The unchanging epidemiology and toll of measles in Burma.

Bulletin of the World Health Organization, 63 (3): 551-558 (1985) (cj World Health Organization 1985 The unchanging epidemiology and toll of measles in Burma* JAMES CHIN' & U MAUNG THAUNG2 The very large contribution of measles to childhood morbidity and mortality in many developing countries is generally not well appreciated or documented. Because of this, an evaluation ofthepublic health importance ofmeasles in Burma was carried out in mid-1983 toformulate recommendations regarding the need andfeasibility ofincorporating measles vaccine into Burma's expanded programme on immunization. The findings, based on a review ofthe available medical, serological and epidemiological data and afollow-up study of childhood deaths, all showed that the epidemiology of measles has not changed during the past two decades and that measles is a major, if not theforemost, killer of infants and young children in Burma. This situation is probably very similar to that in many other developing countries where measles vaccine has not been routinely used. Similar assess- ments of the epidemiology and toll of measles should be carried out in countries where measles vaccine is not a current component of the national immunization programme. Any population that is not protected by measles vaccination will invariably be exposed to the risk of measles among all the infants and children. Since measles vaccine was never included in the national childhood immunization programme in Burma, it can be assumed that the natural epidemiology of this extremely infectious virus disease has remained essentially unaltered. Although health care workers in many areas of the world may realize that measles is a universal infection of children, the majority among the general public have an inadequate perception of the immense numbers of serious illnesses and deaths that can be attributed to this infection. Thus, one of the prime obstacles to the initiation of a measles control programme in Burma and other countries in south-east Asia is the lack of an accurate assessment of the full health implications of this vaccine- preventable disease. This article describes the epidemiology of measles in Burma and its main effects on morbidity and mortality among infants and children. The epidemiology and toll of measles in other countries in south-east Asia where measles vaccine has not been used are probably very similar to our findings in Burma. * This article is based on a report by J. Chin to the World Health Organization (Regional Office for South-East Asia) after a Con- sultation to Burma in May/June 1983. Chief, Infectious Disease Section, California Department of Health Services, 2151 Berkeley Way, Berkeley, CA 94704, USA. Requests for reprints should be sent to this author. 2 Deputy Director (Epidemiology), Disease Control Division, Ministry of Health, Rangoon, Burma. METHODS Data on the occurrence and clinical severity of measles in Burma were obtained from both published and unpublished studies carried out during the past couple of decades and from specific studies carried out during aWHO consultation visit in May and June 1983. During the course of this consultancy, two additional measles studies were carried out: a serological survey of children and a follow-up study of childhood deaths in selected areas of Rangoon. The methods of these two studies are described below. Serological survey As part of an investigation on plague in Burma, serum samples collected by intravenous bleeding were obtained in early 1983 from 3559 persons of all ages residing in plague-endemic and plague-free areas of the Rangoon Health Division. Filter paper discs, capable of absorbing 0.4 ml of sera, were soaked in the sera of 383 children aged from 1 month to 14 years. The discs were air dried and transported with- out refrigeration to Berkeley, California, USA, where measles haemagglutinin inhibition (HI) antibody tests were performed in November 1983 by the Viral and Rickettsial Diseases Laboratory of the California Department of Health Services. The discs were eluted by a method previously described (1) and the HI test was performed by a standard technique (2). 