THE NATIONAL MEDICAL JOURNAL OF INDIA

CorrespondenceVOL.4, NO.4 205

A no-smoking policy in hospitals

Sir-The Environmental Protection Agencyof the United States of America considersenvironmental tobacco smoke to be a humancarcinogen. Although cigarette smoking iswell known to be a causative factor in coro-nary artery disease, chronic bronchitis, pep-tic ulcer and oral and lung cancers, surpris-ingly the habit continues to be fashionableamong medical students and hospital staff inIndia. There is, therefore, an urgent need toadopt a 'no-smoking' policy in hospitals.

The best way to put this into action wouldbe to print clearly, on the patient's admissioncard, the message that the hospital is 'Smoke-free'. Smoking during outpatient clinic atten-dance should be forbidden and public healthworkers should counsel smokers who attend.

A hospital no-smoking policy will com-municate a potent messa&e to patients andvisitors that smoking is harmful to health andwill protect non-smokers from the substan-tial hazards of tobacco smoke. All physiciansshould be role models and either give upsmoking or never smoke in public.

2 July 1991 Arvind JadhavHammersmith Hospital

London, UK

Routine chest X-ray

Sir-We have a few comments to make re-garding the study methodology of Rajani etal,' Any study which evaluates the 'neces-sity' of a 'routine' chest X-ray must takespecial care to be free from expectation bias.An expectation bias occurs when radiographsare read with the knowledge of the clinicaldiagnosis. The study must assess the 'inde-pendent contribution' of the chest X-ray, thatis, the relevant diagnostic information notsuspected from a prior clinical evaluation.The study must also assess whether the infor-mation obtained from the chest X-ray re-sulted in changes in patient outcome; theoutcome includes peri operative complica-tions or changes in surgical, anaesthetic ormedical management. At least two of thesemethodological criteria are not adequatelysatisfied in the study. An expectation bias isacknowledged by the authors themselves andpatient outcome has not been assessed. Al-though two observers interpreted the films,

no data on inter-observer variability has beenprovided. Such data would add to the value ofthe study.

Further, group II in their study includes1000 individuals with 'symptoms or signssuggestive of chest disease or who had sys-temic disease which might be associated withradiological abnormalities'. No further de-tails of the exact indications for chest X-rayin patients in this group are provided by theauthors. This assumes importance becausewhen a chest X-ray is obtained in the pres-ence of chest symptoms or signs it may nolonger qualify for a 'routine' chest X-ray. A'routine' chest X-ray has been defined as onethat is done to screen for unsuspected chestdisease or for use as a baseline study." Forinstance 10% of the patients in group II hadradiological evidence of an inflammatoryconsolidation. If these patients had clinicalsigns suggestive of consolidation, as is likely,the chest X-ray obtained in them cannot beregarded as 'routine'. These must be consid-ered as 'diagnostic studies? and their inclu-sion only serves to confound matters.

Any study addressing the issue of 'neces-sity' of a routine chest X-ray must combinedata on 'diagnostic yield' with informationregarding its 'incremental diagnostic value'and its effect on 'patient outcome' or 'man-agement decisions'. A diagnostic yield of41 % in group II does not in itself indicate thata chest X-ray is 'necessary' in this subgroupunless the information obtained from thechest X-ray led to changes in patient manage-ment, an aspect not studied by the authors.The incremental diagnostic utility of the chestX-ray in this subgroup also needs to be as-sessed before interpreting the results.

The authors address the issue of 'diagnos-tic yield' of a chest X-ray and not its' 'neces-sity'! An investigation tool with a low diag-nostic yield can be considered 'necessary' ifit is highly relevant to patient managementand an investigation tool with a 'high' diag-nostic yield may be 'unnecessary' if it has noincremental diagnostic utility over otherexisting diagnostic methods, or if it adds littleto patient management and outcome. Forinstance, a routine preoperative chest X-rayin a patient undergoing thoracic or upperabdominal surgery may have little diagnosticyield; it may, however, be considered neces-sary because of its possible role in postopera-tive patient care when it serves as a baselinefor comparison with chest X-rays obtainedsubsequently.

