MEDICAL WRITING

MEDICAL WRITING

H L Dhar
Director, Medical Research Centre, Bombay Hospital Trust, Mumbai - 20.

INTRODUCTION
Primary purpose of any publication is to advance knowledge to communicate the results of the authors’ work to their colleagues and contemporaries. Sometime scientific personnel are under pressure to publish papers before getting job or grants renewed. Added to this, lack of training in scientific methods and sentiments of english style makes presentation of majority of write up not up to the mark inspite of good material.Purpose of this write up is to give a guideline to the inexperienced authors and to have informations at hand for others while writing a paper. Attempts have been made to highlight the important aspects of scientific paper writing which may serve some purpose particularly with the beginners. Even senior people some time are not careful in presenting data in its right perspective.
How to present your article

Before submitting an article to a journal an author should ask himself the following questions : Have I made any point(s) clear and understandable for the audience? Is it entertaining (or at least readable) as well as instructive? At last, but not the least, am I saving my time at editorial expense, but not attending to the details of presentation? If the answers to all these questions are favourable, the hopeful author may not be an Oslar or a Churchill, but his paper will be published!

The first step

For most doctors writing a paper is a fairly important step in their professional lives. It should therefore, be treated like other important steps. At the initial stage to many writing a scientific paper is a tedious duty and a dull job to perform. However, when the subject of an article is important it is not much to do with its dullness. If a man has something to say which interests him, and he knows how to set it, then he need never be dull. [2]

The search for information

The first question anybody outside the project will ask "what has been done before"? Good sources of information are leading articles in journals, review articles of the kind published in reputed journals. Current contents is a valuable source, quarterly journal of medicine is another source. One has to consult the index medicus of the last four to five years to fill in the gaps. The list of review articles published in the index every month may be short cut to getting references on a subject, while books such as year book and progress in series and medical annual may be used to have a final check that no important original work on the subject has been missed. Currently, a series of review journals are published giving upto date information of the subject which may be a short cut in the review of literatures e.g. current opinion published on various subjects. Next, one can approach the librarian with material already collected for further help; which has become a simple job with internet system available in many libraries.

Types of medical writings

Publications consist of many types, each type has got its own design.

Original articles : Authors who write original articles really do not have a varied readership in mind. On the contrary, they use concepts, language, symbolism and methodologic descriptions that will attract and impress the co-experts. Acceptability of the article for publication will be judged first by reviewers who are specialists and then by similarly qualified and hypercritical readers. Though clinical research is more glamorous and may appeal to the practical doctor, it is much less certain to yield suitable material than rat work. A carefully designed laboratory study should be guaranteed to produce original results for an MD thesis and a good piece of research could easily provide material for a PhD thesis. This may provide several papers published in conference proceedings and two or three original articles of reputed journals. [3]

Clinical research

Too many clinicians in different institutions spend years recording everything imaginable; for example, height, weight, age, sex, blood pressure, urine analysis, retinal appearances, ECG and chest X-ray from patients taking one or two established drugs. The findings of such a study are predictable, in most cases such work has been done before even if they differ from all earlier works, then very good reasons will be needed to convince the editor and his answers that the results are reliable. One should keep in mind that usually only positive findings are of interest.

Investigations should be kept as simple as possible, but studies of biochemical value in groups of patients must be properly constructed, so that the results are scientifically reliable. This usually means that the correct control group must be studied and a statistician’s advice is valuable here, and that the right ethical decisions are taken before one starts the work. Occasionally, original work of scientific interest is turned down by a good journal on the grounds that the study was unethical.

Drug trial

If preliminary screening of a drug is published there is a grave temptation to clinical trial. Poor trials are easy, good ones are time consuming and difficult to arrange. If the pilot study suggests that the trial is likely to give positive results, then speed is important as trial may be done in several centres at the same time. The first two or three papers on the subject will find much readier acceptance. A perfectly designed and scrupulously performed double blind cross over study is of negligible interest if two or three reports have already appeared. Any body undertaking a drug trial may consult the text book of pharmacology by Lawrence4 which gives a detailed concise check list of all the items that should be considered.

