 FT 12 Led Mort the other pepe CT etcon the heart that's the at the Trente eh dicted to the photo the barcha al Thabo Soseslodesidio BET entrancatican Broccasenitaries нс Аустасавет Зерттеге таарах ez chestate canadiant Lites and k the dar um bonor posthe Promet astrained summedaalam and 24 se ne med for thurs ttasca come I USTIATI bit co the bhes was to ad the nel desupun tas HEAnta in Tirumsr paped upabar 2o me poes the imm A sors to fr bes (data) opblaas at frequht used me for this. To be the ben the terms od same to be in the art 1EL SX 13 8 mm 2303 od poetry te besteed td. stered ifis the et PM 8.5 mees betere one of 2 p Thesen kellemes 3 podlahos 11 hentai one possible to tre o determinis hexorstbetyds idrottendo te kind nicht met fontos ud xuram

·56
Medical Walang
ABCDEFGHIJKLMN
OPQRSTUVWXYZ
1 2 3 4 5 6 7 8 9 0
abcdefghijklmn
op q r s t u v w x y z
By airplane, New York is only a
short journey from Chicago-,,?
A B C D E F G H I J K L M N
OPQRSTUVWXYZ
1 2 3 4 5 6 7 8 9 0
abcdefghijklmn
opqrstuvw x y z
By airplane New York is only a
short journey from Chicago 1,?
ABCDEFGHIJKLMN
OPQRSTUVWXYZ
1 2 3 4 5 6 7 8 9 0
abcdefghijklmn
opqrstuvwxyz
By airplane New York only.
short journey from Chicago
Fic. 14. Hand lettering by Lucille Cossell Innos in 1 ample style suitable
for labels. The middle and bottom groups are shown reduced one-third
and ond half respectacly
Marginal labels, whether hand lettered set in type or posted
on necessarily increase the size of a picture and therefore may
influence the cost of engraving and the degree or reduction (Fig.
13) Costs often may be lessoned by placing labels and key lettc73

·wdrotrons 57 directly on the illustration and by uilizing marginal Indentations of the Mostraron dhon the m-21 dimensons The syk of letters used in kabels should be umple. For hand- Kettering (Fig. Ithanyle sonen bat emibr to typesenpt is ccellent and pridoned by many illustraton. The product of ibe Kem mechandai ketiermee sets AS ausfactory for nearly all purposes Honder handletteries allons more detibility in ate and place unt. Letis reduced to less than I am un bashit ere barely readable, men in line ilustracions. One millimeter height in halftone reprodactions us too small and there ibc carubasipo should be made when the onpinal illustration is prepared. At tupes to sutvantageous to prepare the labels and leaders separately frota the illustration. When this is donc Accurate teister is assured of the labels and leaders are draan on trans parent materal placed 01 the illustration Separate preparadon fucrates changes on leaders, labels and terminology of as may happen chances are required by the publishes or aubor Another adizotage of the separate preparation is that when the combina lton engraving us made the labels and leaders will be umbrolen by the attend in malmg the ordinary halftone. The disadvantage of the method is the extra DOT110g pense. Ladets or fud fides crord labels to object should be of swt able thickness bea reproduced for the age to follow asily but DOT SO Pronounced as to pubordinate the illustration proper COTty best mugtuned by the use of black in light arras and brte in light area. Broken los disturb the illustration less and art smelo COLSXT to focom than continuous times. Heavy leaders tend to drxde the Ilustration in an underable way Tyre set for babek s duadiantageous because of a ponse and some lack of tributary of placement lindoled border spaces balltore ilustrator ue to be alitird, outline cuts l be required. Partaalt jusufiod (alipad on one side) type in some of the apose of Ktitog la the case of halftove topropy, leader locs noust be clended to the type by ube coptohet The Wnuing of legends Lepends should be bext one she subject of tile of an illustration i de tipuod tough. What should be and depends on the descnpine and destrung lesers, bbels and marks on the picture, and on the deteco 10 11 in the textThe legends must be sumbered in sequence lo Corespond ath the figure oumbers marked ·58 Medical Wnting on the illustrations, most publishers prefer to have them type written together on a separate sheet Permissions to Publish Courtesy demands that the author secure permission to republish an illustration belonging to another author or even one of his own which was onginally reproduced by a different publisher If the illustration was published under copyright, and most illustrations are there is a legal obligation to obtain permission As a general thing, permissions should be requested by the author (rather than by the publisher) and it is best to do this well in advance of the date of expected republication Authors and publishers almost invanably grant permissions and willingly com ply with requests to borron a limited number of illustrations and sometimes engravings as well Publishers, however may be loath to lend engravings which they may need at any time and may offer to furnish clectrotypes at a nominal cost instead Written permissions from patients to publish their photographs should be obtained by the author whenever the photograph reveals the identity of the patient when the patient is a minor written permission should be obtained from the patients parents or guardian as well If the patient is dead, written permission should be obtained from the patient s sun11ang spouse, or if none, the closest relative, and in the case of a minor from the parents or guardian. But if the identity of the patient can be obliterated in the photograph in some way it is unnecessary to obtain permission to publish the picture Usually masking the face on the photograph by painting over the eyes on the negative or covering them on the print is sufficient to conceal identity but in fase instances some other means of rendenng the patient unrecognizable must be devised Copyrights The rights of authors in their published illustrations arc ſully protected by the copyrights obtained by publishers upon applica tion to the Section on Copyright Library of Congress Miscellaneous Information Illustrations which have been drawn on brittle chalk papers should be mounted on heavy pressboard and the surfaces of the pictures protected with cover flaps.

·Austrations 59 Soft mayon or any other medium that is easily smudged should be sprayed with fixatire to make certain that it mll not be disturbed it touched Insurance on a valuable shipment of illustrauons s advnable. Unless others ise directed by the author publishers As a rule return llustrations to the source from which the illustrationt were recehed. However the publishers must keep the illustrations on hand until after all the related editonal work has been completed. Photoongraving Photocograviag is copy of process with the inherent limutan tions of photography whicth, as such, o nethet selectie nor interpretac The satisfactory result of photocograving, therefore, depends entirely on the preparation of the illustration which will be copred Photocagnets are able to work at better advantage of they are spared the encumbrance of wxde margins of paper surrounding the illustrations proper Wide mounts are bothersome Line or Zinc Engravings Pen and ink of libe drawings are the simplest and cheapest 10 engrave and print and are usinally welcomed by the publnbet Line drmniaps reproduce faithfully if property prepared. If they are sufficiently open u to spacing, reduction to a small size is possible Witbout loss of detail. Honger the lines must be black and wide and strong enough to withstand proportionate loss by reduction lines loo close together berge and print as told black. Vanou black and white technics are suitable for line engray mg. Hand-stippling in black mk reproduces well. Lithographx crayon und drybrab (Inked brush, partly drym) on the specia) surfaced papers are satisfactory methods when the reduction is not to be great Droning in lead pencil must be reproduced by the halftove process most instances. Ben Day patterns (named for the inventor Ben Day) applied by the engaver and anos mechankal patterns which may be ated by the illustrator are useful adjuncts to simple line drawing ind charts and often grie udmunbly lo plast of color The Techanical patterns require no aun erroving cott, hike Beni Day adds appreciably to this expense La douings may be combined with tone illustrations and etcted for printing from a single plate. There are two methods ·60 Medical Writing combination plate and double print which are described under Halftone Engravings. Halftone Engravings Halftone engraving us the process used to reproduce photographs roentgenograms and the vanous types of drawings in tone such as wash and brushed crayon As the term halftone implies, the reproduction is without tones of pure white or solid black The process differs from the line engraving in that the illustration is photographed through a ruled glass screen which breaks up the image into tiny dots. The varying size and density of the dots account for gradations of tone (Fig. 15) Because of these dous, an engraving made from a printed reproduction in halſtone is unlikely to be wholly satisfactory Square finish (Figs & 8 12) and outline finish are the two principal types of halftone engravings The former is used far more extensively because it is less costly Square finish is described as a halftone with an unbroken screen surface, finished in rec tangular shape with or without a border line outline finish, as a halftone with the removal of all background surrounding or showing through or within the outline of an illustration Outline cuts emphasize the illustration because, when printed the illustration is surrounded by white paper However in case of marginal labels and leader lines on the lustration the routing away of the screen is not feasible and if a white background is wanted, an outline combination plare (Fig 16) is required With this method the illustration and the line work (labels etc) are photographed separately and the two film images recombined on one plate. The copper plate is then engraved at suitable depths for both halftone and solid black Double print is also a combined line and halftone, although unlike combination plate, one film image is not necessantly imposed above the other the two may occupy separate portions on the plate. When superimposition is required the line and halfonc illustrations are prepared separately usually the fine part on a transparent overlay so that exact registration with the halftone is possible. Illustrations in Color Illustrations requiring color are so expensive to reproduct (see table page 62) that the publishes may feel obliged to ask the author to share the cost of publishing them

·Illustrations
61
2
F 15 Dot fotoration de battooc mpning The lotonedinte pac
ture o reproduced in upronin dy the stone in u the original photo
pripta

·62 Medical Writing Cutis- M flexor hallucis brevis M abductor hallucis Aponourosis plantaris -Tubor calcanor 76. 16. Outline combination hallonc reproduction from a transparent wash drawing by Mary Diron. (By permission of Dr Barry J Anson) Table of Erroring Costs including the Proof Sheets (Philadelphia 1957) S 742 23344 9.87 LA " X 4 Line Engraving + One Color Oerlay Hanone Square Cut Outine Cut Combination Platc (ascrape) Full-Color Process Plates (color prints or paintm) (from hodachrome film) 40 0300 112.00 ·Mutrotrons 63 Fall-color illustrations must be reproduced by the four-color Mocessbuch entails bany apeose in all of the processes of Fit 17 loe portag showtang yaylate tose other Day patterns hich place ca the plate by the copate The lotermediate lopes of Aty prontok trocbcat contrast ma lo this one at und instead of colors. The cup cost would hang berukund the trutt otoboed had Rrovar patterns been placed on the ordenal Had od butc dry by the Parottor Deba De Bem Amon) publication and authors should not submit them *thout first consulting the publisher Before employing color in em form ibc author and the whestrator should explore the possibilties of a black upd bute subfutute Bos-Day-type poitetas (Fig 17), mechan MAA Fit 11 Various tie of bars, coatuous and to custod. which the ondoln the partion of curt mi parte ha la Mch brod dr feruntament auton calines primed to color kcal or bandmade, and sarved ines (Fig 18) will often odku four differentiation to srie in price of color Mastrations in ·64 Medical Wrang color including hodachrome in some instances retain illustrative value when they are reproduced as halftones. The engraving cost only partly reflects the expense incident to the reproduction of color illustrations in books and periodicals, since such factors as pnnting, binding and location of the color pages are involved. The expense of a single reproduction in color might run to 100 times that of a line cut ·HAPTER STATISTICS EMPLOYED IN MEDICAL PAPERS It is hazardous to enter tightly into the preparation of a medical paper which presents charcal or bboratory data by means of figures-bt statistical method. Yet this is done mery day anh perfect confidence, and the tiles b (requently astonished by the fact that his results and conclus Are By Hocepted. The subject complicated there are xicral merlent books available (p 93) and that can be tiemprod bete i 10 sketch a fem problems to comader ben *TIUNE 1 satiscal paper and call attention to onc the common failings in medical anchos bich Interiere with their effcctctcss. Ain En Uring Statistics What the um in ang sarsues as a means of presedime modal dan Thos bas been defined by Yuk as the use of methods especially adapted 10 the decidation of quantnatnc data affected by malplicity of Causes Put more concretely suppose OTC *ere dahnt "nh the problem of recovery from Lobar pneumonia nach dem ser, Hate of outhlion type of pneumococcus, promphxc rodenice, and perhaps sa stal other factors might be contacted to find a possible infectee on the outcome. It is in a Rady of this sort that the statistical method may be employed profitably eke the multiple souble rooms can be summarizach analyzed and described clasty by the means. If tbc banc madres tre adecure the conclusions can be apressed more acurately by staustal means than in ury other way The first problem, then, to decide whether the cmical or Laboratory dara hixhi ue ip be puhered for a particular study and paper an and stoold be treated by statistical methods. It bs hoped that the absequent decisions in this chapter may help the Ther to each the decision.