4553 -551 J. CHIN & M. THAUNG Follow-up study of childhood deaths This study utilized death certificates that were available in the Rangoon Health Division. Phase I. Concentrating on three selected townships in Rangoon, the staff of the Rangoon Health Division were requested to provide a list of all deaths registered among children aged from 6 months to 10 years during the period I January 1982 to 31 March 1983 (i.e., a 15-month period which spanned two measles seasons). The information requested for each death included the name of the child, sex, age at death, father's name and address, and the stated cause of death. This phase of the study was started on 23 May and the list was completed by 25 May 1983. Phase 11. Starting on 30 May, the staff of the Rangoon Health Division began to visit and interview the parent(s) of the deceased children in order to complete a standard form for each death. A set of detailed instructions was given to each health visitor so that the forms could be filled out in a standardized fashion. This phase of the study was completed by 9 June 1983. A death was classified as related to measles if the following signs and symptoms were present within 3 months of death -fever, rash, and at least one of the following three symptoms: cough, runny nose (coryza), or red watery eyes (conjunctivitis). The health visitors who interviewed the parent(s) were not informed of these clinical criteria for measles. They were instructed to fill out uniformly and completely the follow-up form. The classification of deaths was carried out jointly by a medical officer in the Burmese Expanded Programme on Immunization (EPI) and by one of the authors (JC). FINDINGS Reported incidence of measles in Burma Measles cases have been officially notifiable to the Burma Ministry of Health since 1971. Fig. 1 presents the number of reported measles cases in Burma from 1971 through 1982. From 1971 to 1978, a fairly low level of reported cases is seen -ranging from 1681 cases in 1973 to 7674 cases in 1976. In 1978, the Burmese EPI was started. In the following years, the reported incidence level increased to over 20 000 cases per year. The observed pattern of reported measles during this time span suggests that the incidence of measles has been steadily increasing in Burma. However, from other data on the occurrence of measles in Burma presented below, it will be evident that almost all children in Burma have been exposed to measles 28000 r 24000 F 20000 I- is(5 16000 F 12000 I- 8000 4000 71 72 73 74 75 76 77 78 79 80 81 82 Year Fig. 1. Number of reported measles cases in Burma, 197 1-82. infection by 10 years of age and that the true incidence has not changed over the past couple of decades. Age distribution of measles cases in Burma Several sources of data clearly show that practically all children in Burma are infected with the measles virus by 10 years of age. Serological surveys. Data from three measles sero- logical surveys-an unpublished study in Rangoon in 1969, a published study in Rangoon in 1971 (3), and a 100 r 0 .00 0. 0 e .0 E 0 vl m 0 2 0. w :1 90 80 1 70 1 60 1 50 1 40 1 301 201 10 L/ -*- 1969 (863 children) / . Hi 21:2 / ! 1971 (186 children) 1// Hl1:10 - // , 1983 (383 children) HI > 1:2 "<1 1 2 3 4 5 6 7 8 9 10 Age in Years Fig. 2. Results of three measles serological surveys in children in 1969, 1971 and 1983. IuI f } w 552 MEASLES IN BURMA 50 r C._ Ca: - U, 0 _ m E0gm 401- 30 - 20 - 10 - 0 <1 1-2 2-3 3-4 4-5 Age in Years I <1 1-2 2-3 3-4 4-5 Age in Years Fig. 3. Measles attack rates by age group (on left) and cumulative age-specific attack rates (on right). recent (1983) survey of selected areas in the Rangoon Health Division are presented in Fig. 2. The pattern of acquisition of measles HI antibodies by age is virtually identical in all three studies. Close to half the children had measles antibody by the age of 4-5 years, and nearly 90% of children were infected before the age of 10 years. A multipurpose cohort study of children in Rangoon, 1970-75. In 1970, the Central Epi- demiology Unit of the Disease Control Division of the Ministry of Health, Burma, initiated a study of various diseases in a cohort of 800 infants born in 1970 in South Okkalapa, Rangoon. Children in this cohort were followed up every 3-4 months, up to their fifth birthday, in order to collect information about selected diseases, including measles. Fig. 3 (in two parts) presents data obtained from this cohort on the clinical attack rate of measles by age. One part shows that the clinical attack rate for measles is highest (27.407o) during the second year of life; the other shows the cumulative clinical attack rate up to the age of 5 years. This cumulative attack rate is almost identical with the results of the serological surveys presented above. In the cohort study, 45%o of the children had clinical measles and in the serological surveys, 40-50% of children had demonstrable measles HI antibodies by the age of 5 years. These data corroborate the fact that almost all infections with measles virus are symptomatic. Community outbreaks ofmeasles. Additional data to support the very high measles attack rate in young children are derived from two community outbreaks of measles in Burma. The first outbreak was in the Sagaing township in 1966 (4) and the second, which occurred in May 1983 