A test survey of measles in a rural community in India.

130 NOTES the criterion for antigenic quality being its optimum (=working) dilution. To compare the antigens, the consumption of Toxoplasma mass (economy) and protein nitrogen (purity) were calculated. TwE anti- gens are more economical than FT antigens (see Fig. 4). The most frequent optimum dilution of FT antigens was 1: 8, while with TwE antigens it was 1: 32 (in two instances 1: 16). One gram of wet Toxoplasma sediment yielded approximately 80 ml of antigen by the FT method or approximately 640 ml of equally good TwE antigen. A total of 90 blood donor sera were examined- by the complement fixation test using both types of antigen and by the Sabin-Feldman test. A group of healthy individuals, i.e., who had mainly low anti- body titres, was considered more suitable for assay- ing the sensitivity of the antigen as when high anti- body titres are present there may be good corre- spondence in positivity even with an antigen of inferior quality. Complete correspondence in positi- vity and negativity between the two antigens was found. The TwE antigen gave titres one dilution step higher with two-thirds of the sera. All sera positive by CFT were also positive by SFT (Fig. 5 and 6). An interesting and simple method of preparing Toxoplasma antigen for CFT by carbonate extraction was described by Pettersen (1968). However, the optimum dilutions with this antigen are in the range of 1: 4-1: 8. We prepared two batches of Pettersen's antigen and obtained results analogous with the original results. REFERENCES Afzelius-Alm, L. & Hahn, E. (1953) Arztl. Wschr., 8, 1100 Beverley, J. K. A. & Beattie, C. P. (1952) J. clin. Path., 5, 350 Casey, H. (1965) Publ. Hlth Monogr., No. 74 Cooney, M. K., Kimball, A. C. & Bauer, H. (1958) J. Immunol., 81, 177 Cutchins, E. C. & Warren, J. (1956) Amer. J. trop. Med., 5, 197 Desmonts, G. (1954) Arch. franf!. Pediat., 11, 57 Fulton, F. & Dumbell, K. R. (1949) J. gen. Microbiol., 3, 97 Jira, J., Bozdech, V. & Heyberger, K. (1961) eas. Lek. ces., 100, 1291 Jira, J. & Bozdech, V. (1960) Zbl. Bakt., L Orig., 178, 262 MacDonald, A. (1949) Lancet, 256, 950 Pettersen, E. K. (1968) Acta path. microbiol. scand., 74, 35 Sabin, A. B. (1949) Pediatrics, 4, 443 Sabin, A. B. & Feldman, H. A. (1948) Science, 108, 660 Schuhova, V., ZdvadovA, H. & gtumpa, G. (1963) J. Hyg. Epidem. (Praha), 7, 65 Steen, E. & Kass, E. (1951) Acta path. microbiol, scand., 28, 36 Tos-Luty, S. (1967) Wiad. Parazyt., 13, 41 Uminski, J. (1956) Wiad. Parazyt., 2, 47 Warren, J. & Russ, S. B. (1948) Proc. Soc. exp. Biol. (N.Y.), 67, 85 Westphal, A. (1951) Z. Tropenmed. Parasit., 3, 72 Zastera, M., Huibner, J. & Pokornm, J. (1962) Cs. Epidem., 11, 122 Zavadova, H., Kutinova, L. & Vonka, V. (1967) Arch. ges. Virusforsch., 20, 421 A test survey of measles in a rural community in India USHA SHAH,1 K. L. BANERJI,2 A. N. D. NANAVATI,3 & NASEEMA A. MEHTA4 Measles is a disease of worldwide distribution, usually confined to the younger age groups. Mortality from measles is believed to be high, especially in communities with a poor nutritional status. Many workers further believe that measles is followed by an unduly high incidence of tuberculosis, severe 1 Professor of Preventive and Social Medicine, Medical College, Aurangabad, India. ' Medical Officer, Rural Health Unit, Paithan, India. Present address: Medical Officer, Rural Health Unit, Saoner Nagpur District, India. 'Assistant Director, Haffkine Institute, Bombay, India. 'Diamond Jubilee Research Fellow, Haffkine Institute, Bombay, India. 2786E malnutrition, and, sometimes, by physical and mental retardation that may be of graver significance to the community than is generally appreciated. No reliable data on measles, or on its long-term consequences, are available in India, since epidemiological infor- mation is scanty and unreliable. The alarming pic- ture drawn by Taneja, Ghai & Bhakoo (1962) is too full of approximations and extrapolations to be con- sidered as a factual representation. Long-term pro- spective studies are financially prohibitive while re- trospective studies are of doubtful significance unless the accuracy of the data can be ensured. A significant observation made during a recent MEASLES SURVEY IN A RURAL COMMUNITY IN INDIA 131 Table 1. Classification of surveyed families according to village and age