·66 Medical Wormg Collection of Data Accuracy and Significance The observations on which the statistical structure is built must be accurate or the paper will be like a skyscraper resting on a foundation of sand Reputable investigators acver are consciously inaccurate but powers of observation are fallble and it is wise, therefore, to let some impartial observer in addition to the author go over the basic data This is a safeguard rather than a reflection on the powers of observation of the writer or on his integnty A common problem particularly in clinical papers, is to judge the possible significance of a statistical study in advance. By this I do not mean significance in a statistical sense, which will be discussed later but rather whether conclusions drawn from the use of statutical methods will really mean much on the basis of the data collected If one were studying the effect of some drug on the duration of headaches for example it might be possible, even on a fairly large series to prove statistically that the drug shortened the duration But in dealing with a disorder like this in which it is usually most difficult to statc cxact times of onset and cessationand entirely subjectively at that the statistical proof" might not mean anything It is worthwhile, therefore, in planning a paper which will use statistics to consider carly and in advance whether the figures, when they are obtained, can actually be used to draw conclusions which are really meaningful Sampling Any series of observations, whether made on guinea pigs gastric surce, or patients is a sample of a whole it is never possible to soclude all of the patients, gastric jukes or guinea pigs in the world in any set of data. But the purpose of sampling is to enable the investigator to make deductions which can be generally Appued consequently the aim of this method should be to select A sample which will be truly representative of the whole group c.g. all patients of a particular kind all gastric juices in peptic ulcer all guinea pigs sunilarly constituted This purpose, and the need for a representative sample and not just any old collection of data, is too often overlooked Every sample 13 selected it is hon this selection is made which determines whether or not the sample can be considered to shed light on a larger group Many of our hospital statistics are selected in such a way that the information they supply cannot be said to represent ·Stotists Exployed in Medical Papers 67 E proppy most of the patients may come from a single ecocomklad, the hospital may admit 1 high proportion of one set or age group often odvidual judgment rather than chance determines the choice of treatment, and many other factors ar Lely to enter Lato the picture proienting the sample from being truly representatne of the coure comunity THE PROTLEY OF Bius. The biggest problem in medical ampling as the introduction of brass. Since Ficty sample selected, the method of selection is often determined not only by the use of petunt data (perhaps from a sorte bich us not representaine) bunalo by the human lector of tending to select a sample which is Elely to support the cogiketoons of the author on this matter Pearl sud l' personal gameatstorbato mond to be isdom ore Boned to play any parto the choosing of ladicoabs to po unto the prople the result is actually certam to be based. But in the cootrary rore purely mechanical method of pichborg the Indiab -mony and fittedly told that genriag judios of hub pood md wste to boosiert unh #s operation, the bride ape to be free of bus, proved the mechanical method was aurtfoy planned la o relative to rudomicity Cor der hon ohmibet papers in the medical literature fal in this respect' Almost all follow up studies of the results of some particular operation art, for cuample, brused the potenus oper sted on wete self-schetted in the first place because they presented themschic or CTC referred to the particular surgeon--surely not a random sample here. Then those bo were operated on WCTC segtarted from those who were not, wally because the sureTON expected better this in the formet poup But tha ann 25 selection on the bars of age, som erity of the disease of some other factori. Finally the simple wu influenced by the judgment and skil of tbe sur pon. the enerbete, and buch partkular patients returned for the ponoperatat ealustoor Suct studks may not be valveless, but als clear that they do not fuil the criters oui Iunct my Part od the results cannot therefore be extended to ! palment. The okree to bych spplication in large groups is justified depends on the size of the simple, but not applicable al al miss the sample is truly randon QLTTURES. One method of collating samples shah has hecome aceedingly comition toong Nons of the use of que toapautes. The amnes moy dangers and the results require the most cardul lud of interpretation Hom are the individuals ·68 Medical Wning selected to whom the questionnaires are sent? Are they chosen by some mechanical means which avoids bias? How is the question naire worded? Will the questions tend to produce more answers favorable to the preconceived ideas of the author or will it be equally casy to record unfavorable replies? And it is exceedingly rare for all questionnaires to be answered which patients do not answer those who have died those who are dissatisfied those who are old or what are the reasons? In other words, the sample obtained is not even the whole sample and if the reasons for not replying in the missing group were known the conclusions might be quite different from those actually arrived at In general, there fore, the samples obtained from questionnaires to patients are not random samples and the results must be interpreted with caution Comparabilir In numerous papers involving statistical methods of handling clinical or laboratory data or vital statistics two or more char acteristics are compared or matched It is wise to make sure that such characteristics are basically comparable and this involves critical examination of the meaning of the bask data before they are subjected to statistical calculations Jo July 1935 the United States Bureau of the Census compiled a set of 1032 death certificates cach recording from tho 10 five causes of death together with other relevant information Eightech countries were asked to classify these deaths according to selection principles then current Wide differences in classification occurred one country allocated only 5 l per cent of the deaths to respiratory diseases whereas another classified 23 5 per cent in this calcgory although the two sets of death certificates were identical This situation has now been somewhat improved by wider use of the International List of Causes of Death but the results show how dangerous it is to compare international vital statistics In spite of such knowledge, even professional sutustmans are sometimes guilty of citing as evidence figures which are obviously not comparable The fact that the crude death rate is somewhat lower in Australia and New Zealand than in the Unned States has been advanced as a reason for turning to the medical care program of the former countna. But such figures are not comparable The age distribution of the populations are not alrke the climate and dict are different Australia and New Zealand do not have the Large Negro population of the United States (and it is well known that the life expectancy of the latter is lower than that of the white ·69 Stotistos Employed in Metical Popers mhabitants). Indeed, there are many other factors such as Domy status and wail runzl teenet all of which may play aran in diretta betucce the code depth rates. Those who used this uzloment should hanc lnonn better The same principles apply to final or Laboratory data as to uttal statsua. In the land of andy hech kids itself to the use of controls, comparabibiy can be obuwned by selecting amilar groups for companson, c, perrenta) ammals of equivalent aps. x. and tight sinxtly alternate selection of patients for a non trot ment (and 10 best the results will bave to be seruiunied bier for chance vmations in factors file age ich may affect the comparability of the groups) The problemen may be the bote dificult. Take a hypothetical problem which a new drug for pepti ploer has become available and the movedgator wishes color whether to give up the method bebas been previously following and sich all his parents to the BC drug. He has two borces other to eat a sufficiently brec number of parents chosen by sixt alteration by the old zod nem methods, or to place huis en patients on the Dew mediant and compare the results woh a omlar group of patients he has trased m tbc post with the ol' method. The former a better but the barter o commonly employed. However the un estigator bas 10 be austful that the no poups are smally comparable Art the eco- DOTIK conditions alike' Are the signs and symptoms equally senare the two groups. Besides this one change in medicine horse there been many other changes in the patute of the proups or in the management of the Dnese' All these things must be wetghed before the proups can be condered comparabk and was condu on to be dry1 ts to the relatne sfocacy of the two drust Anatysk of Data Harmot Ang some seriou thoopht to the colection of a KTO of observations ad finally being supplied with a set of bruts, the subor i nett fand sib uninging and analyze do obsentons The figures must be suficant in the statal sense so that the condoskops pot be blonn and le a house of cards by some reader who maintains that the results are meanlooks because the myrtet has failed to malyze the figures properly The walyses Deed to be carried larther for some studies 1520 others, and of course the corucalor babosatory worker bo las retenida int ma tistical methods is better off than most. But there are soos ekmentary aspects of statistical analys's ** hact ·70 Medical Writing every physhian and investigator should be familiar and although employing certain methods will not guarantee that the writer s data will withstand criticism it will at least keep him from falling into the worst errors and give him an idea as to when it is necessary to call for expert help Central Tendenci-drerage Median and Mode The most common calculation presented in medical or biological reports is that of the arithmetic mean of a series of observa tions Among medical workers this is usually spoken of as the average professional statisticians employ the same method as one of several devices for determining the central tendency of the observed data The arithmetic mean or average is simple to calculate it is expressed as the sum or total of the individual observations divided by the number of observauons A simple cxample which constantly recurs in medical papers is age in any study of patients it is almost routinc to record ages and express the average of all of those observed. Suppose that 20 patients were involved ranging in age from 13 to 78 years. By adding all the ages and dividing by 20 the grerage age may be found to be 54 years as shown in the following section Usually the average is the most desirable method of expressing the central point or tendency in a series but sometimes the niedian is calculated which us that number exactly dividing the observations so that an cqual number fal above and below that figure, The nade is another measure of central tendency It is that number which occurs among the observed figures more often than any other Variability Standard Dertation The individual observations of a series will not all be alike in the vast majority of instances. Vanability cxists in all types of observations whether they be levels of hydrochloric acid in the gastric juice, weights of rats, or leucocyte counts for thus reason several statistical measures have been devised for expressing the degree of scatter or venation from the mean of these, the standard deviation is one of the most useful and us therefore usually applied by the statistician It is a method of calculating and express ing how far individual observations doorge from the average either above or below Consequently a small standard deviation shows that most of the observations are clustered farly closely around the anthmetic mean, whereas a large standard deviation ·Statistics Employed on Medical Papers 77 Indicues that there a wide gap between the average and many of tbc loridual obseritrons. Without Folf mo the mathematical background a suple *sy to press the standard dention se foloss a squares of dit from Dreen maal momta a obrany 100% This is a genre of the scatter or variation from the mean. Tale the a suston m the hyporhetical report mentroned under the procedit Keton NON HTUS DATA Nova -12 SLARI 0 HEATION 14 676 +10 100 7 3 6) 64 35 +11 +10 - 19 576 16 191 100 361 60 10 +13 63 71 196 +14 +17 +1 1 35 3,86 Sundard 100 - 161 What conclusions an be dir*n from this upplikation of the standard de won' for it u krpe destation and therefore thers that the Indnduages and nuther for from the alcafe of 4 Second on one of the asedimen are hue ortade the to standard de avons on eher sck of the man (21 8 to $6.2 jaars) whxc matektin moght consider Panificant therefore in the particular bypothetical stady we stems to have Imk or no ·72 Modical Writmg significance in regard to any other attnbute which might be the subject of this study There are two principal aspects of standard deviation the calculation and the interpretation. In small senes, such as the one just presented and the one following, it u labonous though not too difficult to arrive at the mean list the deviations from the mean, and square the latter with pencil and paper, in large series, which are less common in clinical and laboratory studies a slide rule is almost essential. One reason for squaring the deviations is to eliminate the minus signs The interpretation depends on the nature of the study as well as on the absolute figure of standard deviation arrived at For example the following hypothetical study deals with a single type of measurement on tuo small groups for which a comparson is desired IVelehu of 20 Boys and 20 Girls Aged 14 SQUARE -17 +6 +5 +10 WT OF DÖVLATION SQUARE WT OF DEVIATION DOYI (LD.) YROS MEAN OTRL* (LD) FROU MEAN 95 121 104 108 +2 4 116 0 104 4 99 112 36 119 +3 92 196 121 98 -8 64 126 121 +15 275 109 -7 132 +26 676 133 &8 -18 324 109 +3 9 103 116 +10 100 112 105 114 100 -6 36 123 118 +12 144 122 99 -7 49 103 -13 119 +13 169 117 +1 90 16 256 121 +5 103 9 128 +1 107 Its 101 4 119 TOTALS 2120 3428 2120 Alcon (boys) 106 Mcan (pris) 116 2428 Standard deviation (boyr) - 121 4 - 11 20 289 9 25 100 49 289 25 64 16 4 49 36 169 1 さす!!!! 25 144 1 1 +3 1+48 Sundard deviation (is) V72.4 - ·Statahes Comployed in Modical Papers 73 From these data od caktlations sentral deductions oold be possible, awning that the sunples are random and truly representatne of the abo's group (1) the mean of nitragt weight of boys at the age of fourttes ss less than that of arts of the same are and (2) the waght of the boss tend to vzry more widely from the mean than that of gris os sborn by the Larre standard dermaton of the former The comparson between the means iad sandard de bains would not prove that boys are less bealthy thain arts of this age because they tupbed legs and some boys are Canber away from the average than girls the average may not in ether instance be the deal opht for good health. In other words 00€ must be autous n driving dominanted conclusions from this or any other statistical akubtion. Coeficknl of lartation Io order to compare the variabilitses of different frequency distributions another measure for this purpose bas been dessed termed the corffichent of worlation. This is expressed in perontages Coccicat of attoo Standard duration Mtan X 100 Thus to the emple of ages promously ated the coefficient of varaton ould be akulit as follows 162 Cocent of raton x 100 29 per cent If the mean age or bad been 25 mistead of 54 and the sandard des on the same the cocfficent would have been dillerat 16 1 * * 100 - 64 4 per cent A high percentage for the coeficient of antion indicates that barre mur ber of the indical observations are a long way from the unthetic dient In this alubion the original unit of measurement k cincelad and becomo tamportant because it present in both numerator and denominator Thos malos corpersons insolving differeat aarts of Ixasure possible. From this cursory discussion it should be clear that it dangerous to think of terugs or means us presenting a complete maurt, and that anless orr thunks of ruration and frequency ·74 Medical Wning distributions and makes the necessary calculations the writers conclusions will often be fallacious Probability and Chance-Chi Square Test In all samples and statistics always deals with samples there is a possibility that the results observed or collected are duc to chance and therefore are not significant. For this reason it is necessary to study the figures to determine whether the results are significant or not in other words it must be decided for any given set of observations whether it is probable that they arose through chance alone and are valueless for drawing conclusions or whether it is probable that they could not have arisen by chance And therefore are significant. The statisticians have developed several methods of weighing the probabilities by mathematical means in complex problems it is often advisable to use more than one of these methods One of them is the chi square iesi onginally devised by Professor karl Pearson and considered to be particularly useful in the handling of the usual kinds of medical and biological data Thc chi square test is especially valuable in comparing the characteristics of two or more groups. The question at issue is whether the differences between such groups are more than would exist by chance of the differences are considerably more than might be expected it is probable that there is a relationship The chi square test partially answers this question it is not absolute and its value depends in part on the number of groups as well as on the figure produced by its calculation Furthermore, it is valid only for frequencies and not for ratios, and it does not supply information on the meaning of any possible associations or dif ferences There are many complex mathematical aspects of this test but it aan be applied as simple anthmetic to many laboratory or clinical problems and thus us useful to the medical unter in dcter mining the element of chance in his observations Without going into great detail I can think of no better way of discussing this valuable test than by illustrating its use in a hypothetical problem Suppose the question is to determine whether left handedness in the mother has anything to do with left handedness in the off spring. The first step is to obtain a random sample of children record which hand cach child uses and then inquirt concerning left and right handedness in the mothers of those children The hypothetical results may be apressed as ·75 Statistics Employed n Medical Popers UCHILDID MOTHER LIHAVED MOTHER Left-handed contro (a) 10 (b) Recht beaded culturen 19 (c) (d) At first glansippers that a high proportion of hell-handed children hmeldt-handed mothers and that lelt handedness in the mother almost certainly be some influence on bether a child will be neht or left-handed But this may or may not be true from the figures pien, since there is no lformation supplied on bow much of the apparent different may be due to chance. This is where the ctu square to como The first step to akulate the number which might be mpacted in tich group on the base of the group as a whole. This theoretical or erpected frequency s urned at by the following formula the mathematical background for which it not here Decessary to toplore As the rected sker to (1) sboro b the tabutadion Bo the correcte monta mbt shown to the tabutafon Cube opet doceber betsbon ba the tabulation Do the card punter m (d) shonam the tabulation the toni pula of obtenons to this barnet 200 The akulation of the expected ngores thurs obtained as follows + b) 1 + 10 A- X х 1 + 12 X - 144 ( be) B. N (b + 3 10 + 30 х 10 + 8 100 XN. X 20 - IC + 2) (c + d) XN 139 +1 X 19+ 200 X 200 - 156 (d + b) (d + X XV. 0 + 10 +12 200 x 200 - 164 The follors ing table of theoretically experted frequences thus umed at UGHT HIO HOTO UNTHANDED MOTOR Lah-buddha H4 (A) 3 6 (P) Rustat buoded duren 143 610 36 4 (DI The formula for ctu square can not be applied Itu (C-F + ICH + B D .