in a village in the same general area, was investigated and reported by the Sagaing Health Division. The measles attack rates by age from these two outbreaks are presented in Table 1. The data from these outbreaks also support the findings of the measles serological surveys and the prospective cohort study. In these outbreaks the highest measles attack rates were in children under the age of 5 years and, as would be expected from the results of the serological surveys, very few measles cases were seen in children aged 10 years or older. Clinical severity of measles in Burma The vast majority of measles cases, even the seriously ill, may never receive medical attention, much less be admitted to a hospital. However, a review of available clinical data from the paediatric wards of three hospitals in Rangoon gave some indication of the severity of illness. Table 2 presents for three paediatric wards in Rangoon the percentage of total admissions attributed to measles or its com- plications during the first 3 months of 1983, a period which coincides with the peak of the measles season in Burma. This percentage ranged from 450o in the Infectious Disease Hospital to 16%7o in the Children's Hospital, but for all the hospitals, measles accounted for nearly 20% of all paediatric admissions during this period. Thus, annually, measles can place a very large burden on the hospital services in Burma. 30 25 0 0 :0(U 20 15 10 5 0 553 J. CHIN & M. THAUNG Table 1. Measles incidence by age group, in two community outbreaks in Burma in 1966 and 1983 1966 outbreaka 1983 outbreak" Age No. of No. of group measles Attack measles Attack (years) No. in population cases rate' No. in population cases rate' < 1 633 82 13.0 47 11 23.4 1-4 2 082 705 33.9 196 80 40.8 5-9 2 794 504 18.0 257 75 29.1 > 10 14 353 49 0.3 1429 16 1.1 Total 17 982 1340 7.5 1929 182 9.4 a Source: ref. 4. b Source: Outbreak report on measles, Sagaing Health Division, 1983. ' These rates are per 100 persons for each age group. Table 2. Measles and total paediatric admissions in the Rangoon hospitals, January-March 1983° Total Measles admissions admissions January 1980 237 (12.0)" February 1748 375 (21.5) March 2200 540 (24.5) Total 5928 1152 (19.4) a Data provided by the hospital Medical Superintendents. b Figures in parentheses are percentages. Clinical data on measles cases from selected hospitals in Burma (Rangoon and Mandalay) were variable, possibly because of such factors as differences in the physicians' diagnosis of measles complications, in the hospital admission policies by age, etc. However, among the measles cases seen in three large hospitals in Burma over a 15-month period (January 1982 to April 1983), respiratory conditions constituted the majority of the severe complications associated with measles and were noted in 50-90% of all measles admissions. A wide variation was also noted for gastrointestinal complications in measles patients, ranging from less than 5% to over 25% in different hospitals. No data were available on the number of outpatient visits for measles, but it is assumed that measles and its complications must place a strain on these services during the measles season. It is also likely, with such large numbers of measles cases in the hospitals, that susceptible children admitted to these hospitals for other problems during the measles season are at very high risk of acquiring measles during their hospital stay. Measles deaths in Burma Data from several sources point to the major con- tribution made by measles to infant and childhood mortality in Burma. The case-fatality rate (CFR) of hospitalized measles cases. A review of 2569 measles cases admitted to hospitals in Rangoon and Mandalay from January 1982 through April 1983 showed an overall CFR of 8.2%o. This rate in the different hospitals ranged from 4.507o to 13.10o, but the variation was probably related to admission policies: the highest CFR was in the hospital that usually treats young infants and children under the age of 2 years and the hospital with the lowest CFR tended to receive mostly older children. The CFR observed in community outbreaks of measles. Table 3 shows the measles CFR for the two community outbreaks of measles in 1966 and 1983, which were mentioned above. The rate was 2.3%7o in the larger community outbreak in 1966 and 8.4%o in the smaller village outbreak in 1983. The highest CFRs were seen in the youngest children. Of special note is the 1966 study finding that, of all deaths in children under 10 years of age recorded during the 6-month study period, more than half (30/51 or 58.8qo) were very probably attributable to measles. Follow-up.study ofchildhood deaths. Table 4 gives by age the causes of death for the 280 children identified by death certificates. About