composition Total Adults Children Children aged under Total Average Village no. of aged over aged 5 years no. of size ofbetween 6families 16 years and 15 years Male Female Total persons family Pimpalwadi 309 961 546 251 248 499 2 006 6.49 Dhorkhin 255 758 459 205 193 398 1 615 6.33 total 564 1 719 1 005 456 441 897 3621 6.42 serological survey conducted in Bombay was that, in the overwhelming majority of children tested, the parents' history with regard to measles was reliable; i.e., when parents stated that a child had suffered from measles, antibodies were present in the child's serum; when it was stated that the child had not suffered from measles, antibodies for measles were absent (Mehta, 1970).1 To determine whether this observation has a wider significance, a test survey was planned in a rural area to obtain data on the epidemiology of measles, together with serological studies on a representative sample of the surveyed population. The survey The Department of Preventive and Social Medi- cine of the Medical College at Aurangabad has a Rural Health Unit at Paithan, 64 km away, which serves that town and surrounding villages. The entire population of two villages, Pimpalwadi and Dhorkin, situated about 16 km apart in this area, were included in the study. Method. The survey was carried out over a period of 1 week. Interns, assisted by regular workers of the health unit, who had previously been briefed about this programme, visited each house in the two vil- lages. The information obtained was entered on the forms provided. A general form was used for each family, showing the number of persons, type ofhouse, educational status, occupation, etc. Deaths among children below 5 years of age during the past 5 years were noted, with the cause of death. Details of measles and other illnesses were recorded on an in- 1 This survey, and the serological studies described in the present paper, were carried out by one of us, N.A.M., as part of a study presented to the University of Bombay for a Doctorate in Philosophy. dividual form for every living child below 5 years of age. Blood samples were taken in Pimpalwadi village only; as far as possible, samples were obtained from every tenth child. Bleeding of a statistically random sample was not attempted on account of practical difficulties. Results Type of population. The population of the two villages comprised 564 families, containing 3 621 individuals. The age and sex distributions of the children are shown in Table 1. There were 897 children below 5 years of age for whom individual forms were completed. The socio-economic status of the population was uniformly poor, and malnutrition was common. Al- together, 477 out of 547 heads of families had no formal education and 409 were either agriculturists or unskilled labourers. There were 2 university grad- uates and 5 high-school graduates among the heads of families interviewed. Morbidity. Table 2 shows the incidence of the common infectious diseases in the child population. It is seen that 43.37% of children up to 5 years of age had suffered from measles. Table 2. Incidence of measles, chicken pox, smallpox. and whooping cough in the surveyed children Disease Total no. Reported Percentage _________ of children no. of cases affected Measles 897 389 43.37 Chicken pox 897 158 17.61 Smallpox 897 5 0.56 Whooping cough 897 41 4.57 NOTES Table 3. Age distribution of child population, and percentages of affected cases in each age and sex group I______________ Total Affected cases Percentages AMales Females Total Males Females Total Males Females Total 0-3 21 26 47 - - - - - - 4-6 29 32 61 1 3 4 3.45 9.38 6.56 7-9 21 12 33 1 - 1 4.76 - 3.03 10-12 37 39 76 4 6 10 10.81 15.38 13.16 13-18 51 38 89 22 9 31 43.14 23.68 34.83 19-24 39 44 83 21 12 33 53.85 27.27 39.76 25-30 38 37 75 22 17 39 57.89 45.95 52.00 31-36 73 70 143 42, 41 83 57.53 58.57 58.04 37-42 18 15 33 12 7 19 66.67 46.67 57.58 43-48 59 56 115 38 33 71 64.41 58.93 61.74 49-60 66 71 137 46 48 94 69.70 67.61 68.61 >60 4 1 5 3 1 4 75.00 100.00 80.00 total 456 [ 441 [ 897 [ 212 177 | 389 { 46.49 40.14 43.37 Table 4. Incidence of measles according to size of family Percentage of Average no. of No. of family No. of cghildren Children affected/total children aged members families age under affected children under 5 years (2) (3)~ ~(4)rs(5) -(4) x 100 per family(1)_________ (2) (3) (4) (3) (6) = (3)/(2) 6 3 38 38 10 26.32 1.00 4 89 125 54 43.20 1.40 5 110 160 70 43.75 1.45 6 95 148 64 43.24 1.56 7 85 138 61 44.20 1.62 8 50 84 37 44.05 1.68 9 40 76 41 53.95 1.90 10 19 34 13 38.24 1.79 > 10 38 94 39 41.49 2.47 total [ 564 897 389 } 43.37 [ 1.59 132 MEASLES SURVEY IN A RURAL COMMUNITY IN INDIA Table 5. Age at onset of measles and family size Age at onset 1 No. of family membersAge at onset.Toa(months) 1-2 2-3 3-4 4-5 [5-6 6-7 7-8 8-9 [9-10 > 10 Tota < 3 - - - 1 - 2 2 - - - 5 3-6 - 2 7 6 3 - 1 2 2 2 25 6-9 - I - 4 1 2 3 3 - 1 15 9-12 - 1 13 22 15 13 13 9 4 10 100 12-18 - 2 7 4 6 9 2 4 - 3 37 18-24 - 2 16 13 14 15 5 12 5 10 92 24-30 - 1 2 3 4 3 - 3 - 2 18 30-36 - - 7 8 15 12 4 5 - 7 58 36-42 - - 1 3 6 1 3 - 1 1 16 42-48 - 1 - 4 - 4 4 3 1 3 20 48-54 - - 1 2 - - - - - - 3 total affected~ - 10 54 170 64161 ]37141 113 1391389 Age and sex distribution. The age and sex distribu- tions are shown in Table 3 together with the number and percentage affected by measles in each group. There is a slightly higher incidence among males. The overall incidence of measles was unaffected by size of family, except in single-child families where the incidence was significantly lower (Table 4). Table 5 shows the number of children by age of onset, the highest number being between 1 year and 3 years of age. The age of onset of measles was not affected by the number of persons in the family. Seasonal incidence. The higher total number of cases seen for every alternate 6 month period (Ta- ble 5) reflects the seasonal incidence of the disease, which reaches a peak during the winter months. Duration ofillness and complications (Table 6). The duration of illness shows two distinct phases. The majority, 312 cases, recovered within 15 days. In the remaining 77 cases the duration was between 20 and 99 days. When the duration is compared with the incidence of complications, it is seen that there were 122 cases in which complications ofvarious types were reported. Among the 312 patients who recovered within 15 days, there were 59 cases with complica- tions. Among the 77 cases with a longer duration, there were 63 cases with complications. A retrospec- tive history with regard to the duration of illness and to the complications (assessed by symptoms) is notoriously unreliable, and no effort has been made to evaluate this information statistically, or to assess the relative severity of the complications recorded. It appears reasonable to assume, however, that cases in which the illness is prolonged beyond 15 days are usually cases with more severe complications. As regards the types of complication encountered, little can be stated. About 22% of the cases showed gastroenteritis, and 14% respiratory complications, alone or combined with other symptoms. In view of the overall frequency of episodes of gastroenteritis and respiratory illness in this age group, the signifi- cance of these findings is impossible to assess. Otitis media (ear pain or discharge or both) was recorded in 7 cases. No correlation was found between age of onset of measles and duration, or age of onset and complica- tions, i.e., the severity of the disease was unrelated to age. Treatment. The beliefs and taboos prevailing in the surveyed population, as also the treatments given, were varied and diverse, and no correlation could be obtained between all these factors. The treatments given, reflecting the beliefs of the population, could be grouped as dietary (milk, glucose, jaggery, etc.), external applications (bathing in neem leaf infusion was common), use of charms or mantras, medical treatment, and various combinations of these treat- 10 133 134 NOTES Table 6. Complications and duration of illness Weeks Total for Weeks Total for Complication short a_long Total al1 t 1-2 duration 2-3 3-4 4-8 9 durationa fever and cough 2 2 4 - - 4 - 4 8(1.0) (1.0) (2.1) respiratory complications 2 9 11 3 - 2 - 5 16(2.8) (1.3) (4.1) gastroenteritis 1 29 30 6 1 10 4 21 51 (7.7) (5.4) (13.1) fever, respiratory, and gastrointestinal 1 3 4 1 - 10 1 12 16 (1.0) (3.1) (4.1) conjunctivitis, respiratory, and gastrointestinal - 3 3 2 3 12 4 21 24 (0.8) (5.4) (6.2) otitis media - 7 7 - - - - - 7(1.8) (1.8) total cases with complications 6 53 59 12 4 38 9 63 122(15.2) (16.2) (31.4) total cases 21 291 312 15 4 45 13 77 389 (80.2) (19.8) (100.0) a Values in parentheses