Hepatitis B antigen and viral hepatitis type B in India*

VIRAL HEPATITIS Hepatitis B antigen and viral hepatitis type B in India* T. JACOB JOHN,' ROBERT H. CARMAN,2 & PETER G. HILL 3 Surveys were conducted to determine the occurrence of overt hepatitis manifested by jaundice in groups ofhospital patients who had survived their initial illness. Ofthose who had been given blood that had not been screenedfor the presence ofhepatitis B antigen (HB Ag) 9.4 % reported that they had hadjaundice during the 6 months following transfusion. During a similar periodjaundice was reported by only 1.1 % ofthose who had received bloodfound to be negative for HB Ag by the cross-over electrophoresis test. This difference is highly significant. Jaundice was reported by 25.5 % of recipients of HB Ag-positive blood. Of a control population of hospitalized but non-transfused patients 2.2 % reportedjaundice within 6 months of hospitalization. The risk of overt hepatitis from HB Ag in transfused blood is high in southern India and is similar to that in temperate countries where the occurrence of antigen carriers, andpresumably of immune status, are lower than in the tropics. In temperate countries the risk of developing hepatitis following the transfusion of blood contain- ing hepatitis B antigen (HB Ag) is about 500% or more (1, 2). Therefore, it has been recommended that HB Ag-positive blood should not be transfused. The prevalence of the HB Ag carrier state in apparently healthy individuals in many tropical and subtropical countries is about 5-20%, in contrast to 0.1{0.60Y in temperate countries (1). It is not known, however, whether the increased prevalence of HB antigenaemia results in a higher occurrence of im- munity and thus reduces the risk of hepatitis in transfused subjects as suggested for the population around Harlem Hospital (3). Before the adoption of routine screening of all donor blood for HB Ag and the subsequent elim- ination of positive blood we sought to document the risk of overt hepatitis following transfusion of unscreened blood and the effect of the elimination of HB Ag-positive blood on the incidence of such hepatitis. * From the Christian Medical College Hospital, Vellore, Tamil Nadu, India. 1 Professor of Microbiology and Chief, Enterovirus Labo- ratory. Present address: Department of Pediatrics, University of Arizona, Tucson, AZ, USA. 2 Professor of Pathology and Head, Department of Clinical Pathology and Blood Bank. 3 Reader and Acting Head, Department of Clinical Biochemistry. MATERIALS AND METHODS The Christian Medical College Hospital is-situated in the town of Vellore in tropical southern India. A description of the blood donors and the technique of cross-over electrophoresis (COE) used for the detec- tion of HB Ag have been described previously (4). Information on the occurrence of jaundice or death within 6 months of transfusion was collected both by postal questionnaire and by house visits. All hospitalized patients who had received one or more units (500 ml) of blood between November, 1971 and July, 1973 were listed and assigned to the house visit interview group if they lived within about 20 miles of the hospital, or to the postal question- naire group if they lived elsewhere. Those who had died in hospital were not included in the list. Patients in the house visit group were visited about 6-8 months after discharge from hospital and the number of cases of jaundice or death that had occurred within 6 months of transfusion was docu- mented. A diagnosis ofjaundice made by responsible members of the family or by any medical -attendant, based on characteristic discoloration of the urine and conjunctivae and other symptoms and signs of hepatitis, was accepted for the purpose of this survey. Patients in the postal questionnaire group were sent a reply-paid questionnaire 6 months after their discharge from hospital. 3307 - 617 BULL. WORLD HEALTH ORGAN., Vol. 51, 1974 T. JACOB JOHN ET AL. Table 1. Results of the postal questionnaire survey Population No. Replies Jaundice Deathsent out received N % No. % transfused with unscreened blood 1 000 520 46 8.9 46 8.9 transfused with HB Ag-negative blood 600 247 4 1.7 19 7.7 not transfused 1 000 613 18 2.9 47 7.7 In addition to the study population, a control population was similarly surveyed. The control population consisted of 1195 patients who had been admitted to the hospital between April, 1972 and March, 1973, and who had been in hospital for at least 1 week but had not received a blood transfusion. Prior to July, 1972 blood donors were not accepted if they had had jaundice within the previ- ous 5 years, or if their icterus index was higher than 6 units. During this time some samples of blood were tested for