a third of the deaths were attributed to fever and convulsions, and about a quarter of the deaths were listed as due to respiratory problems such as bronchopneumonia, 554 MEASLES IN BURMA Table 3. Measles case fatality rates (CFR), by age group, in two community outbreaks in Burma in 1966 and 1983 Age 1966 outbreak" 1983 outbreak' group (years) Cases Deaths CFR (%) Cases Deaths CFR (%) < 1 82 5 6.1 11 0 0.0 1-4 705 1 9 2.7 80 11 13.8 5-9 504 6 1.2 75 3 4.0 Total 1291 30 2.3 166 14 8.4 Source: ref. 4. Source: Measles outbreak report, Sagaing Health Division. During the 6-month period of this study, a total of 51 deaths was recorded in children under age 10 years. Of these, 30 (58.8%) were attributed to measles. Table 4. Childhood deaths, by diagnosis and age on death certificates, in three selected townships in Rangoon, from January 1982 to March 1983 Age group Diagnosis on 6-12 1-4 5-9 death certificates months years years Total Fever, fits, convulsions 18 62 15 95 (34.0)a Respiratory problems 19 44 8 71 (25.4) Measles 2 34 10 46 (16.4) Malnutrition 4 24 15 43 (15.4) Other 3 12 10 25 (8.2) 46 176 58 280Total (16.4) (62.9) (20.7) (100.0) a Figures in parentheses are percentages. chest infections, bronchitis, etc. Measles was the third most commonly listed cause of death (16.40o), with malnutrition ranking fourth (15.4%0). Among the 280 deaths, it was surprising to note that only three (1.%lo) had gastroenteritis or diarrhoea noted as the primary cause of death. As expected, almost two- thirds (62.90o) of the deaths occurred in the 1-4-year age group. Parents of 249 (88.907o) of the 280 deceased children were interviewed and a survey form was filled out. The following analysis is based on the 249 deaths for which this form was completed. As shown in Table 5, of 43 deaths which were listed as due to measles on the death certificate, 36 (83.7%o) met the established criteria for measles. The other 7 children had fever, but did not have a history of having had a rash within 3 months of death. Of 206 deaths which did not have measles mentioned on the death certificate, 51 (24.87o) met the clinical criteria for measles. Of these 51 cases, 46 (90.2%) probably had measles immediately preceding death and 5 (9.8%) within 3 months of death; these 5 deaths were listed as bronchopneumonia, meningitis, or mal- nutrition, which may be considered to represent post- measles complications. Thus, of the total of 249 deaths reviewed in the survey form, a total of 87 (34.9%o) could very probably be attributed to measles or post-measles complications. Table 5 also shows that, of the 51 measles deaths diagnosed retro- spectively, the majority had been attributed to fever, fits, convulsions, and respiratory conditions like bronchopneumonia, chest infection or congestion. Table 6 shows the type of medical care received by the 87 fatal cases considered to be due to measles or to post-measles complications. Less than 20% of these 555 J. CHIN & M. [HAULJN(i Table 5. Results of follow-up on 249" childhood deaths identified by death certificates in Rangoon, June 1983 Cause of death on No. confirmed as measles death certificate No. by follow-up Measles 43 36 (83.71" All others 206 51'124.8) Total 249 87 134.9) " The parents of 31 decedents 111.1%) of the 280 deaths reviewed could not be located. Included in this group of 31 were 3 children whose death certificate gave measles as the cause of death. All 31 deaths were excluded from the follow-up analysis. Figures in parentheses are percentages. The causes of death given on the death certificates of these 51 cases, diagnosed retrospectively as probably measles, were as follows: Fever, fits, convulsions (in 25); Respiratory problems (in 18); Malnutrition, meningitis, etc. (in 8). Table 6. Measles deaths, as identified by the death follow-up study, and the type of medical care received, Rangoon, June 1983 No. Type of medical care of deaths None (home and/or traditional treatment) 29 133.3) Outpatient care or visits to a doctor 41 (47.2) Hospital" 17 (19.5) Total 87 (100.0) Figures in parentheses are percentages. Most of the hospitalized cases also had outpatient care previously. children had been about a third of physician. hospitalized before death, and them were not seen by any DISCUSSION Among the major causes of illness and death in infants and children throughout the world, the measles virus is probably the single most important infectious agent. It has been estimated that the annual worldwide toll of measles is close to a million deaths (5), with at least an equal number of very severe clinical illnesses. Compared with other major killers of children (e.g., diarrhoea, malaria and mal- nutrition), measles stands out because it alone is relatively easy to prevent. The reported numbers