are percentages. ments. About 43% of the patients took medical Altogether, 10 of these deaths were due to measles, treatment either at the health unit or privately, while giving a case-mortality rate of 2.2% (total cases = about 40% took other treatments; there was some 389 + 43.37% of 139 = 449). overlapping (Table 7). A glance at Table 7 suggests that complications were far more frequent in patients Serological findings. Bleeding was confined, for taking medical treatment than in others. However, convenience, to the village of Pimpalwadi, and 61 the stage of illness at which medical treatment was sera were obtained, representing 6.6% of the total sought is not recorded, and it seems that medical child population of the study area or 12% of the advicewassought because the ilness failedtosubsde child population (of 499) in Pimpalwadi village.advice was sought becauethillnsHaemagglutination-inhibiting(HI),neutralizing(SN),and complications had developed, and complement-fixing (CF) antibodies were esti- Mortality. A total of 139 children below the age of mated in these sera by standard procedures and the 5 years had died in the last 5 years in this community. tests were interpreted according to previous experi- Table 7. Type of treatment given and complications recorded 1- ~~~~~~~~~Chrs+ Total no. Medical Diet External Diet + Chatrm+ without treatment Noapiainapplication application medical and other treatment| [ appli ation {treatment |meas'ures| no. of cases with complications 15 7 5 8 35 72 12 total no. of cases receiving the treatment 65 21 29 38 153 165 60 percentage showing complications - - - - 22.9 43.6 20 MEASLES SURVEY IN A RURAL COMMUNITY IN INDIA Table 8. Results of serological tests on children of Pimpalwadi village Group Group Grouup Total A B c total no. of children 221 250 28 499 sample tested 28 23 10 61 no. positive for antibodies a 27 1 - 28 no. negative for antibodies 1 22 10 33 a Positive indicates HI titre above 1 :20 and neutralizing antibody titre above 1 : 16. ence (Enders, 1964; Mehta, 1970; Rosanoff, 1961; Rosen, 1961; Toyoshima et al., 1965). It will be seen from Table 8 that the blood samples represented three groups: (1) children who had suffered from measles; (2) children who had not been exposed to measles; and (3) children who had been exposed to measles but had not contracted the disease. (" Ex- posed" means having household contact of several hours or days with a case.) The sera were numbered, with no indication of the group represented, and were tested in Bombay (by N.A.M.). All except two sera, one each from groups (1) and (2), showed corroboration of the history obtained. On general grounds, this indicates a high degree of reliability, and these findings are statistically significant at 95% confidence limits. Detailed results are presented in Table 9. Contagiousness of measles, and subclinical infec- tions. Among 897 children surveyed, 389 had suffered from measles, and 42 (20 males and 22 females) had been in contact with a case of measles without contracting the disease, i.e., 431 children had contact with a case of measles, and 389 acquired the infection by such exposure (Table 10). Thus, 9 out of 10 non- immune children exposed to a case of measles devel- oped clinical measles. Children below 6 months of age, considered separately, showed no difference of susceptibility. Maternal antibodies, therefore, were either absent from this small number of infants or had declined too rapidly to have any protective effect. The absence of antibodies from the sera of 10 children who were exposed but did not suffer from measles (Tables 8 and 9) suggests that subclinical infection with development of immunity is not a likely occurrence in a nonimmune population. Discussion A major problem in elucidating epidemiological processes is to ensure that events can be correctly traced and recorded. This depends on the extent to which the interviewing and interviewed persons have a common language in terms of concepts and ideas, or, alternatively, on the availability of a specific laboratory test that is simple enough to be applied on a large scale. The difficulty of ensuring these conditions has discouraged many workers from undertaking such studies. Recent experience in Bombay indicated that a retrospective history, with regard to measles, could be reliable. The present survey was planned