In conclusion, the question which the au-thors have so painstakingly answered is thediagnostic yield of the chest X-ray in certainspecific clinical groups. Further analysis wouldbe required to settle the question of the'necessity' of a 'routine' chest X-ray.

25 July 1991 R. S. VasanS. Seshadri

Departments of Cardiology and NeurologyAll India Institute of Medical Sciences

New Delhi 110029, India

REFERENCESRajani M, Sharma S, Sonekar P. Is the routinechest X-ray necessary? Natl Med J India 1991;4:24-5.

2 Hubell FA, Greenfield S, Tyler JL, Chetty K,Wyle FA. The impact of routine admission chestX-ray films on patient care. N Engl J Med1985;312:209-13.

3 Tape TG, Mushlin AI. The utility of routine chestradiographs. Ann Intern Med 1986;104:663-70.

II

Sir-Rajani et al,' have done a commendablejob by analysing radiographic abnormalitiesin a large number of patients undergoingroutine checkups. A diagnostic yield of only3.1% in asymptomatic patients indicates thatin these patients the procedure is not cost-effective. In a similar prospective study? ofpatients who presented to a casualty depart-ment with acute exacerbation of chronicobstructive airways disease we observed thatafter a thorough clinical assessment had beenperformed the utility of routine chest X-rayswas limited to a small subset of patients. Onre-analysing our data we found that carefulclinical assessment was 100% sensitive and90% specific in predicting lobar and segmen-tal lung consolidation in the sub-category ofpatients with complicated chronic obstruc-tive airways disease (COAD). Furthermore,the positive predictive value of such an as-sessment was 80%. The sensitivity, specific-ity and positive predictive value of clincalassessment in making a diagnosis of uncom-plicated COAD was 74%, 86% and 90%respectively. However, the routine chest X-ray is helpful in changing treatment in a smallsubset of patients who have a pneumothorax.These findings are also in agreement withanother recently published retrospective study.'

It would have been helpful if the data

206

regarding sensitivity, specificity and predic-tive value had been presented and to knowwhether the study was retrospective or pro-spective.

30 July 1991 Anoop MisraPrasoon Jain

D. MukhopadhyayDepanment of Medicine

All India Institute of Medical SciencesNew Delhi

India

REFERENCES1 Rajani M, Sharma S, Sonekar P. Is the routine

chest X-ray necessary? Natl Med J India 1991;4:24-5.

2 Jain P, Misra A. Routine chest X-ray in chronicobstructive airways disease: A myth. NZ Med J1990;103:163.

3 Sherman S, Skoney JA, Ravikrishnan KP.Routine chest radiographs in exacerbations ofchronic obstructive pulmonary disease: Dia-gnostic value. Arch Intern Med 1989;149:2493-6.

Indian Medical Institutions

Sir-The article by Sudhakar Reddy et al,' onResearch in Indian Medical Institutions wasinteresting, thought-provoking and timely andwe fully agree with the views expressed.

The authors have included the researchoutput from institutions like the Tata Memo-rial Centre, the Sree Chitra Tirunal Instituteand the Chittaranjan Cancer Institute togetherwith the research output from medical col-leges. However, these are specialized centreswhich have few departments and deal withfewer patients than medical colleges. Theyshould have been considered separately.

We are surprised that the Regional CancerCentre, Trivandrum does not find a place inthe list. This institute was established in 1981by transferring some departments of the Medi-ca~ College, Trivandrum, which were mainlyconcerned with cancer care. The aim was todevelop an institute of excellence, in bothteaching and research. We have divisions ofradiotherapy, medical oncology, surgicaloncology, paediatric oncology, dentistry, cancerresearch, pathology, tumour registry, cancerepidemiology and clinical research andcommunity oncology. Annually, we get about5000 new cancer cases and 40 000 outpatientvisits.