Case Reports

Case reports usually are best reported in a short letter to the editor. However, one has to spend a lot of effort searching references and writing up a long discussion of earlier cases of the same kind, but the chances of publication are reduced if the editor is reminded that large number of such cases have been reported already.

Rare conditions are however, seen rarely but an account of an unfamiliar condition is more reliable if based on several cases. An alternative to writing up the case at once is to present it at an appropriate meeting of a local medical society, when there is fair chance that other clinicians will come forward with similar cases and a more authoritative paper can then be written by several authors together.

Adverse drug reaction

Reports of side effects of drugs are frequently submitted to medical journals. The first thing the editor does is to check with the latest edition of the excerpta medica and side effect of drugs where every reported side effects of all common drugs is listed and if there is more than one listing, the chances of a single report being published are very poor. Even if the report is original, it should be kept short, what matters is that the adverse effect should be recorded so that others may look out for it. If it is serious then any evidence supporting the link between drug and danger should be quoted. But as with case reports single patient events rarely make more than letters.

New technique

Sometime a new technique is devised mostly for experimental purpose or for improved medical care. Articles of this type must be description of truly new inventions and that one has successfully used them. The article however, should be very short.

Selection of journals

Before writing an article one should decide where to send it and should go through a copy to see the pattern of presentation and then go through the instructions to the authors.

Among reputed journals published globally mention may be made of two popular British journals which publish papers from outside, the British Medical Journal and Lancet.

British Medical Journal (BMJ) is firmly rooted in clinical medicine, articles are both original and highbrow; most of them have fairly, obvious and immediate application to the patient. Section for original includes short reports (limited to 600 words, 5 references and a table or figure) and side effects. It also contains a medical practice section, which includes unsolicited review, articles and papers on the organization of medical care and the supplement, which may print, talking point; articles on medico political topics. Because it is committed to referring almost any article potentially suitable for publication, the BMJ takes a little longer times for publication. Articles are often regretted not for their poor scientific quality but because they are quite unsuitable for that journal as regards subject matter or treatment.

By and large, if one has some original work concerned more or less directly with human disease and the account is fairly brief without much abbreviation, it is advisable to try BMJ.

Lancet : because they usually do not believe in referring articles, the editors usually let you know very quickly whether they are about to publish the article or not and after acceptance they publish it within few weeks. The lancet has a section for middle articles which covers a wide variety of subjects, while many scientific findings may be published as long letters to the editor.

Lancet publishes very preliminary work, often in animals with potential clinical implications. This means that an important but exciting experimental work in animals if wanted to be put to a wide audience at an early stage, the lancet probably is the best to try first.

Two more journals of repute Brit. Journal of Physiology and Brit. Journal of Pharmacology and experimental therapeutics publish mostly outstanding experimental works.

Other journals

Post graduate medical journal, again from Great Britain publishes well written piece of general interest, that does not report original findings. British journal of clinical practice is another journal of interest with trying if the material is clinical including clinical trial. The quarterly journal of medicine contains a mixture of original and review articles. Journal of American Medical Association (JAMA) having largest circulation of any medical journal in the world publishes articles mostly from USA and Australia, New England Journal of Medicine, sometime does not publish work from outside.

Indian journals

Among the various journals published from India, standard ones have review committees and are referred to various referees who take quite sometime to assess. Once it is accepted publication also takes its own time depending on the papers already pending, such journals are J. of Indian Council of Medical Research (ICMR) publishing wide variety of scientific studies primarily aided by ICMR; Indian journal of physiology and pharmacology and Indian journal of pharmacology primarily publish experimental works. Ind. J. of pharmacology also publishes works on clinical pharmacology, however, Ind. J. Clin. Pharmacology and therapeutics primarily publishes clinical evaluation of drugs and review articles.

There are however, number of journals mainly publishing clinical works, original articles as well as review articles; reputed ones are Journal of Association of Physicians of India (JAPI) and Journal of Indian Medical Association (JAMA). There are quite a few specialised journals i.e. J. of Neurology of India, Diabetes, Journal of surgery, Anaesthesia, which publish works on respective fields. Others are journal of chest diseases, psychiatry, paediatric, allergy and immunology etc.