·76 Medicol Wnting Using the figures available chi square works out (14 4-8) (3 6-10) (145 6-152)? (36 4-30) + + + 14 4 36 145 6 36 4 13 6 How is this then interpreted? It depends on a complex mathemati cal foundation, including the number of degrees of freedom (in this instance only one) as well as on the value of chi square itself But there are tables which help at this point (Appendix, p 103) From the table it may be observed that for a chi square of over 15 and a number of degrees of freedom of one (n) the P or probability is less than 001 which means that there is less than 1 chance in 100 that the figures could be the result of chance. In this instance, therefore, the initial impression concerning the rela tonship between left handedness in mother and child may be considered statistically significant as judged by this test. It does not prove the connection or how it may act it merely shows that the figures obtained have only a slight possibility of being the result of chance. Another hypothetical example may be presented to show a different type of problem in which the chi square test is applied to a larger aumber of subgroups The question is which of five preparations is of more value in the treatment of pneumonia. It must be assumed that the patients in each subgroup are random and representative samples and the chi square test is applied to determine whether the differences in results are probably due to chance or are likely to be significant The basic information (hypothetical) may be expressed as Rendis of Treau Parlexis tik Patronomia RECOVERED (EXTCTED) PIED (ARCTED) NO OF DRUO URD PATLET Sulladlozinc 102 Aurtomycin 83 Penicillin US Chlonmphenicol Taramycio 63 89 (a) 67 (1) 103 (c) 33 (d) 60 (e) 86 9 (A) 70 7 (B) 98 0(0) 40 9 (D) $5 4 (E) 13 16 12 13 S 351 12 3 17 0 7) 96 It will be noted that 413 patients were involved in the study of which number 852 per cent recovered The expected values shown in the table are reached by applying this percentage to the patients in cach subgroup ·77 Statistics Employed i Medical Papers ( Bb (C-cy (D-dy + + + B D (E-c) E + (1699 369 (10 7-677 70 7 + (9-1039 94 + + (409339 409 (55 60 334 + 46 11 06 9 37 707 + + 19 209 + + 19 OOS + 0 26 + I 33 + 0 38 + 2.41 Tuming to the x* table (Appendix, p. 103) it s found that a chi square of 2.41 with en (number of depress of freedom) of 4 mai Parcater than 0.50 (actually about 0.70), bich means that there are about so chances in ten that arry differences beineen the results with the different preparations were due to chance. Uoder these curcumstances it would be bazardous to conclude that one preparation 25 superior to another usstuning tbat the patients were alilc la relevant respects except for the drug used. The chi sqmre text can be applied to a larper number of subgroups or samples by the same formule, but the anthmctic becomes IDCTeasingly leophy and the aid of slide rule of a statistically trained colleague may be needed for the more coraplex applications of this mportant tool for testing for significance Coefficient of Correlation Many of the statal.cal devices are emned essentially at deter mwing whether the sampling of observed data may be significant One of the post mportant of these, especially when dealing with the construction of cunes, the correlation cocient. This one way to measure the degree of relationship betwten two or more characteristikos, as for example, left handedness in the mother and that in the child. It measure this one characteristic only und cannot be coterpreted Indicating any ause and effect between the cturacteristics, no matter bow close relationship is realed. Agun, the chest way to show bow the correlation coefficient mory be used is to itlognute with a small sample it may be used for Large simples also) and follow the steps of the calculation, omit ung the mathematical background and using sumple terms instead of the symbols commooty employed in mathematical formulas The cumple presented & of hypotbetical obser atlons on hasal metabotic rates and weights the question *hether weight ·76 Medical Wnting Using the figures available chi square works out (14 4-8) (3 6-100 (145 6-152) (36 4-50) + + 14 4 36 145 6 36 4 15 6 How is this then interpreted? It depends on a complex mathemati cal foundation including the number of degrets of freedom (in this instance only one) as well as on the value of chi square itself But there are tables which help at this point (Appendu p 103) From the table it may be observed that for a chi square of over 15 and a number of degrees of freedom of one (n) the P or probability is less than 001 which means that there is less than I chance in 100 that the figures could be the result of chance In this instance, therefore, the initial impression concerning the rela tionship between left handedness in mother and child may be considered statistically significant as judged by this test It does not prore the connection or how it may act it merely shows that the figures obtained have only a slight possibility of being the result of chance. Another hypothetical example may be presented to show a different type of problem in which the chi square test is applied to A larger number of subgroups. The question is which of five preparations is of more value in the treatment of pneumonia It must be assumed that the patients in cach subgroup are random and representative samples and the chi squarc test is applied to determine whether the differences in results are probably due to chance or are likely to be significant. The basic information (hypothetical) may be cipressed as Resides of Treating Patients with Premona RECOVERED (EXPECTED) PILD (LXMCTED) NO OF DRUORD PATIINTI Sulfadlubne 102 Aurcomycin 83 Perlaillin 115 Chloramphenicol 48 Terramyan 65 89 (1) 67 (b) 103 (c) 33 (d) 60 (c) 86 9 (A) 70 7 (B) 93 0(0) 40 9 (D) 55 4 (E) 13 16 12 15 5 151 12 3 170 7 96 It will be noted that 413 patients were involved in the study of which number 852 per cent recovered the expected values shown in the table are reached by applying this percentage to the potents in each subgroup ·77 Statotas Employdd n Medical Poport (B-BY (D-d1 + + + + B C (E-P E A (46 99 169 99-103) 99 + + + (55 +601 SS 4 + 77 29 11 3 + 07 + + + 97 09 354 137 005 0.3 + 019 O + 41 Tuming to the table (Appendix, p. 103) it is found that a che square of 241 *Th 23 (number of degrees of freedom) of 4 Nea P greater than 0-90 (actually about 0.70, which means that there are about seven chances in ten that any differences between the results with the diferent proportions were due to chance. Under these arcumstances ionld be hazardous 10 conclude that one preparation «zs superior to another assuming that the patients were ale mrebeant respects except for the drug wed The square fest an be applied to a larger stamber of subpoups or streptes by the sume formula, bert the unithmctc becomes LNCreampty lengthy and the rid of a slide ruke or a statically trained colleagu may be needed for the more complet applications of this emportant tool for testing for umbounce Coxofficial of Correlation Many of the statistical desmes ut aimed essentially at deter mming whether the sampūne of obseried data may be nificant Come of the most important of these, especnly when dealing with the constructing of cunos, the correlation coefficient. This is one vny to measure the degree of relationshup betweco two or more characteristics, us for comple, Lall-bandedors in the mother and that in the child. It measures this one canctorustic only and mot be merpreted as indicating ady cause and effect betucto the charactersiks, no matter bow close relationship is trakd. Apm, the casert kay to sbor bor the correlation coefiamnt may be ned is to llustrate with a small ample it may be used for Large amplos alsoand follow the steps of the calculation, omit ting the mathematical baci prouod and using smpk terms mstead of the symbols commonly comployed to mathematical formulas. The compk presented Bakta of hypothetical observations on basal rocabolic rates and neights the questions bether weight ·78 Medical Wnimg and metabolism are correlated or not and the method cmployed 1$ the correlation cocfficient The hypothetical data may be expressed as LASAL METAC RATE DAVLATION FROM KLAN CARE DEVIATION FROU MEAN WFTOHT 116 1 2 3 145 110 S 6 7 8 9 10 +5 -1S -22 +30 +16 --23 -10 +28 +9 +18 Mean +34 +16 - 18 4 -25 4 +266 +126 -28 4 -134 +24 6 +56 +146 105 150 140 -53 +87 +23 7 -113 -16 3 +28 7 +18 7 - 103 --23 3 -13 3 93 108 Maan 121 3 At first glance it would seem that the patients with low metabolism tended to weigh more, while those with high metabolic rates tended to weigh less. But are the basal metabolism and weight really correlated? The first calculation is to find the mean or average metabolic rate and the mean of the weights These come to +34 and 121 3 respectively The next step is to record the difference between each individual observation and the mean This is shown in the table under the columns labeled "deviation from maan For each case, multiply the deviations from the mean for basal metabolic rate by the deviation from the mean for weight taking account of plus or minus signs The sum of the products is added, giving in this example a gure of - 2920 20 The icon product is found by dividing this total by the number of observa tions (N) less one here there were ten observations so that the divisor would bc 9 giving a mean product of - 324 47 (For lorec samples, it is proper to use N or the whole number of obscnations in the standard deviation but for small samples N - 1 is correct In either case the same number must be used in the mean product in the denominator of the formula for the correlation coefficient as is used in the standard deviation ) The next step is to calculate the standard deviations for the basal metabolic rates and for the weights, using the method described on page 71 but the divisors used in each instance (1h15 being a small sample) are one less than the number of cases, in this instance 9 This calculauon gives standard deviations of 20 2 ·Statutics Employed in Medical Papers 79 for the metabolic notes and 18.4 for the weights. Now the figures are available to calculate the correlation coefficient bach may be cprossed us Mean protect of debate Corretation coefisicot - Product or standard rations Substitatut tht trguntos prepusty obtained, we have -1945 -394 43 Coundation codent 3X1 Hon this to be interpreted? The coefficient may vary from +110 -1 the closer t approaches to the + or -1 the otact the degree of correlation. Corisequently the closer tt 1 to Othe hess Lelihood there that the relation bety con thc chunacteristie mohad o significant in the instance ghen the correlation cocfi- Gentis close to -1 and it is therefore probable that the metabolien and makht are significantly related Homme this woold merely od zic the probabiller that the less of wesght and of basal metabolism at correlated and would casino licht whatever on the question of helber a high basal metabolic nie reduces the eight or xe cruz Contempor Pitfalls The medical nier corned with attempting 10 prone thai the data which bebas coloctad man something and he uses Haustic # a 1001 to Ripport his case. Suausis actually do not prove anything, but when properly employed they may support the unter's contention that there atually an ASSOCHLOn ben een thxects obserie. Even when applying sainixal tooh, bonec the medical ter must be on the alert to mond a number of traps but may lead him toward conclusions which are curely onwar ranitd. FALLOUS COMPANION The greatest of the pitfalls # noncorandrebilty of the records. Its fallecious, for cuatuple, 10 COTDpare the frequency of 1 partrubas disease in mea as opposed to omen anless the muss ue corrected not only for the etpected incidence in the Norts, but also for the number of each included n the sample A staple rumpic rould be to find that a police dodiet contarded the more of 100 mo and 59 mon ho C mohod in trafic dents and to condude that men were therefore in tutes us able to have functi cadents us women. The only *a) ona the oroplied question would be to obtain figures on ·80 Medical Writing the relative number of men and women drivers and to determine their risk to accident by the average number of miles doven in a specified time by each ser Suppose then that only half as many women drive as men and that in an average month the men davers were each at the wheel 1000 miles whereas the average woman drove 250 miles. By simple anthmetic it an therefore be shown that men by their greater daving, are erposed to traffic accidents cight times as much (on the average) as women One would then conclude not that men were six umts as prone to traffic accidents As women but that on the basis of eposure to risk of accident women had one and one-third times the lability to accidents that men had! This question of exposure to risk anses with great frequency in medical studies and must always be considered when comparison between different groups is made The most remarkable example of figures dont mean what they seem that I know of was given in the form of an examination question to which I was exposed years ago The question went something like this How would you explain a decrease in the death rate of New York City and a decrease in that of New York state outside of the aty but an increase in the death rate for the state as a whole?' Most of my friends on whom I have tried this say it is impossible. But here are figures showing hop it can occur 1920 1990 MOPULATION DEAN DLATIM DEATIL POPULATION RATE DFAT31 LATE TER 1000 TER 50,000 4 000 000 149,000 333 New York Chty alone 2,000,000 New York State excl. 2,000,000 State as whole 4000,000 25 000 75,000 12 5 18 75 3 000 000 9000 000 35 000 11 66 175 000 19 47 The point illustrated by this example is common in medical problems There are in the example nio variables, the size of the populations at risk and the attack rates, both of which may vary Thus the time elemeat becomes of great importance and must be taken into account in making the interpretations It would be possible to continue almost indefinitely with hypothetical and actual cxamples of fallacious comparisons One more only will be mentioned here because cramples of its kind are hard to avoid unless the medical unter is consciously on the look ·81 Stotistas Employed in Medical Papers out. This is the question of attributes the series of observations are compared In order perhaps, to decide beta een the efficacy of 1w0 unds of treatment for cancer of the lip If one series, bowever contains a different percentage of far adinaced cases than the other it is not reasonable to drar Lay conclusions concerning therapy since the attributes of the two are not tic. This is the we lbad of problem prepously alloded to in discussing cialis tion of clipal results by such terms as pood, Tur or oudleat, which are so bileły to differ from obstri er to obscti et or by inter natronul