The year-wise record of our publicationsis listed in Table I. There has been a steadyincrease in the number of publications overthe years. The average number of publica-tions for the lO-year period (1981-90) is 15,but for the period 1981-8 it was 11.4 and

THE NATIONAL MEDICAL JOURNAL OF INDIA

Table I. Publications from the Regional CancerCentre, Trivandrum (1981-90)

Year Publications TotalIndian Foreign Books!

journals joumals Proceedings

1981 7 0 0 71982 3 0 I 41983 5 1 I 71984 3 1 I 51985 3 3 0 61986 7 7 4 181987 6 12 0 181988 4 18 4 261989 9 16 3 281990 4 18 9 31

Total 51 76 23 150

during 1989 and 1990 it has risen to 29.5. In1991, 26 articles have been submitted forpublication.

Our centre insists on a minimum numberof research publications for selection to vari-ous posts. We are also considering monetaryincentives for research activity.

8July 1991 R. SankaranarayananCheri an VargheseM. Krishnan Nair

Regional Cancer CentreTrivandrum, Kerala

India

VOL. 4, NO.4

Though the Medical Council of India"recommends that 'published research workshould be given due weightage in makingappointments to the teaching posts', this isseldom followed in practice. At times, eventhe candidates are far superior-both in thequality and quantity of their research achieve-ments-than the so-called experts who arecalled to interview them.

2 July 1991 Kush KumarDepartment of Orthopaedics

Krishna Institute of Medical SciencesKarad, Maharashtra

IndiaREFERENCES,I SudhakarReddy K,SahniP, PandeGK,Nundy S.

Research in Indianmedical institutions.Natl MedJ India 1991;4:90-2.

2 Kumar K.Critical evaluation of' original papers'published in Indian Joumal of Orthopaedics.Health Services 1987;4:11-16.

3 Kumar K, Singh P, Nayak M. Indian orthopaedicresearch (analysis and guidelines). Indian JOrthop 1989;23:104-11.

4 Medical Council of India. Recommendations onqualifications required for the appointment ofpersons to the post of teachers in medicalcolleges and attached hospitals for undergrad-uate and postgraduate teaching. New Delhi:Medical Council of India, 1982:2.

III

Sir-Let me congratulate you on the serviceyou have rendered to medical education andresearch in India by publishing the article'Research in Indian Medical Institutes'.'

I notice that you have used the CorporateIndex of the Science Citation Index whichlists the papers it carries according to theirinstitution of origin and ranked them in adescending order. However, you have notmade a list which takes into account theimpact of the papers published by each ofthese institutions. Sometimes, even thoughthe number is small, an institution's researchmay have a greater impact because of itsquality.

I suggest you make a rank list on this basisat least of the top 20 medical institutions.

8 July 1991 J. Prakasa RaoDepartment of PhysiologyChristian Medical College

Vellore, Tamil NaduIndia

REFERENCE1 Sudhakar Reddy K,SahniP,PandeGK, NundyS.

Research in Indianmedical institutions.Natl M edflndia 1991;4:90-2.

REFERENCEI SudhakarReddyK,SahniP, PandeGK,NundyS.

Research in Indianmedical institutions.Nat! MedJ India 1991;4:90-2.

II

Sir-I read the article by Reddy et al,' withconsiderable interest. The findings need to bepublicised widely as they expose the unfortu-nate state of many Indian medical institu-tions.

I have also been interested in the qualityand quantity of research work in orthopaedicsin India and obtained an almost identicalpicture.i-'

Both the recommendations made by theauthors-to forbid private practice and to linkresearch achievements with selection andpromotion-are very important. I would alsosuggest that the system of 'honorary' teach-ers which exists in some of the medicalinstitutions be abolished. None of the top sixIndian medical institutions who have pub-lished more than 50 articles in a year inScience Citation Index covered journals 1 havehonorary teachers.