Review Journals

There are three types - those that publish only review articles, those publish articles proper and a few original articles or case reports and others that publish articles which are both a review of the literature and a report of a series of personally studied patients. The prototype of first category is the current opinion. The second type of review journal is typified by practitioner and the third type by the quarterly journal of Medicine, American Journal of Medicine and Annals and Archives of Internal medicine etc. The character of all of those particularly the first, differs considerably from the other types., that an author is usually given enough space to present his findings in the perspective of earlier ones. These articles are frequently the standard reference on a particular topic. To be accepted, therefore, an article of this type must deal either with many patients with a fairly common disease, studied from a particular aspect or with a rare condition studied intensively.

How to present an article

Simplicity and clarity are the features of good scientific writing. No body is asking to write great literature, but the meaning must be readily understood. Good points to remember are that doctors not working in the subject should be able to understand the article, clear thought can be expressed clearly; and a man with something of value to say has no need to pad out bore editors (who are likely to reject them) and bore their readers (who are unlikely to finish them). In other words, most writers are failing to communicate the object writing in the first place. [5]

The structure of a scientific paper is well established to accommodate the "why" of the project (the introduction), the "who", "when", "where" and "how" (the methods) and the "what" (the results). The discussion (the "so what") should be in three parts - discussion and explanation of results, disputation and disquisition (implications for practice and for future work). [6]

In any article there are five critical features which consciously or unconsciously will influence an uncommitted reader to read the article and the editor to accept it for publication. These are : the total length of the article, the title, the introduction, the first few sentences, the discussion and the summary.

Summary

Although most articles begin with a summary, this section is usually written after the rest of the paper. By this time one tired of the rest of the whole article only want to be rid of it. This may explain why submitted summaries are almost universally bad, whatever the authority or capabilities of the author. All too often an editor will find out more about the work from a covering letter or from the first four sentences of the discussion than he will from a summary.

Title

An ideal title to a medical article demands the impossible. The object is to give the reader as much specific and intelligible information as you can in as few words as possible.

The Introduction

Introduction should be short except when writing MD or Ph. D. thesis (because the examiner expect it) or a review article, indicate the aim and scope of the paper. This should answer "why the work was started, if this cannot be said in a few sentences it is unlikely to have been worth doing. Except for the review journals, a long survey of the literature, is almost always out of place in this section. Explain how investigations move forward from closely related previous work on the same subject.

Material and Methods

Unless the overall experimental design or theoretical approach is already obvious from the introduction, describe it in broad out line before giving details of the method. This should aim mainly to answer the question : what has been done? Detail is necessary only if the method is original, otherwise reference will do. Never mix comment with this section or with the results. The word ‘material’ should be used only in articles dealing with laboratory work or animal experiments. When the article deals with patients, the heading should be called ‘patients and methods’. You should never refer to your patients as case material.

Results

This section answers the question : what answer did you get? It should be presented in logical order chronologically or in the order of the complexity of the test. Try to put as many results as possible into tables. Numbers given in the text should agree with those in the table and figures. Finally the statistics; do not quote statistical jargon unless it is correct.

Tables and Figures

Many journals print some instructions to authors in each of their issues and usually quite definite about the way in which they want tables and figures, so if you intend to send an article to a particular journal you should follow any instructions it gives. The Lancet publishes a leaflet on ‘writing for the Lancet’ while BMJ publishes instructions to authors in the first issue of every year.

Tables should be used to present facts in the clearest possible way. They should not repeat information that is already in the test. The information that you finally put in to a table should be unambiguous. Variable measures should be spelled in full e.g. serum protein mg/100 ml, serum alkaline phosphatase, king Armstrong units etc. The editor may insert abrreviations if his house style uses them, but manuscript must be clear. Do not include units in the body of the table. If the paper has only one table, it should not be numbered and should be referred to it in the text as table. Otherwise, each table should be numbered usually in Roman numbers. Every table should be typed on a separate sheet and should have a title, patients should be described by letters or numbers, but not by their own initials.