comparsons of deaths ben the classifications in different countries are not alike Any attempts at comparisons will prove futile unless the groups compared haic 6c attributes PERCENTAGES. Percentages are merely a means of a pressing the resulu of a study involving a number of patients or observa trods. There appears to be no absolute role as to when percentas should or should not be used, but 10 ctpert in statistics advised me some years ago that m hus opinion percentages should not be toployed when one than 100 observations were involved. The reason for thus is twofold that petoenlages are unnecessary for a Amalt pumber of observations because it is just as easy for the reader to interpret the basic figures, and that they are a false uopresion of reliability If one is referring to 20 patients, 2 of bon died, it is just as clear to the reader to say that 2 of 20 paixents died, BH that 10 per cent of the patients died its esscouually more precise also among the nett 20 paucis observed 6 might have died. If ibs bad occurred, tbc death rate would have been 30 per cent instead of 104 rosuh which would have glica un epurely different picture to the reader For these reasons it seems pood policy not to employ percentages encept ben the number of obsessions of parlents has greded 100 thus fring a large enough number to warrant this book of expression A few years epo Campbell pointed out znother frequest fait top in the use of percentages. He llustrated it ths way to a series of patients with detayed operation, the mortality was 1.5 per cent, hetess in the series with bromediate operation, the mortality was 48 percent. From this may be concluded that death occurs more than three time u frequently after wamedaic operation. Now this Bi common error Carapbell says, which k brought about by the sell-delosion of the unter In this case the figure 48 had been talon and the condusion dm that is more than three unos 15 oct of course is inde Hosenet u sbould bave been posted out that regardless of whether the operation was delayed or ·82 Medical Wolmg immediate the recoleti rate was 98 5 and 95 2 per cent respectively a difference of only 33 per cent which gves an entirely different impression than the conclusion ongnally expressed Prosentation of Statistical Data The final problem in preparing statistical data for publication is the decision on how the material is to be presented Unless this question 15 properly answered the reader may not accept the writer s observations and conclusions as completely as would be the case otherwise. Tables and graphic representations constitute the principal methods used Their proper application is an art in itself and only a few of the high spots can be touched on here Tables The purpose of tables is to express the nature of the relation ship betwten certain observed phenomena in such a manner that the reader can understand the data and conclusions clearly Pearl has said The purpose of tabulation is to so arrange observations that like cases shall be put together and their frequency of occur rence in the whole group thus be made apparent Unfortunately many medical writers do not seem to have the aim of tabulation clearly in mind Many published tables are merely lists of individual observations One such which I saw not long ago consisted of a list of cases of a particular disease recorded over a four year period it occupied twelve full pages in the medical journal in which it was published. In this table no effort had been made to put the 64 cases together in such a manner that the reader could draw any deductions concerning average age, duration or symptoms or any other common features. Yet to do this is the purpose of tabulation Before constructing a table it is well to ask oneself first what its purpose is to be and what it is supposed to accomplish in the mind of the reader After the table has been prepared the uniter may again ask himself how well the table has sened these objectives Pearl who made these suggestions. recommends that the unter who wishes to gain skill in the construction of tables might hell apply these questions to each table which he sees in print After deciding the aim of the proposed table, the date must be sorted into groups of like characteristics. The degree of eness depends on the purpose of the table it might for example consist of all patients with leprosy if the study involved gallstones the grouping of like with like might require subdivisions classified by ·83 Stathaa Employed in Medical Popers afe und sei la other words, the decision as to what characteristics to consoler as 'lile depends on what is to be demonstrated in the table. There are a len geral principles to which wery table should conform 1 The table should be dearts mtclugible by itself without detect to the tatt. This means that it must have a urtle hxch fuly stalas the subject and describes the materal presented Each column must be completely labeled. It is not sufficcat, for example, to bed a column Hoght without lochuding information on Whether it um inctes, meters, or whatever the wordt of measure happens to be 2 The actual unit of observations should always be und cated, smer the fisting of proportions or rates alonc may pic in colorety false picture and without the basic figures the reader is depth ed of al information 3 Whea rates are price in a table the base must be clearly stated it makes a lot of differace whether a fatality rate is per thousand or per hundred thousand' + Tabubuon of alcance of means alone o inadupable of 11 an be norded frequency distribution gres a better picture of the dara. S Finally a table should include mformation on what is left out, and the reasons for the omissions. If a follow-up study for wample, ortuts akrpe number of patients who bave died thus tertiog a potentally tremendous mfluence on the results, the reader is entitled to know about it. There are many cucelent szentific workers who un reporting tbeur toestenbons. Gail to follow these basx principles. The cumple sbond n Fiz 19 B the first half of a table in whxcb the suthor has saumonarzed his observations, but bas not Intly analyzed the cable s purpose of the effect it will produce on the reader It is difficult to tell from the table alone rig what the autbor considers really important, or bat tbe data prone or disgrove. The table sborn in Fire 20 a better Tbe subject as adequately stated and the purposes car The oumber of cases m each proup stated, and the indo tocomiats no difficulty in tooning bat the Tutet nts hon to see and to conclude. But me this could be mproved somewhat. The percentage column moder Relapses could be omitted entirely the one figure appears in that column D01 only ones 20 mpression of fake riability considering the Aumber of obsautons moted but adds Dothrag hxcha 7 out of ·84 Medical Writmg Hondolk Dato y Ceres of twik Aly down First horizontal column in auch itsedmiles blood count *hen fort sett second colupn her theypy had been completed. (This period of time does not necessarily correspond to the duration of tatment odoted in Table 3 Tbe thone clapsed between blood counts here is actually the thing from the first blood count to the final dimifiant blood count after therapy had been completed and bad had time to become chectare.) HEHO ELAMUD CAE ALORIN RKC X POLY PLAIE W X TE NO (months) (mm 10*1210X Murs 10 hou) LYH- MHO- CYTES HOVO CYTE) LAVA 0 13 $6 7.03 S 102 75 9 3.36 236 240 190 2 89 31 13 12.4 12.5 12 441 2.98 635 265 56 57 3 439 67 54 3 6 134 conda 4 45 0 80 14.2 10.2 4.32 385 9_2 11 75 53 31 2 6 5 0 45 85 7.8 2.75 2.39 160 45 68 69 26 23 S 7 4.2 6 2.52 220 0 2 76 10.0 6.4 89 66 78 30 18 377 7 0 2 6 99 54 31 1 74 10.4 1.6 22 33 ܕ ܝܚ 10 8 13 0 1.5 122 104 365 301 3.8 70 310 95 73 66 9 36 71 0 9.3 16.9 9.3 150 743 365 120 -95 بيا - 10 9.0 0 5 10 36 2.14 743 44 | 115 26 30 5.9 275 66 Fig 19 Emple of a table which fail to Lalyze its propose and thus not clearly intelligible by itself 13 patients does not supply at a glance Also the word affected in the left hand column under Relapses could be omitted Never theless this table compares favorably with most of those which appear in the medical literature, There are some practical if minor points in the construction of tabla which deserve attention One is the sice of the table. If too ·Statistics Employed a Medical Popers 85 Korteri RafnnDerror of Cubreak Tram TUNAY WITH CHLOMHOOL ULAR PANT Dorthog of Tretot Adobention Camil Group Nota Tatod Day of Doces Tout Tool Nicoder fer ao Gn.co of Sunt stopped Ort rood) Milents th Arvoeren Adid Affected 1 om 1 6 A 7 or los 12.5 204 VO 11 10 14 un 19 B 112 137 0 49 0 110 с 20.1 1 0 0 14 dagi F14. 20. A food table. moch materals focused in a single table the date may be bord for the reader to compreend. Furthermore, in reproducing such a uable for publication the type may need to be reduced so greatly in order to make the table fit the past, that it becomes practically zlepke. It is therefore, important to anticipate ibc final appear ance and perhaps to use two or more tables in some insunces obce only one on rally contemptated. Finally it is helpful to the reader if the type can all be read horizontally ni nnoying to be required to turn 2 bole journal or book in order to read the coturan beadings such as those sborn in Fig. 21 An Trude Modbe Upon artib AROUX UN UHKAN MUNITAR11131 11 Okrutuo MwXyist - Lunctnut Mwmnat KHST PLAINTS 1 J100 100 200.000 7.00 67 J 40 01 0.000 on 100 1 11 2 Normal 510 $5 4170.000 111 13,250 13 13 J3 13 Narron 6710 $1 4420.000 0 11.50 31.3 135 31$ 1$ Normal Ft 1 The product codemo honchs in the table ore 10) A to the act Grabendral cebes tector ·84 Medical Wnling Hematolopte Doto 24 Cores of Vfxhipte Myelong First hortzonul count in each cake smiles blood count han art en second column ser therapy had been completed. (This period of timo don not neocharity correspond to the duration of treatment as poted Table 3 The time clapsed between blood counts here is atually the unc from the first blood count to the final dgniant blood count after therapy had been completed and had had time to become effective) CAR NO. ILAMED THE (months) HEHO OLODIN Аках FLATE RX X POLY LETI X X (mm 10-02 103 (10 LY CHO CYTES HON CYTUS PLANYA 7.05 102 7.5 0 1.3 $ 9 3.36 2.36 36 240 190 2 31 13 0 12 12.4 12.5 298 633 2,65 57 3 67 S4 3 0 6 134 Ananda 4 0 14.2 10.2 432 3.85 9.2 11 73 45 63 2 6 31 68 5 0 45 8.5 78 2.75 2.39 40 4.2 160 45 6 23 6 220 4 0 2 76 10.0 2.32 3 77 6.4 8.9 66 78 30 18 7 or 99 34 31 1 74 10.4 16 7 64 22 30 6 3 7 13 8 0 7 15 122 10 3.63 3 11 5. 7.0 310 99 12 4 9 0 9.5 16.9 9.3 56 2.95 15.0 745 365 120 71 86 11 10 90 745 26 0 5 36 214 113 275 66 4 Fim. 19 Example of a tabk which fail to analyze 10 purpose and thus not clearly mtelligibk by itself 13 patients does not supply at a glance Also the word affected in the left hand column under Relapses could be omitted Never theless this table compares favorably with most of those which appear in the medical literature, There are some practical if minor points in the construction of tables which deserve attention One is the site of the table iſ too ·87 Statistics Employed in Medical Popers - Untreated postita Oer0 Trrotat patients 103 102 Body Iap 101 F OOH 99 98 97 2 4 5 6 7 8 9 1o Days F 2 Avans of body compartir bn 75 untrated patients with lobur poemon modo frated with perta.co. reasonably the significance of the materal presented. A series of hypothetical data has been down on three diferent cules in Fig. 23 It should be noted that the different effects are created by changing the sales of the ordinates and abscista. In addition to choosing a saak wuch best relect the nature of the study the ter bas to decide which characteristic to plot on the ordinate and burch on the abscis sometimes this makes a great deal of dulicence. BAR DLACRAIS. Bar dugus are also commonly wed in presenting medical data. They serve a useful function, particularly for mpamadeot phenomena such u nauka or judat m contrast to body temperature or pulse rate hurch continuous Ima can be donn. When properly Constructed (scale omportant in bar dagrams also) and fodequately labeled, they are ay on the ges, roddy interpreted and particularly useful m lanter slida. Figure 24 shor's two typal cumples. PTE D14CRAVES. Another common form of grups presentation stbe pre diapoto. Thu lomo of graph is used to show the propor LOTS bxcha separate characteristics have to the whole. It is popular m public hauh education und is used ertetety in portancial ·86 Medical Writing The presentation of observed data by tables is not simple and deserves more effort than it commonly receives But adequatu tabulation is extremely important and shili in constructing tables may be acquired by analyzing their purpose and by practice. Graphic Presentation Atvareness of the uses and limitations of charts and curves and of the different kinds which can be employed to serve particu Jar needs can be bighly useful in the preparation of medical articles The purpose of all graphic representauons is merely to organize statistical observations in a manner which can be comprehended stually and thus and the reader to understand the meaning of the figures more readily Examination of the medical literature reveals that this objective is often not appreciated and many graphic presentations fail to make the reader s task casier Sometimes, indeed they give an erroncous impression of the meaning of the computations. Statistical diagrams are constructed by setting up a system of coordinates after which points lines, or surfaces can be inserted connected, or blocked out to provide the visual picture. The graphic presentation cannot be considered as a substitute for the statistical data but only as a means of making such material more comprehensible hence the writer who wishes to employ this device can choose which of several graphic methods is best suited to his date and to the audience at which he aims. RECTANGULAR DIAGRAILS in medicine the most common form of diagram is the rectangular in which one characteristic is plotted along the horizontal coordinate and another along the vertical coordinate The example shown in Fig. 22 an be readily recognized as the prototype of innumerablc graphs in the medical Intcrature. This is not good graph in a sense it does not reflect Variation 10 individual observations, a question which should be considered in any statistical study but it does shon what appears to be a definite trend and it shows it in a way which is easy for the reader to grasp There are many vanations possible in this type of diagram It is always essential that the labeling be kept full and clear and not so many lines inserted as to make the diagram hard to follow with the cye. SELECTION OF SCALE. The scale on which a diagram is constructed is important because it affects the conclusions which the reader will draw Scale must always be considered carefully so that the result will give the reader a visual picture which reflects