the presence of HB Ag subsequent to their transfusion. Commencing in July, 1972 all units of blood were screened immediately after collection and only HB Ag-free blood was used for transfusion. RESULTS The number of patients to whom postal question- naires were sent, the number of replies and the number of reported instances of jaundice and death within 6 months of transfusion (study population) or hospitalization (control population) are presented in Table 1. The difference in the incidence of reported jaundice between patients transfused with un- screened blood and those transfused with HB Ag- negative blood was highly significant (P = < 0.0005). However, the difference in the incidence Table 2. Results of the house visit survey No. Jaundice Death Population visited No. % No. % transfused with unscreened blood 207 22 10.7 20 9.7 transfused with HB Ag-negative blood 114 0 - 4 3.6 not transfused 195 0 - 6 3.1 of jaundice between those given HB Ag-negative blood and those given no transfusion was not statis- tically significant (P = > 0.2). The rates of reported mortality were similar in all 3 groups. The results of the house visit survey are presented in Table 2. The incidence of jaundice and of death among the recipients of unscreened blood in the house visit survey was not significantly different from that of the postal survey (P = >0.2). How- ever, no jaundice was reported among over 300 people who were either not transfused or were given HB Ag-negative blood. The mortality rates in these 2 groups were lower than in the other groups followed up. Of the people surveyed, 78 were known to have received HB Ag-positive blood, 68 in the population surveyed by postal questionnaire and 10 in that surveyed by house visit. Reports were received on 33 of the former group, of whom 8 (24%) had devel- oped jaundice and 3 (9.1 %) had died. Of the 10 who were visited 3 (30%) had developed jaundice and none had died. DISCUSSION The results of these surveys reveal a much higher risk of jaundice associated with the transfusion of unscreened blood than with the transfusion of blood tested and found negative for HB Ag by cross-over electrophoresis. For convenience, the results of the 2 surveys have been combined and are presented in Table 3. About 25% of those who received blood containing HB Ag developed jaundice within 6 months. The incidence of jaundice during the follow-up period in those who received blood with- out detectable HB Ag and in those who received no transfusion were very similar. A hospital based follow-up with laboratory con- firmation of hepatitis was not considered feasible since the patients came from different parts of the state or from outside the state. If hospitalization for 618 VIRAL HEPATITIS Table 3. Combined results of both surveys Bloodtransused No. of patients JaundiceBlood transf followed up no. c% HB Ag-positive 43 11 25.5 Unscreened 727 68 9.4 HB Ag-negative 361 4 1.1 No transfusion 808 18 2.2 jaundice had been required they would have been more likely not to report jaundice than to return to Vellore. Moreover, most patients with jaundice in southern India seek help from indigenous practi- tioners and would be unlikely to enter hospital unless they were severely ill. House visits to all the patients would have been ideal, but since most of the patients lived some distance from the hospital this was not possible. It is possible that the replies to the postal questionnaire may have been biased towards positive answers (jaundice, death) whereas follow-up by house visit survey would not have led to such bias. There may be other differences between the 2 populations; it is possible that those who lived some distance from the hospital were more sick to begin with than those who lived nearer the hospital. This hypothesis is supported by the fact that whereas 13% of the population in the postal survey had received 6 or more units of blood, only 4% of those visited in their homes had received as many units. In studying mortality, it was in many instances very difficult to ascertain whether death had been caused by hepatitis or whether jaundice had been overlooked in dying patients. Although instances of death as a result of HB Ag-positive hepatitis follow- ing transfusion were known, such information was very incomplete in the groups surveyed. For this reason mortality was not classified into that asso- ciated and that not associated with jaundice. There were variations in mortality rates between the two surveys, but those who were transfused with un- screened blood generally had a higher mortality in both surveys than those given either no blood or HB Ag-negative blood. There was little difference within each survey between the mortality rates of those