Figures should also be used sparingly and for a definite purpose, which cannot be obtained with a table and most graphs and histograms can be better Most accepted system, is to name first three authors followed by et al for the subsequent authors. However, one must study the way the journal in question cites its references and give a full bibliography in the list of references.

Advices for Reprints; Acknowledgements

Many journals now ask for the name and address of the person to whom readers may write for information and for reprints, and its a good idea to include these routinely in any article.

One should not forget to thank people who have really helped including technicians, colleagues, whose patients have been studied and the patients and volunteers themselves if the study was an experimental one. A good rule is to thank anybody who has done more than the routine work for which they are employed.

Acknowledgement can be written on a separate page at the end of the paper before the references, unless the journal prefers them as foot note or elsewhere.

Who Writes the Article and How?

Most articles carry names of two or more authors; either each author can write the draft of one or more sections or one person can do the job. The latter method usually produces a better article. One solution is to give this task to the first authors (who has done major portion of the work) and make others particularly who did least work on the subject to collect upto date references, but it is important that everybody approves and signs the final version. For a post graduate student or a beginner, it is the usual practice that the senior author usually the guide makes the final write up which forms the part of teaching programme and his name should come last among the authors.

It is a sound advice to begin by concentrating on the title, the summary and the tables (how many, their designs; their content). This method will expose any errors or omissions in the work itself and provide a frame work for the rest of the article.

Finally one should remember that no article is likely to read well unless it has been through atleast five revisions. Good writing is deceptively simple; but best ones have to go though many drafts.3

Better English Style

Any text book of English grammar will tell the six rules of writing clear English; choose the correct word (i.e. anticipate as expect; assist as help; commence as begin or start; due to as owing to; demonstrate as show; greater as more; elevated as raised; excess of as above; approximately as about; reveal as disclose; limb upper as arm; limb lower as leg; large number as many; majority as most etc.), prefer the simple word to the circumloculating, prefer the short word to the long, prefer the word derived from Anglo-saxon to that derived from Roman and the use of active rather than passive voice.

Other simple advice is to use short simple sentences; write with nouns and verbs and not with adjectives and adverses and avoid the few obvious howlers such as floating particles and less than happy constructions such as separating a subject too far from its verb.

Using short simple words and sentences one can write in clear prose. Good prose is like a windowpane.8

Commas

Wrong use of commas is a frequent error and one of the more important sources of confusion. Commas cannot be used in defining clauses but must be used in commenting clauses; one of a commonest causes of flabby writing is too many words (e.g. instead of toddler age child, toddler can be used). Construction of a sentence is also important. The use of verbs where one will do is a common cause of dead writing.

How to Present Article

Care and time spent in the actual presentation of an article are well worth while. You must send the article to only one journal at a time, but the fact that a paper has been presented to a conference does not prejudice its chances of being accepted if it fulfills the journals criteria for acceptance which in any case is likely to be rewritten for publication possibly with additional material for which there was no time at the conference. However, in this case you should use the formula based on a paper read to the international conference (year on).

Covering Letter

Every article should be accompanied by a letter to the editor. It is always good to say in your letter why you think article should be printed. Many journals ask for a statement in the covering letter that the material in the article is not ‘already published’ and that it is not being simultaneously submitted for publication elsewhere.

What to Do If Article is Rejected

Once the article has been submitted, editors reply will state either the paper has been accepted or rejected or that it will be accepted or reconsidered if you revise it in ways suggested by the editor or referees. However, it is unusual for a paper to be accepted without queries of any sort. But if it happens you are lucky and wait until the proofs reach you except perhaps write a note thanking the editor.

More often than not your paper will be returned by the editor of the first journal you sent it to. They do this usually on the advice of referees mainly because the articles do not meet the criteria for publication or because they are unsuitable for their journal. If the editor says that the paper will be accepted or reconsidered if the changes are made, consider the suggestions carefully, modify or rewrite sentences or sections as necessary, retype the article before you return the paper to editor, but enclose the original corrected copy. On the other hand if it is a profile no or a sincere suggestion that you should try another type of journal, have another look at it, rewrite and send it to another journal, good work usually gets taken up by least one of the general journals.