·89
Statisties Employed in Medical Papers
trant
10
*
03
0
A B с D E F B H
1
findoro
LA
2. Renato
DICATA,
Met 73%
IntU US
KARO QOST
I OMEK ODST
000
NUMBER OF INDIVIDUALS
illi.
DH
OH
ASE IN YEARS
Ft + Typload berdirin.
manis in unele patient along time absor. Lo essence na
prapblc bostory of cintaal and hboratory observons on one
patront. Presentations of this sort, honte almox in amably fail

·88 Medical Wntag 190- vt 180- ha TH 101 160 ISO- Stort ! 234 6676901112 of dret The Inn 200 190 A. 180- Ibu. 1 70- 180- 160 Stort I 2 3 4 5 6 7 8 9 10 11 12 oldet The wako 200 A 100 wt 100 170 lbs 100- ISO- Stort 2 3 4 5 6 7 . P10 lt 12 of dret Time in this Fig. 23 Dhgroms drawn on three different sales to show difference in visual effects. - expositions. In schotfic medicine and in the biological sciences it is generally considered of restaed value and should be used sparingly A typical example is shown in Fig. 25 CLINICAL HISTORY DIAGRAMS In recent years another form of graphic presentation has become increasingly popular with clinical invesuigators This consists in showing several develop- ·89 Statidas Employed in Medical Papers 43% PINOYTAXT OF TOTAL CAKI 107 10 A B C D E F G к Le Z14 rochie hadi LE more C AL Ctr, N. WHITE ST LOWN IL 1 000 CHR DOZ 1199 n 1000 241 153 00- NUMBER OF INDIVIDUALS - H- 40-44 to 00 10-14 AGE IN YEARS Fe. Typial burde ments wogle patient along a time absciss. In axence it is a graphic hotory of cliakal und laborttory observation on one paucal Presentations of this sort borence almost evarably ful ·88 Modical Writing 200 190- L 180- In Ibr. 170 1604 150 t Start I 2346 6 7 8 9 1112 of duet Time in tu 100 AY wt in 1804 1 70 160 150+ Stort 1 2 3 4 5 6 7 8 9 10 11 12 of ditt Tuone mweek, AY 110 t. 100- 170 s 160- ISO- Start 2 3 4 6 7 othet Time in the Fig. 23 Dagramu drawn on three different scales to show difference in visual effects expositions in scientific medicine and in the biological sciences it is generally considered of restricted value and sbould be used sparingly A typical example is shown in Fig. 25 CLINICAL HISTORY DIAGRAMS In recent years another form of graphic presentation has become increasingly popular with clinical investigators This consists in showing several develop ·Statedra Employed in Aledical Papers 91 cumples the method pahars rocks its greatest heights of con fogon in the field of clinical endocrinolop The graphic representations mentioned are perhaps the most compon of those used in medical papers. There ure many others which occupy an important place a certain types of problems histograms, frequency pohyeons, logarithmre cunes, and others, but a deset stort of these belongs in a book on saustis rather than one on medial Titung The writer who wishes to delle more deeply into this subject to therefore referred to the list of references In the Aprenda (p 93) ·90 Medical Wnting 30-40% Rheumatol.de Earthritis 30% other types -30% Fig. 25 A pie diagram showing the proportions of all types of arthrth. LUMAN PUTUT LADACHE PTTRA COCA 20 HO HLODO RESUME 100 to 00 Ort HOROCHLORIDE WORLD 7 tor M STOWO WICI 0001 1 LOOD Umla 1 of tof 41 Cena nvt 4000 1000 OUTPUT 000 Fig. 26. A clinical history dlagram showing the hospital course in a case of severe hypertension treated with dibenamine to fulfill the one most important purpose of the graphic method - case of interpretation Figure 26 is only one of innumerabic ·APPENDICES L Rafwonics The following rerences basc becan selected because they may be partxukrly useful to the medical writer General Croby Henry 5 Better Httmy 1 York, Hartoon, Bruce Coonary 1926 Charting to helplo pot predich Gomar, Sir Ernest Muin Bords. London, Ko Alajesty Stationery Ofia, 1977 Delightful, Iberally quoted to this book Fletch, Rodolph. The Art of Plate Tuik. Ma York, Harper & Brothon, 1975 Much common sense about writing coonins stapter on ito WTML Vindical Album, T Cuijord Mots on the Composition of Schentic Paper London, Alaconian & Co Ltd.1905 An arabirabite book on metaal tutios, but hard to get Foto Monte Aledical song d Ed Toba delphia, The Bllbstvo Company 16 A raroodity wable for los restos on style Ham Richard ML Better Medical Writing Some Typical Errors and How to Arad Them. J Urol +454 (Sepl) 197 Short but tety valuable ducson Statustk HA A Deadford Puuples of Mecal Statistich Ed. Loodon, Orford Upn xt) Pro, 1940 An nitrocty col book for any medical or dance witer in Hatua Soon Coorothee On Standards for Graphik restoration Prefabry Report Quart Put Am. Sat Axoc, H 790-197 1915 Vanlund, Doca Ecruter Medal Station Pladehida, HB vonden Cocoon 1952 Come od thorooghly up to date. Vanlind, Donald The Treatment Chical and Laboratory Dao London Other Boyd, 1944 Adertad rdu daculoa. 93