to test the reliability in a less sophisticated rural population with much lower educational levels than prevail in Bombay; the entire population of two villages was included. The reliability of the information obtained at interview was checked by a sample study for serum antibodies and a high degree of reliability was found. Until recent times, measles was considered inevit- able; only palliative and supportive treatment to prevent serious complications was possible. Today, an effective preventive measure in the form of vac- cine is available. However, the cost of the vaccine is relatively high and its administration is not easy. If measles causes considerable loss of life, or if its sequelae cause considerable disability, then the cost of the vaccine is not important, and mass vaccination must be given due importance in our plans. If, on the other hand, the morbidity and mortality are low, it has to be seriously considered whether the large outlay required for mass vaccination is justifiable. Hence, an understanding of the epidemiology of measles, and of its effects, immediate and remote, on the health of the community is essential. The requirements for the present survey were modest. The interviews were conducted by 9 in- terns working for 1 week. Subsequently, a day was fixed for bleeding and 4 physicians attended to take blood samples. It must be added that there are two essential prerequisites for carrying out such a programme; (1) the participation of persons who enjoy the confidence of the population, and (2) care- ful briefing of the interviewers before the work is started. The overall incidence of measles appears to be similar to that reported from other parts of the world, and the mortality is close to the lower limit reported for this disease. No statistical significance is 135 NOTES Table 9. Antibody titres obtained in the sera tested Subject Age Antibody titre No. (years) HI CF SN Group A, children who had suffered from measles 2 160 8 4 40 <4 4 320 8 2% 160 4 2 160 4 5 160 <4 3 320 4 2% 640 1 6 4 160 4 1 640 16 2 160 4 4 640 16 4 320 4 5 160 <4 2% 320 1 6 4 80 <4 4 320 4 3% 80 < 4 4 320 < 4 4 320 4 1% 640 32 3 160 4 3 80 <4 a 2% 10 < 4 2 160 4 4 80 < 4 3 320 4 3 160 <4 Group B, children not exposed to measles 2% <10 <4 11 months < 10 < 4 256 64 512 256 256 256 512 1 024 256 1 024 256 1 024 512 256 512 128 512 128 512 512 1 024 256 128 < 4 256 128 512 256 < 4 < 4 Subject Age No. (years) HI 12 4 <10 16 11 months < 10 17 1% < 10 1 9 b 1'1 640 23 10 months < 10 24 3 <10 25 1 <10 26 2 <10 29 1 % <10 33 2% < 10 34 1 <10 36 1 <10 38 1 <10 44 1% <10 46 10 months < 10 50 2 <10 51 1 1 months < 10 52 5 <10 43 4 <10 65 2 <10 71 3 <10 Group C, children exposed to measles but 2 5 < 10 18 1 < 10 37 2 <10 40 4 <10 42 2'% <10 45 8 months < 10 55 2% < 1 0 70 1 % <10 72 3 <10 74 3 <10 Antibody titre CF SN <4 <4 <4 <4 <4 <4 16 1 024 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 showing no clinical disease <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 <4 a Case with history of measles, did not show any antibody titre. b Case without history of measles, exhibited high antibody titre. 136 11 151 21 28 30 35 39 49 53 54 56 58 59 60 61 62 63 64 66 67 68 69 73 6 10 MEASLES SURVEY IN A RURAL COMMUNITY IN INDIA 137 Table 10. Susceptibility to measles at different ages Category Age 1 Total| < 6 months > 6 months no. exposed 32 399 431 no. affected 30 359 389 percentage affected 93.75 89.97 90.26 attached to the higher incidence among male children (Table 3). It is interesting to note that the number of persons in the family does not affect the incidence or age at onset of the disease, and that the severity of the disease or its complications are unrelated to age. Complications are of particular interest in measles. It would be reasonable to consider a complication as serious when it prolongs the course of the disease. Only 77 cases of such prolonged and complicated illness are recorded. If we add the 10 fatal cases, 22% of all cases showed a serious or prolonged illness with complications, while 78% recovered within 15 days. The assessment of late sequelae presents a real problem in such a study. As already stated, the people are very poor and malnutrition is widespread among the children. To assess relative degrees of malnutrition calls for specific objective criteria to which numerical values