Letters to The Editor

A general journal nearly always includes a correspondence column especially where the journal is a weekly. On an average week both the Brit. Medical Journal and the Lancet devote atleast 12,000 words to letters to the editor and Nature has a large correspondence section for the biological sciences. As such the chances are much better, rejection rate for letters being much lower than for formal articles. Almost all monthly journals in India have columns for letters to editor. It is better to try to publish preliminary data in this column if the results are good.

Letters that are infact short articles or case reports should be presented with the same care as a full version. Yet for some reasons, doctors who would never think of writing anything that had not been fully checked in an article will do so in a letter. A letter to the editor may be trickier than an original article. To make their points in 300 or 500 words (the usual length), most letter need some sort of formal structure, even though this is much less obvious than in a formal original article. This will make the task of writing a lot easier and help the reader as well.

State in your first sentence what you are talking about? Does it refer to a previous article or letter (in which case give the reference) or is it a statement about something in its own right? Next, state your point of view : are you refuting a previous claim or modifying it and is your evidence personal experience or other people’s published work? If you really need to give a case history, or use a figure, do so, be brief and confine to what is really relevant.

There is usually no need to send a letter to the editor with note requesting publication. Letters must be published quickly and editors realise this better than authors. However, if the letter does not appear in a reasonable time (which depends on the type of journal and the country of publication) there is no harm in asking why. But write polite letter rather than telephone. Nobody has the right to get anything published and letters to the editor are no exceptions. But, after all, editors are interested in good outspoken, fighting letters and your approach may be at fault if they decline to publish yours.

The MD Thesis

Unlike an examination which is essentially a test of knowledge, pursuing MD or Ph. D by thesis judges a candidates ability to investigate a problem in depths. The thesis must show a high standard of scientific and scholastic excellence and reveal not only the candidates knowledge of the subject but his powers of deduction and reasoning. This involves a thorough study of the literature, careful recording and analysis of collected data, followed by discussion and conclusions which show candidates complete command on the subject. [9]

Preparation

The way in which one should approach the thesis is little different from that in an article. Having decided on your subject you should start by asking : what am I trying to find out; what has been done before; and what is going on in this subject at present?

Before starting the work you should check thoroughly using review articles, reference books, index medicus and finally taking help of medline service if available.

Formal Structure

The easiest way of creating a structure is to go through several successful thesis and copy down what they have done.

As in journal articles, never mix comments with facts. This is why so many subheadings may be needed in the discussion section, so that you may consider each aspect of your findings in turn. A typical order might be as follows. [3]

Title page : title, your name and degrees and the year.

Outline of work done : Object of the work and the method used (not exceeding 60 words).

Review of literatures : This pertains with survey of other publications in considerable depth, with everything relevant mentioned, even if some of the previous work was wrong.

Material and methods : This includes details of definition and methods used and should be described in full. Patients studied and case histories are given here.

Results : Presentation of data in chronological order or in order of complexity, this is the one place in medical writing where nobody minds if you repeat the results in the table form and in figures.

Discussion : State briefly what you set out to do, and what you found, then discuss each topic under a separate heading.

Summary : A fairly full (400-600 words) account of your findings and conclusions.

Conclusion : A brief summing-up, putting your present work into perspective.

References : Authors (upto three and then et al); title in full, journal, year, vol. page (first and last), cited reference should include the source.

Lastly the final version must be read by several people including the guide. You must not leave any obvious answerable questions unanswered otherwise an examiner may raise them or thesis may not be accepted. The copies must be immaculately bound, in the style prescribed by your university. Once your thesis has been accepted you have a fairly rich source of journal articles and papers for conferences.

Papers At Meetings

Most medical meetings are dull. Many of the doctors who regularly attend them do so out of a sense of duty to meet colleagues or to start their holiday cheaply. [3]

You should know by heart what you are going to say, yet be ready to introduce new sentences on the spar of the moment to deal with points made by other speakers. However, nervousness mostly among the beginners or real lack of time for learning one may some time have to read a script.