·No text on this page.

·APPENDICES L References The following references have been selected because they may be parixularly useful to the medical writer General Cob Henry Better Rey de York, Harcourt, Brace & Company 39% Charming a belpful por medical Goners. Sur Ernest Polo Monde London Ho Varesty, Statuery Officer 1943 Dictaphorid kerally gooted in this book, Flach, Rodolph The Art of Path Tan Mon Tor, Hoper & Brotben, 1976, Afact common some about writing contam chapter od KKTC withog Medical Allbutt, T Chord lote e lhe Composition of Scientik Pares. London Lacmim & Co., Lid., 1903 Animalmble boot op mentical thing, but hard to get. Flubben, Morris Aledal nature Dd Ed Philadelphia The Bllbion Company 1910 Acta paruruburty zabke for Lo Risto on style. Hema Richard M. Better Sledical Writing Some Typical Erron and Hom to A on Thema. J Urol, 454(Sepe) 1933 Short but ratty yabalik dacosta. Statistics HIL A Bradford Procpla or fedeal Statistics 'th Ed. Loodon, Ocord Utversity Ho, 1940 An ettrond, wdut boot (or any meetical excience the wint Lommt Coromittee on Sundards for Graphes Presentation Preliminary Report Quart PURI Am St. Anxo 14 790-07 1915 Yuntal Doral Blanetary Materal Santo. Pollacktpha, B. Saunden Comraep 1942 Come i thorough up to date. unlod, Doon The Treatment of Chalud Laboratory Data. London, Olno & Dord, 1935 A duwdd docenioa.

·94 Appendices Pearl, Raymond Introduction to Medical Biometry and Suustics. Phila delphia, WB Saunders Company 19-40 An excellent exposition of the subject but rather complex and mathematical for the average medical writer Miscellaneous Webster i New International Dictionary Springfield, Mass., G &C. Merriam Company 1947 Webster u Dictionary of Synonyml Springfield Muss., G. C. Miertam Company 1942 Dorland' American Illustrated Medical Dictionary 22nd Ed. Philadelphia, WB Saunders Company 1951 Standard Nomenclature of Diseax and Operations. Philadelphia, The Blokiston Company 1947 Peppers Medical Etymology Philadelphia, W Saunders Company 1949 II. Style for Usting Reforoncos Medical periodicals and publishers do not all use the same stylc in publishing bibliographies Consequently whenever on author knows in advance where his arudle or book is to be published he should follow the form of that particular publisher For the convenience of many writers two styles are given here Publications of American Medical Association 1 Forsham, P H Thorn, GW Prunty FTG und HUL, AO Clinical Studies with Pituitary Adrenocorticoiropin, ) Cun. Endocrinol 8 13 (Jan) 1948 2. Boland E. W Oxygen in High Concentrations for Reller of Pain in Coronary Thrombosis And Severe Angina pectoris, J.AMA, 114 1512 (April 20) 1940 3 Wiggers, CJ Physiology in Health and Disease, ed. 3 Philadelphia Leo & Febirer 1939 p 431 Publications of 1 B Saunders Componi 1 Adams, CJ, and Connell, WP The Symptomatic Rexel of Emphyx3 am) MSc. 187 687-692, 1934 2. Baker F L Management of the Pneumoptas JAMA., 103 84-47 1938 3 Jones, O F. Carson, LJ, and Hoopo AR Recent Advances in Sulfanilamide Therapy Am. ) Digest Ds & Nutrition 60 659-691 1939 4 Smith, WF Surpal Dugou and Treatment Id Ed Philadelphia WB Saunders Company 1973 p 273 ·95 Appendices m. Common Abbreviahons in Medical Monuscripts Ulternating of EC Aoptrum med alone tatrack tudom. contbter como cubic ctutora att foot ou il coblc och Cu ba cubic meter duobd do. dengan dom decutor d de curcot chc F It. Fahrenhet foot, fect Ako pam hoor nch Logo kilomcher tulon ott quod Onanion TOO nho liegen alcat hero indrer minder otocoa nc. Fa ounct ob mta. MO az M. ih road ant matund KU ad بهر ·Appendicos 96 I Proofreaders Marks Stort porograph rood The orked The proof No porographi nunn Rolse The proof Lower C More lett ? The proof. The proof The 'proof More right 11 Align type Three of Two Women Straighten fline The Proof Imert period 0 Iuert CODANO 기 The proof, The prool, The proola The proof Insert colon ><ti Coll muert renkolon Insert opostrophe muert quotation morks. mert hyphen. Insert Interior figure Insert superior Rowe Incert exclamation Dark. V The boxes proof Marked by prooly ./ A proof mark. Water HP **+ P =CA Promin ? hut good was The proof stod by c/] The lone, boy (1) The prooba The prool ILL The proof, 12.1 The proola 2 Insert question mark. Query for orthor Luert bracket. 43 (/) huert parentheses Insert len doch Insert I-en dash 111 Insert 2.en doch ideat len The proof Indont 2 ani The proof Indent I. The proof ·Appendice 97 Kidrider beste I The prooft х Lokalite 1 Lokert proof * Theol. Ten by 9 The pool. fe proef Lda The rool eg good proot. The proot. Con il De pobol to Aprilgood Where of The proft. Loc. le the front lo $I T **0c The prool rool 40 Сю Carpets and med cph The prool Crec prool The proof pro È Conrin. The proot. The pool ital pool. Am The rol Le pret prool Min lume ha. s od 07 Sni stat the prool outs Heprost ·100 Appendicos VI Useful Equivolonts Weights and Measures 31 1 ANOTHECARLES 1 couple 1 drom 1 oude 1 pound METRIC 1 milligram 1 centigrum 1 decorum 1 rum I klogram AVONDURO 1 Ounce 1 pound 1 ton Weights AUTHECARO HETIC 70 graine 1.296 TL 60 mins 38 460 prabes (8 drams) 5760 muins (12 ovoces) 373..pm. APOTHECARIF MTRK 1/65 grain 0.001 po. 1/6 prin 001 pm 1 1/2 mun 15 432 mm 0.001 Wognuti 22 pounds avdp. APOTHECARIOS METRIC 437.5 grains 23.35 m 7000 mins 433.59 pm 2000 pounds 907 184 Milograms 01 g 1000 PIT- AVODNO 1 duldrun I fuldounce 1 pont HENRK 3.697 cc. 29.573 cc 473 167 cc. 946.333 cc. 3785 cc. 1 quant Liquid Measures APOTHECARIES 60 mdoins 8 nurdrams 16 Mudounces 32 fuldounces 128 Duidoum.cos CUTIC MEASURE 0.061 cube inch 0.61 cubic inch 61 cubac Inches 61.0271 cubic inches 1 duidnam 2 Qudruna 4 furtum 2 Muldounces 4 durdouros 8 nurdouka I pullon HOUC 1 murter 1 cendhter 1 decirter 1 Inter 1 teaspoonful I dessertspoonful 1 tablespoonful 1 winculusul 1 tencupful 1 classrul METRKC I o. 10 °C 100 a 4 cc. 15 cc. 60 cc. 120 cc 70cc Linear Measures 1 inch 25 4 run 254 centaneter I foot 12 actes 30.48 centimeter 1 yard 36 Incha 09144 meta 1 rod 199 incha (16 in recti 3.029 miles 1 maite 5200 feet 1 609 Iulomatt 1 min. 0.03937 inch 1000 MTOTT 1 ceptimeter 0 3937 nch 10 MOTI 1 damcter 3 917 inches 10 centoncte 1 meter 39 37 mches 10 danctors I klometer 2001 feet (062 mikel 1000 meter Approximate equivalent ·mala 20 prains prios pal Pros 065 pm prums 06 prin ரம் 101 Appendices Dasares (Approvincie Endralenis) Weights Om AUTHECALS UUTA 001 PD 57 -77 pata 0 0001 pranime-1640 pm 006 prin 10 - 70 pala 0 -17320 pua 0065 FT 65 -1 00003 pon-Trio prin 00? mo Ome-11 malo 00014 po-04-19 pain 0 000 HOS-1/120 per 00 Prim .0006 06 mg-1/100 pron 009 pm 90 m I prima 01 100 m-1 ir prins 0 0007 om 07-19 m oooo po-01-10 prin Q. L. gram 0.0009 07-1775 pro 0001 pomy -154 pm O po 03 300 00011 port ng-140 pain ir pum 00013 to 13 -1/0 pa 033 PM 310 00014 pon-14 ne- 4 pub 04 400 mg 0 2016-16-10 pin 045 pm 40 m.-7 00011 pm-1.5 -1136 pin OS "00-7 In mulos O prim- - pin 03 pm 000m - 100 pm 06 600 mg 00026 pon--26 og 1823 prin 60mg -10 0003 pon me-1) po 0.73 prio 730 mg-11 ரம் 0004 por ng-1116 puh OP pon FOO my 0003 pre-Se-1/1 60m-13 0006 pm- 6-1/10 pun 0.93 pro - 910-14 OCT -7 -19 1 mm 1000 -15 000 mar 04-11 pun 1.06 pag-IOKO my -16 0009 poro-In Fun 113 pm-11-17 O 01 man-10-16 mun I 13 pm 112 my -13 0013 pto-13 -13 pum 1.3 pm-1260 mg 12 0016 - 16 mg 14 prim L30 pm1000 m OC preman 13 MT 11 ரம் 4 dnt 003 pan-30 m - 3 pun S O pre-1. me-1 pen 1 004 Pro -JS mun 10 I do 000 per neprun 15 PID drums OO port-34 pun DO ound Loud Measures n APOTHECARIS HILC APUTKECAUS 000 cts around - Intru I cobi tutto பபன் 00d at numetti 34 pinan to ahe tnt co- ir Andron 20h crotocrt -1 15 coby tatoor Auditors 01 harimaket-11PRO o cubre cuir Buty O2 cohes contact - 15 cubkatatonra 36 Andance OD at conanon cabi combanon fundource 03 he counter- OLD Taht statutor-1 3 4 udownicy 0.3 Otarttoch tornar Dundon 06 ate antema - 10 JOLLO 100 cutk TM-Jin Budouc 075 aux autre -12 ILON 120 konton Cu alter-15 200 abitude 7 Jouro cubu anlar- 0 ch contacto 3 ahசாட்-45 Lokbonds Hou could 4 cutit! aut cuero-1 Atact-11 1900 as coton quart hot Atlet) the house analet online (N) The one wedste asta penctal pedale. mura mm pra fuis Prins Po 03 pm 1 $ ·102 Appendices Temperature Conversion Fahrenheit 10 Centigrade +2 +3 +5 +6 +8 +9 0- 10 20 - 30 - 40 - 60 - 70 - 80 - 90 - 100 - 110 - 120 - 130 - 140 - 150 160 - 170 - 180 190 200 210 - C с C. C C с. C C с -177-17.21-1661-16.1-15-1501-14 -13 8 -133) -127 -122 -116-11 11-103-1001 - 941-8.8-1.3/77 - 72 - 6.6 - 61 - SS1-50 44-3.8 - 3.3 - 271-22 - 16 - 11-O.S 0 +03+1! + 16 + 2.21+ 27 + 3.3 + 3.8 +441+ 3.0 + 3.3 6.1 6.6 7.2 77 8.3 94 100 10.5 111 11.6 12.2 127 13.3 13.3 144 ISO 15.5 161 16.6 172 177 183 18.8 194 22.0 20.5 21 1 21 6 22.2 22.7 23.3 23.8 244 25 0 25.5 96.1 266 272 277 2.3 28.3 294 300 30.5 311 316 32.2 32.7 33.3 33.8 350 35.5 3.1 366 17.2 377 38.3 38.8 394 40.0 40.5 41 6 42.2 42.7 43.3 438 44.4 450 45.5 46.1 46.6 472 477 43.3 43.8 500 30.5 S1 51 6 S2.2 32.7 33.3 S3 / 544 350 55 S 56.1 $6.6 57.2 577 583 58.8 $94 60.0 60.5 611 616 62.2 62.7 63.3 63.8 650 65.5 66.1 66,6 67.2 67 7 68.8 700 70.3 711 716 722 72.7 733 73.8 744 75 0 75.5 761 766 77.2 77 7 733 78.8 794 800 80 5 8111 816 .2.2 827 83 3 83.8 850 $55 86.1 86.6 87.2 877 88.3 88.3 $94 90.0 90.3 911 91 6 9221 92.7 933 938 944 950 955 96.1 96.6 97.2 98.3 98.8 99 41 1000 977) NOTE Each of the figures after the decimal point in this table brouting decimal. Consequently if rounded off to the pares tenth of a degree, each decimal figure pretter than 4would be increased by oo ch 114F - 45.555 + C, which would be alled 45.6 C. (From "Medical Physios" by Olio Guant Yar Book Publisher, Ioc., 1950) Temperature Conversion Centigrade 10 Fahrenheit C. +1 +2 +3 +4 +3 +6 +7 +8 +9 0 10 - 20 - 66 40 F 32 SO 68 86 104 122 140 158 176 194 12 F "F F F F F F 338 356 37 41 392 41.0 48 446 464 482 S18 5361 354 577 590 69.8 71 6 73 41 752 770 788 806 24 878 96 914 932 950 468 97 61004 102.2 105.8 1076 1094 111.2 1130 1146 116.6 1184, 120 123.8 1296 1274 129 131.0 132.8 1346 136 4 1 138 141.8 | 143 6 1454 1472 1490 10.8 1426 1544 156? 159 8'161 6 1634 165 1670 1888 170.6 1724 1741 1778 1796 1814 183 1850 1868 1896 1904 197 199.8, 1976|1994 101 2010 2018 06 6 4.4 10 60 70 - 40 - 90 - 100 - (From "Medical Phystan," by Orto Gissser Yeas Book Publishers, Inc., 1999) ·103 Emmy OUT n QlH*Sing,T$148nmpmnNU%和 6310SbHm1, +8+3%AB%刀pm=a=%af% ps 68打打i3Fred H85mngHeg8 + xinma为以阳打女minHn 企 b d 9. | 11} 15 13 w bupi*f3af99%,RSD为打比如nt n938#de%%m%n&n*门" alway. Panad nut0w LITT) 10 TPE For larger abord-soy be referred tormity to Fra Leone thermia a Serena Dan Motrend Co Aner 「Nigh remand dumpsd drum) ROTTO produkty ole Appendices VIL Table of ? Probability Scale f=25>-in-HFn r83333858 只 Sn=81#tun msLFIR仁ngn unmunmingn pm-8Fngts Illulum LI 14 FL 16 九 1 中p BTti-TTFF - II 1 14 11 可加 n Fup