given HB Ag-negative blood and those of the non-transfused group. We have previously documented the prevalence of HB Ag carriers among our blood donor population (4, 5). By COE alone it was about 3% and by COE and complement fixation tests it was over 4 %. It has been shown that COE and complement fixation tests do not show positive results in every case found to be positive by more sensitive techniques such as haemagglutination inhibition or radioimmunoas- say (6). Therefore, the actual prevalence is likely to have been somewhat higher than 4%. Jaundice was reported in 25% of those who received blood known to contain HB Ag and in 9% of those who received unscreened blood. We may therefore assume that about 36% of the latter group received HB Ag-positive blood. Since the average number of units of blood transfused per person during the study period was 3.8 it appears likely that there were about 36 HB Ag-positive units among 380 units of blood. The prevalence thus calculated is approximately 9%. Since screening of blood was done by COE, about 5-6% of the blood samples would have been considered negative in spite of containing HB Ag, yet the incidence of jaundice among those who received blood found negative by COE alone was about 1 %, which was no more than that observed among the non-transfused control population. It is possible that this phenom- enon is related to the immune status of the general population, there being less than the anticipated inci- dence of hepatitis when the virus inoculum was rela- tively small, as would be the case if HB Ag was not detected by COE. However, these considerations are merely conjectural; the facts as documented point to the high risk of hepatitis following transfusion of unscreened blood, and to a marked drop in risk if COE-positive blood is eliminated. In the light of these results it is strongly recom- mended that screening for HB antigen be carried out routinely in all blood banks, including those in areas of the world where the prevalence of antigen carriers. is high. 619 620 T. JACOB JOHN ET AL. RESUME ANTIGENE DE L'HEPATITE B ET HEPATITE VIRALE DE TYPE B EN INDE Des enquetes faites a I'aide de questionnaires envoyes par la poste et de visites a domicile ont 6t6 menees pour determiner la frequence de 1'h6patite patente, se mani- festant par de l'ictere, dans des groupes de malades ayant survecu a la maladie initiale pour laquelle ils avaient e hospitalises. Parmi les malades ayant requ du sang oiu l'on n'avait pas cherche a deceler la presence de F'antigene de l'hepatite B (AgHB), 9,4% ont presente un ictere dans les 6 mois suivant la transfusion. En revanche, l'ictere n'a ete constate, pendant une meme periode, que chez 1,1 % de ceux qui avaient requ du sang trouve exempt d'antigene de l'hepatite B a 1'6preuve de 1'electrosynerese. Cette difference est tres significative. L'ictere a ete releve par 25,5% des receveurs de sang contenant I'antigene de l'hepatite B. Un contr6le effectue parmi une population tdmoin de malades hospitalises mais n'ayant pas subi de transfusion a revele que 2,2% d'entre eux avaient souffert d'ictere dans les 6 mois suivant leur hospitalisation. On peut donc dire que le risque d'hepatite patente due a l'antigene de l'hdpatite B dans le sang transfuse est dlev6 en Inde meridionale, et sem- blable a celui que connaissent les pays a climat tempere ou la prevalence de sujets porteurs d'antigene, et sans doute celle de sujets immuns, sont plus basses que sous les tropiques. A la lumiere de ces constatations, les auteurs recom mandent vivement que la recherche de l'antigene de l'hepatite B soit faite regulierement dans toutes les banques du sang, y compris celles des regions du monde ouf la prevalence de sujets porteurs d'antigene est elevee. REFERENCES 1. WORLD HEALTH ORGANIZATION. Viral hepatitis. Ge- neva, World Health Organization, 1973 (WHO tech- nical report series No. 512). 2. HOLLAND, P. V. ET AL. The infectivity of blood containing the Au antigen. In: Prier, J. E. & Fried- man, H., ed. Australia antigen. Baltimore, University Park Press, 1973, pp. 191-203. 3. CHERUBIN, C. E. Risk of developing post transfusion hepatitis in recipients of blood containing SH antigen at Harlem Hospital. Lancet, 1: 627-630 (1971). 4. HILL, P. G. ET AL. Australia antigen in blood donors in Vellore. Indian journal ofmedical research, 61: 378- 382 (1973). 5. SHANMUGHAM, R. V. ET AL. Comparative sensitivity of cross-over electrophoresis and complement fixation tests for the detection of Australia antigen. Indian journal of medical research, 61: 521-524 (1973). 6. RocHE, J. K. & STENGLE, J. M. Comparison of the sensitivities of the newer detection systems for hepati- tis B antigen. Transfusion, 13: 258-267 (1973).