The Audience

Remember you are speaking on deaf morons. [10] It emphasizes that material which to you was familiar, even stale, was to your audience new perhaps difficult. If so speak slowly to the extent of not exceeding the time limit.

The Paper

The entire pattern of your paper should depend on what you are asked to do. If you have to talk for an hour, then the paper can be of the form : Themes, exposition, development recapitulation etc. The emphasis here should be on repetition, driving your points home various ways. But most contributors at the conferences are limited to about ten minutes, one of these is lost at beginning (people talking, coughing and coming late); another is taken up at the end, summarising, what you have said and yet another may be lost in turning the lights up and down and in waiting for the next slide. So you have only seven minutes to get your ideas across. The BBC recommends that only one new idea should be introduced every three minutes. [3] Only by careful planning you can make the presentation worthwhile and your paper a real contribution. Most of the best speakers who appear to talk spontaneously at meetings spend hours in detailed planning rewriting and rehearsal. Whatever your audience, you should devote much time and effort to what you want to say.

Delivery should aim at adaptability to audience (maximal speaking and limited writing) with arresting introduction (no writing) and visual aids (relevant slides/or minimum table and figures). Jokes (minimum) may be desirable and the delivery should be over little before time.

Contributing Chapters to Text Books

Most text books are now written by several authors and if you are asked to write a chapter or two, ask for a careful brief. You need to know whom your message is aimed at (undergraduate or post graduate, general or specialist audience) and to know exactly what the other contributors are going to deal with. Most editors now send detailed frame work, length, content etc. including maximum number of references as well as tables and figures allowed and the dead line of submission. There is nothing ruder in medical writing than being two years late with a chapter that quotes references upto the month. [11]

Medical Ethics

Research combined with treatment is ethical if it is justified by its therapeutic value for the patient. It cannot be ethical, if it entails withholding the best treatment from the patient. It is justifiable to conduct a controlled trial of a new drug if, first, there is established evidence that the new drug is likely to be least as effective as conventional treatment; and secondly that the control patients are treated with orthodox drugs.

A trial that uses placebo treatment should be done only with the consent which should always be obtained if possible. Occasionally it is not consistent with good management to ask for consent from a patient or his relatives as this might distress them.

Consent

If the research is of the kind that requires full and informed consent, it should be obtained in writing and the fact should be recorded in the paper describing the work. Should patients or subjects refuse consent, their fact should also be recorded and if the work is ethically objectionable then it should not be done.

Patients must be fully capable of assessing choice. Children and mentally defective cannot give consent (nor can others on their behalf) to any procedure that is unlikely to be direct benefit to them. Ethical controlled trials using placebo therapy are almost impossible in children.

Any element of risk must be clearly explained to the patient, nature and purpose of all procedures described. The subjects must be free to withdraw from the experiment at any time. Even if the risks are accepted by the subjects, the importance of the objective must be in proportion to the risks, based on scientifically established facts to justify the procedures. [12]

REFERENCES

Rayan P. Cited by medical journal of Australia 1973; 1 : 779.

Asher R. How to present your article. BMJ 1958; 2 : 502.

Stephen Lock Thornes’s better medical writing, Pitmen Medical, 2nd Ed. 1977; 6.

Lawrence DR, Bennett PN. Evaluation of drugs in man, therapeutic trials. Clinical pharmacology Churchill Livingstone 8th ed. 1977; 44-65.

Luttikhuizen F. The ins and outs of scientific writing Microbiologia 1996; 12 (2) : 477-80.

Fraser HS. Writing a scientific paper. West Indian Med J 1995; 44 (4) : 114-24.

Bradford Hill A. Logical order for a scientific paper. BMJ 1965; 2 : 870.

George Orwell. Collected essays, Secker and Warbug, London. 1968.

Williams WO. British Journal of Medical Education 1969; 3 : 171.

Pyke DA. How to write papers. BMJ 1970; 1 : 420.

Wilson Edmund A. Atlantic monthly. 1935; 155 : 67.

Medical Ethics Today. In practice of philosophy consent and refusal from BMA’s ethics. Science and information division, Tavistock square, London. 1996; 1-35.

To Section TOC