·No text on this page.

·INDEX ARMATIONS, best of 9 AIA Antholowe 3 the of, Bolthee tyre me of, to lady 10 on todes, 9 Brushed cry drana rtpro- Aconkpment of pure quot duction of, 60 mattor 2 Adinice to malonul 35 Adjatnes, the o, 0 CAN Frank Cook of our obok 9 odpactna, 1 DO 1 Campbell, on con in the offer Adhorts, time of centres, 31 Alphabetizm or boder, 40 Citalon tog la de 35 American Medical Awodabon, pub Cank, Dory, 37 laboos ol, bibbographic style in C mumkerk te of A case it 24 portu, Amahan, kucal. reports, 3 Apothecaro dichts and measures Kentications of patients h. 3. equnalents of, 100 Deol yntry od conclusions Ardomcuc mean, 70 77 Arucht. See Vapor Carchards, cron referred from, Artidos send the, omkoon of a foder 39 modelu mdevine undet 37 79 Atom Aumery tak 0,99 Oration of In connectne ide opts, the of curi, 41 Attributes of butthal Amples, 31 Conte Drach, laminal LM Autoc, autonols Author duties of In Mpon of Cario Bartın. Unded States, danh mare 33 cettuales con pidetty Averpo, D Catrace Fuhrenheit com viac no, taktes. U tahle 10 Avonduras vapus i measures Central point 70 equn alents of 100 tendency D Chance, 74 Changes to yoor 35 Bural proprio unice, 9 Charts, 16. Sot allo DOTTED, to Dar daarns 17 dand door Dunk lowed) mg for 63 beo Day Fantana Chu quan formab for 73 Braintal ruha, 67 probablity cal, tant of 103 Buttgart Woute 3 4, 74 B.Kartucs, fometendo myk for listing from in, 4 for partidoa of 76 Curlocutor 103 ·106 Index Citation of rcferences, method of 31 Data statistical presentation of in Clinical history daagrams, 88 graphs, 86 Clinical observations as part of in tables, 82 structural outline, 6 review of by impartial obserier Coefficient of correlation, 77 66 alculation of 78 significance of 69 of varlatron, 73 sorting of for the in tables, 82 Color Ilustrations, 60 See Also Illicos treatment of 69_91 frarlants Death Causes of International / overlays, 45 of 23 68 Combination plate halftone en gray certificates, compilation of by inøs, 60 United States Bureau of the Comparability of data, 68 Census, 68 obtaining by alternate selection Degrees of freedom, 76, 77 of patients, 69 of relationship measurement of by selection of similar proups, 69 of scatter 70 Comparisons, fallacous, 79 Deviation from mean 78 Condusions as part of structural Handard, 70 outline, 6 27 alculation of 71 Conjunctions in alphabetizing. 40 interpretation of 72 Controk, 4 Dagtams, statistical, 86 Conversion tables, temperature, 102 bar type, 87 Copy form of for submission for dinical history type, 88 publication, 34 ple type, 87 Copynghted material, 100 of 32, 58 rectangular 86 Copyrighting of fllustrations, 58 selection of scale for 86 Correction of final draft, 35 Dictionary medical aus aid to selec Corrchation, coefficient of T7 tion of words, 19 alculation or 78 Discussion us part of structural Costs of photocogravlag, 60 62 outline, 6 Crayon drawings, brushed, reproduc Diseases Standard Nomenclature of trou of 60 23 Lithography, for line engraving, Distributions, frequency 73 39 in tables, 83 Cross references, 38 Dobell, on errors in citing refer Cuts. See Illustration crices, 31 Dosies, equivalents of 101 Double print halftone engraing Dara, statical, accuracy of 66 60 analysis of 69 Dralt, final, additions and correc collection of 66 tions in, 35 compenbility of 68 firxe, 7 obtained by alternate selection donts un preparation of 8 of patents, 69 Drump, brushed anyon, reproby selection of similar duction of 60 un tone, reproduction of by half meanindfulness of 66 tone process, 60 roncomparability of 79 wash, reproduction of 60 presentation of 82 Dryburth, 59 groups, 69 ·Index 107 Enten dutid 33 Heading Ichtion of your mmodmoritu Hents, cal atteo Hipon o portal structural ou Eorta cetacing d. 10 Enfant See Photoessante Herytur tot klation of, 46 Ecria, mremat ot, Epos, Equales dopod mostoh LY 100 Ideat patients ancealment of "Excluso sectorlot Horredor A So alto Do STATO, starostire terror Eraloxotal casato span n Photo Portiera of structuulut, 6 greaty md Rara Experimental procedure descrit tanto o math for bogoy Elroute to tak, 80 cost 60,0.61 met to yricted fredaction of Fouva-cuk tersia tak, 10 abort of artebongoth writica lec, 76 Food company cod of reproduction of, 60, for wine 14 empeza Frame, Folo murher can oh, lot bades, i factors atrocem preparation of, ) fute of soft mata ors, 99 Folerup Kuder, Was . 67 Fontos, ru-cobx 63 Fett 19 pose of 45 Fow-ci proas 63 hamtoot. 160 kontro ol, 16 Fitchen def. 6.77 Frequency anontuous, 73 and foc, 45 bume for 19 Fot.com Worttitog. 63 ud for 4 flaconent of (dispod 5 de of kattenne 45 Gott TTT 1 mailomity 45 Cowok Sir Erbe, toute kepada fer 57 na 11.19 * orname Gratur tips for botocopintot. Graph, 6. Sa tho Daro, ott month Jentry of Nations in eta (, doo, 15 un kr 63 mendo o Goede ser lateicar, pol wedi bod cars, u Fanar tid for 43 protica wrotih Hurmione to Fhotoenproof 9 Hoibhelp mes of 6 Halott balcopa Macarthe Sce Prostat MOTO Kurich Hal is, med 3 UD hairp AUS ruration of pol metr ·108 Index Illustrations, return of by publisher Initials, patients use of in care 59 reports, 32 saling of size of (diagram) 46 Inserts, tise of with illustrations, 45 screen Overlays for 46 Insurance for ustrations, 59 selection of 43 Intentallonal List of Cartes of Death Size of 45 23 68 space for 43 Introduction as part of structural tone, combination with line draw outlme 5 ings, 59 It, 20 methods for 45 Indentations, marginal on illustrations, used for labeling. 57 JARGON 14 Indention of paragraphs, 33 use of in index, 41 Inder Catalogue of the Swoon hoy letters, C of on Illustrations, S6 General's Office 3 Indexer chancteristics of 37 Kodachromc illustrations, costi ol Indexes, 37-41 reproduction of 62 abbreviations in, 39 reproduction of as halftones, 64 adjective us catchwords in, 39 alphabetting of conjunctions in, LABELS for illustrations, 45 51 53 40 consistency in terms used in 35 anatomical adjactive m, 39 guide lines for 57 boldface type in, 40 hand-lettering for 57 capitalisation in, 38 headers for 57 cards for 37 lettering for 57 atchwords in, 39 merginal 56 adjectives as, 39 placement of (diagram) 55 crot references from 39 smi mechanical lettering for 57 ommssion of in consecutivo site of lettering in, 45 entries, 41 symbols used in 55 cross references in, # terminology for 55 indentlons in, 41 type ined for 57 loglaal varunts as entries in 39 uniformity in, 45 nouns us entra a, 39 for photographk prints, 31 om son of Articles end for roentgenograms, 53 the ... 38 Laders for labels on illustrations, 57 preparation of by suthor 37 Legends for illustrationL 57 from page proofs, 38 Length of artick, 7 by publbher 37 Lettering for Labels on illusin reference tymbols in 39 trons, 45 S6 selection of headings for Line drawings, 43 59 specual type frees in 39 on pavings, 59 subentries in, placement of 3.8 Lincar measures, equn alents in, 100 subordinate subjects in, cross refer Line for chants and graphs, 63 axco from, 38 Literature, gunie to, 3 termunology in, congstency of 38 review of atation of in brblog sphy 30 varant formi u antries in, 39 summary and conclusions in, 17 Infinitives, split, 22 xarch 3 2 typing of 41 ·De, Index 109 Lithographic aim for bec caps N (frumber of obxerittom), 78 A (depes of freedom), 76, 77 Logal wants, of, bo odes, 19 Memes, personal o babets for dis a, etc., 23 Nepthe results, toportiog of 17 MAMI condensation of 9 Nepathes, photopraphies, como draft of, Anal, 35 too or pod og 49 bir, 7 Noorendaturg media, thodard. form of for roundsdon (or pub tatlog or 23 Lcttion, 34 Noocoonpanabiity of record 79 henfiop , from oom, 6. Notas, recording of Thestrukom in. See Dlar nouns * adfecte, 21 gerlok Matrero Photo Cance In sumber with serta reptes etc. md modern 30 kogth of, 7 cotnes binoku 19 orporation of 3 10 Number troncot of pouns with pepat Dunbaring of 35 arbs und modifics b, 20 proportion and balance of, 9 Numberins of pures, 35 quoted materal in, 3 Numbers Atomis, table of 95 rdeco n. 3 0,94 pops cotry of, for der, 38 buon af, 33 ASU,9 Acond, 11 Oct Titok 11 durd, 17 Observations, number of, 78 spool problemos la contraction Esting of tn tablet, 33 recording of importance of mi myke of 11 form system for 3 tablomt Otosalons, checking for 10 Largon on a tradot for pboto from date, Deccity for explana mpaving. 