of measles cases in most countries are usually a gross underestimate of the actual incidence. This is especially true in non- industrial countries where the reported incidence may be only 1/100th of the true incidence. However, the annual incidence of measles can be estimated fairly accurately in areas where little or no measles vaccine is used. The number of measles cases that could be expected, on average, to occur each year in these areas would be the annual number of births minus those infants who die within their first few months. This is because the infectivity of the measles virus is so high that almost all persons in a given population will be infected at some time during their childhood. In Burma several serological studies, a prospective cohort study, and investigations of measles outbreaks carried out in different populations at different times since 1966 have shown that about 50%7o of all children had clinical measles by the age of 5 years and that virtually all children had measles before the age of 10 years. Thus, measles can be considered to be a disease affecting all children in Burma and it can be estimated that, on average, about one million cases of clinical measles occur annually. This number is the equivalent of the estimated current annual birth cohort in Burma. Thus, prior to the start of the expanded immunization programme in 1978, measles had been underreported in Burma by a factor of almost 200 and after 1978 by a factor of almost 50. Since practically all measles cases are symptomatic, it can be expected from the known, documented clinical spectrum of measles that each year in Burma, hundreds of thousands of infants and children will experience relatively severe clinical illness consisting of rash, high fever, and some respiratory symptoms, which will last for weeks or months. As documented in Rangoon in early 1983, up to 2007o of all paediatric admissions to hospitals can be attributed to measles during the measles season. In addition to the large number of severe measles cases that occur annually, the study of childhood deaths in Rangoon indicates that measles is a major contributor to childhood mortality in Burma. Over a third of the total deaths of children aged from 6 months to 10 years, during the 15-month study period, could be directly attributed to measles or to post-measles complications. This study spanned two measles seasons so that the proportion of deaths found to be associated with measles (34.9%) was probably higher than if the study period was limited to one season. However, even if the percentage of measles deaths was halved (i.e., 15-20%), this would 556 MEASLES IN BURMA still be an unacceptably high mortality rate. Other clinical and survey data available in Burma also support the conclusion that measles is a signifi- cant contributor to childhood mortality. Study of the community outbreak of measles in the Sagaing town- ship in 1966 indicated that more than half of all deaths in children aged under 10 years during the 6-month period of this study was probably due to measles. The clinical data from paediatric hospitals in Mandalay and Rangoon indicate that the CFR among measles admissions is over 87o, but, as shown in the death certificate follow-up study, less than 20% of measles deaths may be hospitalized. From the available data in Burma regarding measles incidence and measles deaths in young children, some estimates of the total annual number of deaths due to measles in Burma can be made. Taking into account all the available survey and study data, the overall CFR for measles in Burma can be conservatively estimated at about 1-2%. Since approximately a million cases of measles occur annually, such a rate would yield from 10 to 20 thousand measles deaths per year in Burma. All the available data from studies of measles in Burma since 1966 indicate that the general toll of this disease and its epidemiology have not changed. Measles continues to infect virtually all Burmese children by the age of 10 years. It also continues to be responsible for a large proportion of the total deaths in young children, and each year tens of thousands of children may have very serious complications due to measles, especially bronchopneumonia. Besides morbidity and mortality considerations, the measles burden on both outpatient and hospital resources justifies, on a cost-effective basis, the use of meAsles vaccine in Burma on as large a scale as possible. The major problem for the prevention and control of measles in Burma, and many other developing countries, apart from resource considerations, is the general perception that measles is a relatively mild and unavoidable childhood disease. Collection of data on the morbidity and mortality due to measles in these countries is often