can be assigned. Since no such criteria are available today, no relative assess- ment of malnutrition was possible. Tuberculosis is not uncommon in this area, though cases in the age group surveyed are few. There is no mention of obvious pulmonary infection in the indi- vidual records of the children. However, in the absence of a special enquiry on this point, no opinion on the relative incidence of tuberculosis can be given. Regarding mental or physical retardation, slow rate of development, weight loss, and other such manifestations, only subjective judgements are possible in retrospect. Hence, specific conclusions on late sequelae of measles cannot be obtained from this survey, though it is unlikely that any obvious defect or disability known to be associated with measles has been missed. A more careful study of these aspects, and the establishment of objective criteria for assess- ment, should be attempted for any future survey. Variations in the pattern and severity of the disease due to socio-economic differences could not be ascer- tained in this population, which was uniformly poor. In spite of the poor general health of the population, the disease does not appear to be as severe or to carry as great a risk of mortality as in other parts of the world (Morley, 1962). The question whether this difference is due to differences in the strain of virus or to factors in the environment requires investigation. TIhe possibility of subclinical infections of measles followed by immunity has been considered by various workers to explain some of the epidemiological phe- nomena that are encountered in successive epidemics (Bech, 1960; Krugman et al., 1965; Medical Re- search Council, 1938; Stocks & Karn, 1928). In one study (Snyder et al., 1962), antibodies were found in the sera of 45% of 372 young people examined, though a history of measles infection was obtained in only 27 %. Most of these studies have been concerned with children of school age, in whom other factors may be involved. In the pre-school child population surveyed here, it appears that subclinical infections do not occur with any degree of frequency, and, as a natural corollary, the history given by the parents can be relied upon. Similar observations were made in 200 sera of children tested in the Bombay survey already referred to (Mehta, 1970). These findings need to be confirmed by similar surveys in other parts of the country. ACKNOWLEDGEMENTS Without the help of the Health Unit staff, and in parti- cular of Dr V. N. Ingle, Professor of Paediatrics, Auran- gabad Medical College, whose regular paediatric clinics at Paithan have won the confidence of the people and ensured their co-operation, this study and the collection of blood samples would not have been possible. The authors express their gratitude to those workers, to the interns for their enthusiastic participation, and to the Dean of the Aurangabad Medical College and the Direc- tor of the Haffkine Institute for permission to publish this study. They also thank Glaxo Laboratories, India, Ltd, who kindly arranged for a computer analysis of the data at their expense, and the Gokhale Institute of Politics and Economics, Poona, and Shri D. B. Sardesai in particular, for the statistical analysis and the prepar- ation of tables. REFERENCES Bech, V.- (1960) Acta path. microbiol. scand., 50, 322-330 Enders, J. F. (1964) Measles virus. In: Diagnostic proce- dures for viral and rickettsial diseases, New York, American Public Health Association, pp. 540-550 138 NOTES Krugman, S., Giles, J. P., Friedman, H. & Stone, S. (1965) J. Pediat., 66, 471488 Medical Research Council (1938) Epidemics in schools, London.His Majesty's Stationery Office (Special Report Series No. 227), pp. 144-150 Mehta, N. A. (1970) Seroepidemiology of measles, Ph. D. Thesis, University of Bombay, India Morley, D. C. (1962) Amer. J. Dis. Child., 103, 230-233 Rosanoff, E. I. (1961) Proc. Soc. exp. Biol. (N.Y.), 106, 563-564 Rosen, L. (1961) Virology, 13,139-141 Snyder, M. J., McCrumb, F. R., Jr. Bigbee, T., Schlueder- berg, A. E. & Togo, Y. (1962) Amer. J. Dis. Child., 103, 250-251 Stocks, P. & Kam, M. N. (1928) Ann. Eugen. (Lond.) 3, 361-398 Taneja, P. N., Ghai, 0. P. & Bhakoo, 0. N. (1962) Amer. J. Dis. Child., 103, 226-229 Toyoshima, K., K-itawaki, T., Otsu, K., Mutai, M., Omura, S. & Kunita, N. (1965) Biken's J., 8, 87-94