59 tion of intables, 33 Tomy Wentrty of patients in Dot a reference to author 20 photopipha, 38 Open intcr, aurt, arrihmcuc, 70 Order of presentation, 10 mntables, 3 Durint combidabon plate, 60 dontom from, 13 Anboth halftoo wopning, 60 urbation from, 70 Orining of sticks product, TS Osrbyl color 45 Measure, abbinations for % Lone lechino for 46 and aches, equh alat, 100 Mieden 70 Madal nomreodatert, standardiz (probability, 74 76, 77 tion of, 3 maak for chiar lest, 103 ring data on, 9) Metre dos equalents of, 101 Prammbers, coutry of lorder 1 phis and mat, equivalents Paper, mat, nambarios of 35 proofs, we of, toodete of 100 Mock, 70 Tapet too for stolp. 35 Malba aprennent ube rod u factor secting prepart with our ton of Destinations, 4) Mouring of the cons, 13 typed for pbotopiphile prints, 31 hungrate, construction of 17 ·110 Index Paragraphs, indention of 35 Photographs, pants of paper for SI Pathogenais as part of structural stre for $1 ovulins, 6 reproduction of ty halfionc proc Pathology As part of structural outlins, 5 setouching of 3 Patrenta, identification of in ase trimning of 45 reports, 32 Photomicrographs, 51 See also Il photographs of masking Identity Instration and Photographs In, 38 retouching of 33 permission to publish, 58 Pre diagrams, 87 Patterns, Ben Day 59 63 Plurals, 20 mechanical for cograving, 59 POLAL, central, 70 Pearl, Raymond, on blas in som Prepositions in alphabetizing. 40 pling, 67 Prints, photographk, preparation Pearson, Karl, as originator of chi of 31 square tot 74 Probability 74 Pen and ink drag for en grav tale for chi square tast, 103 ing. 59 Problems, specul, in construction Pencil drawing, reproduction of 19 of manuscript, 29 Percentages, 81 Procedures, experimental descrip- Perenission to publish Tlustra tion of 29 tions, 38 Product, mean, 78 photographs of patients, $8 Pronouns, 20 to quote, 32 Proofreaders marks, 96 Personal name as Labels for discose, Proofreading, 35 Proofs, changes and corrections to 35 Photocacra vings, 59 errors in, responsibility of author cost of 62 for 36 halone, 60 marting of 35 mechanical patterns for 59 maps, use in indung. 38 media for 39 reading of 33 Photographer clinical importance Proportion and balance of article 9 of 47 Protocols, description of 29 Photographs. See also Illustrator Puddery 11 Plotomicrographs and Roemi Punctuation, 24 GOTTANUT cropping of, 45 distance from camera to subject Quarterly Cwmtanire Inder Alecta ( 3 distortion in, elimination of 49 Questionnaire as method of collect extrinsic objects in 48 ng samples, 67 lichdog in, 48 Quiller-Couch, Sir Arthur on oht negauna of allmination of spots sure atiting, 12 from, 49 Quoted matter acknon leugment of of patients, permason to pubsource of 32 Inst, 58 paraphrying or 12 preparation of 47 prints of Labeling for 31 markny or for corpung, 31 RANDOYNs or umpte, 67 mounting or si Rates, cupression of in tables, 81 GC, 23 in, 49 ·index 117 Rao, waptation of Huty to Lumpling purpor Of, 66 peat propos Samden Coupeny VB, Wola Reading op importance of 3 Probleke peamed by 94 Kerror tobak, din lockey, Sak, rection for rectingular sta dag, 86 References, completa, 3 Sarter degree of, 70 OC, 93 Static terminolop it of, 22 mentry of one * Scope of det, 7 method of cute. 31 Storects, construction of 17 personal rest of, openbos, 26 recording of 4 Sipicance, chil quart to for 74 Kyk for Istic , 91 of data. 69 Relations.bop, despree of meerse Sh. 14 tent of 77 Sources, bobro priphis, 3 Reports clical zod kontory of gootabno, doorlopement hod..my of ol, 12 Reprecatne pumpk, 66 Spalme 4 Reproduction of copy Sput infinhhes 22 titions, 32 Square Anlah halftone con Ings, 60 Routte, deteripuon o, 30 Standard dobion, 70 nepatne, reporting of n akalarion of, 71 Rocs of beanture, citation of ordadoo of corcent of bblogrupy 30 correlation, 77 Komary and conchrony m, 17 cupression of, 70 Ration, cogol, 35 brerpretation of, TZ duties of author , 33 Standard Aameachanre of Dirar 23 editoral, Noteptance by author Statistical analys, 69 of 4 capris. See Dictions, des for kad NoCopd, 11 methods, 65-91 Chord, 17 Sualtics, 1.91 Rok, yogare 10,0 the boot 65 Rococogums hospital, election of, 66 bben of refTe 00, 93 Lont dupa do TUS Kolactha of, 66 ul, adoparpop of metalon reduction of $2 reproduction of, 60 Sorpping hand, for firm metum. ratochnoy , 3 the 19 ut of 5) Stody poms of 4 StL, 11 Sus collected by quesit bolotnaptis, of the loc labes, 57 Roth, 67 publishek 34 ndoan 67 Subotics mindkey 1,40 Can. 66 Karton of, 67 Subjects, chole of! uno, 67 upract us fector SOL. 46 mintun lector m, 2 bei 0,67 utorduate, cross dators froon, mit 3 ·112 Index Subordinate pubjects, cross refer Type facts, special, in index, 39 ences from, in mder, 38 for labels on illustrations, 57 Subtitle, use of 23 Summary 25 "Summary w part of structural out UNITED States Bureau of the Census, linc, 6, 27 death certificates complied by 68 Symbols, referenct, use of in indez, 39 Synonyms, alds to selection of 19 VALENCE, chemal, table of 98 Cross references from, in Index, Vanability 70 38 Varlants, local use of as entries In Index, 39 Variation coefficient of 73 TABLES, 82 from the mean, 70 construction of 83 Verbs, agreement in number with expression of rates In, 83 nouns, 20 heading in, 85 Vital statistkes, international, com intellidbillty of 83 parison of 69 omussions of data in, 83 purpose of 82 size of 84 WASH drawines, reproduction of by sorting of data for 82 halftone process, 60 uso of avenges in 83 WB Saunders Company biblio frequency distributions in, 83 graphxc style preferred by 94 horzontal headings in, 85 We us reference to single outhor 20 Temperature, conversion tables for Webster s Dictionary of SJTOT 1477 19 102 Wdghts, abbreviations for 95 7 cndenxcy central 69 atomic, table of 98 Terminology consistency in in inand measures, cquivalent 100 dex, 38 Williamson, on obscure writing. 12 scientific, 22 Words, usage of 19 21 standard, for labels, 55 Writing, medical references on 93 Test, chi kare, 74 obscure, 11 interpretation of 76 Thesaurus, us aid for selection of X-RAY See Roentgomurami synonymı, 19 Title, selection of 25 Tono drawings, reproduction of 60 Yuu, on aut in using Hatistics, 65 ilustrations, combination with line drawing, 59 Topis, witable, obtaining Ideas for 1 Zivonpranings 59

·No text on this page.

·112 Index Subordinato subjects, cross refer Type facts, special in index, 39 ances from, in Index, 38 for labels on ilustrations, 57 Subultke, use of 23 Summary 25 Summary as part of structural out UNITED Status Bureau of the Censin line, 6, 27 death certificates complied by 68 Symbols, reference use of in index, 39 Synonyms, aids to selection of 19 VALINCE, chemkal table or 98 Cross references from, lo index, Vartability 70 39 Varants, logical, use of as entnes In Index, 39 Variation coefficient of, 73 TABLES, 82 from the mean, 70 construction of 83 Verbs, agreement in number with expression of rates In, 83 nouns, 20 headines in, 85 Vital statisties, intematonal, com- Intelligibility of 83 panson of 68 ombusions of date in 83 purpose of 82 size of 84 WASH drawings, reproduction or by sorting of data for 82 halftone proces, 60 use of everages lo, 83 WB Saunders Company bibliofrequency distributions in 83 graphic style preferred by 24 horzontal headings in, 85 We u reference to single author 20 Temperature, conversion tables for It obster s Dictionary of SJTION 17718 19 TOZ Welghts, abbreviations for 95 7cndetxcy central 69 atomic, table of 98 Terminology comistency in, in inand maasures, equivalent 100 dar, 38 Williamson, on obscure wriung. 12 soientific, 22 Words, unge of 19 21 standard for labels, 55 Writing, medical, reference on, 93 Test, chi square, 74 obscure, 11 Interpretation of 76 Thcours, 23 and for selection of X-RAY Set Rornigeroranis tynonyms, 19 Title, selection of 25 Tone drowings, reproduction of 60 YULE, on oint in tosing statestic, 65 Mustrations, combination with line drawings 59 Topk, sultabk, obtaning Ideas for 1 Znc engravings. 59

·No text on this page.