necessary in order to generate a higher priority for measles control. This information should convince the public and health care workers that measles continues to be a major killer of children and that this heavy toll will not be changed until measles vaccine is routinely given to all infants and young children. ACKNOWLEDGEMENTS Ths assistance of Dr U. Aye Kyu, Assistant Director of the Burma EPI Programme, and Dr U La Win Maung, Medica! Officer (EPI), is gratefully acknowledged. The work of Dr U Hla Myint, Divisional Health Director of the Rangoon Health Division, and his staff in carrying out the follow-up study of childhood deaths is especially appreciated. Staff of the WHO office in Rangoon facilitated the preparation of the initial report. In addition, we are grateful to all the physicians and hospitals visited in Burma for their contribution of data to this study. The serological testing by the Viral and Rickettsial Disease Laboratory of the California Department of Health Services provided specific data for this report. The assistance of Dr Alan B. Bloch, formerly with the Immunization Division, CDC (Atlanta, USA), in the design of the death certificate follow-up study and for reviewing this article is gratefully acknowledged. The authors would also like to express their heartfelt thanks to Professor U Tin U, Director General, Department of Medical Education, Ministry of Health, for sharing his rich experience in paediatrics with them. RESUME AUCUN CHANGEMENT DANS L'EPIDEMIOLOGIE DE LA ROUGEOLE EN BIRMANIE Pour obtenir les donnees sur la presence et la gravite clinique de la rougeole en Birmanie, on s'est servi d'etudes, publiees et non publiees, effectuees au cours des vingt dernieres anneees et d'enquetes menees en mai et en juin 1983 dans le cadre d'une consultation de l'OMS. Les resultats de trois enquetes serologiques executees durant deux epidemies et d'une etude de contr6le effectuee sur une vaste cohorte d'enfants de 0 a 5 ans, ont tous montre que les taux d'atteinte les plus eleves se situaient dans le groupe d'age de I a 4 ans et qu'a l'age de 10 ans, pratiquement tous les enfants avaient eu une rougeole clinique. Bien que, depuis 1978, le nombre de cas notifies annuellement en Birmanie n'ait e que de 20 000 a 25 000, l'incidence annuelle reelle, fondee sur les resultats de l'enquete de contr6le et des enquetes serologiques, est estimee a environ I million de cas par an en moyenne. Durant les periodes de forte incidence rougeoleuse, comme le premier trimestre de 1983, la rougeole a &6 la cause de pres de 20% de toutes les admissions dans les services pediatriques des h6pitaux de Rangoon. Les maladies respiratoires comme la bronchopneumonie et les bronchites ont constitue la majorite des complications 557 558 J. CHIN & M. THAUNG graves et ont e observees chez 50 A 90% de tous les cas de rougeole admis dans les hBpitaux. L'6tude de contr8le sur tous les deces d'enfants de 6 mois A 10 ans enregistres dans 3 communes de Rangoon pendant 15 mois, de janvier 1982 a mars 1983, a montre que la rougeole etait directement ou indirectement responsable de plus d'un tiers de ces dc&s. Cette etude a egalement fait apparailtre que moins de 20% de ces enfants avaient e hospitalis6s avant leur deces et qu'environ un tiers d'entre eux n'avaient requ aucune assistance m6dicale. Toutes les donn&es dont on dispose et les resultats des etudes men6es en Birmanie depuis 1966 indiquent que l'epidemiologie de la rougeole ne s'y est pas modifi&e. Cette maladie demeure une cause majeure des deces d'enfants dans les pays en d6veloppement oii la vaccination anti-rougeoleuse n'est pas assuree. En plus de la morbidite et de la mortalite tres elevees associees a la rougeole, le poids que cette maladie fait peser sur les services ambulatoires et hospitaliers justifie, sur le plan du rapport couit/rendement, I'administration aussi largement repandue que possible du vaccin anti-rougeoleux dans des pays comme la Birmanie. REFERENCES 1. CHIN, J. ET AL. Filter paper disc method of collecting whole blood for serologic studies in children. American journal of epidemiology, 84: 74-80 (1966). 2. GERSHON, A. A. & KRUGMAN, S. Measles virus. In: Diagnostic procedures for viral, rickettsial and chiamy- dial infections. 5th ed. Washington, DC, American Public Health Association, 1979, pp. 685-686. 3. C. KHAI MING, ET AL. Serologic survey for certain child- hood viral infections in Rangoon, 1971. Journal of tropical medicine and hygiene, 77: 260-266 (1974). 4. BANERJEE, P. C. Report of a measles survey in Sagaing. Burma medical journal, 4: 219-223 (1968). 5. WALSH, J. A. & WARREN, K. S. Selective primary health care: an interim strategy for disease control in develop- ing countries. New England journal of medicine, 301: 967-974 (1979).