Academic journal article Bulletin of the World Health Organization

Impact of Measles Vaccination on Childhood Mortality in Rural Bangladesh

Academic journal article Bulletin of the World Health Organization

Impact of Measles Vaccination on Childhood Mortality in Rural Bangladesh

Article excerpt

Impact of measles vaccination on childhood mortality in rural Bangladesh


While measles remains an important cause of death in young children in many developing countries [1,2], controversy persists about the extent to which measles immunization might improve overall child survival in such settings [3]. The limited number of available studies have not resolved this debate. A study of measles immunization in Zaire concluded that it had only a limited impact upon longer-term survival [4]> however, a re-analysis of these data [5], as well as recent studies from Bangladesh and Senegal [6,7], suggest that measles vaccination has significant and sustained effects on childhood mortality.

The above-mentioned study from Bangladesh, a case-control analysis of 536 deaths and 1072 age-sex matched controls, found that measles immunization was associated with a 36% lower child mortality in the period April 1982 to December 1984 [6]. The present paper supplements this earlier analysis in three respects. First, we used a cohort approach to estimate the cumulative probabilities of survival up to late 1985, thus extending the post-vaccination period of observation for almost an additional year. This allowed us to examine more fully the replacement mortality hypothesis--i.e., whether the effects of vaccination are sustained or whether competing risks ultimately offset the short-term improvements in mortality associated with measles vaccination [3,4]. Second, we examined the differential impact on mortality by age at vaccination. Third, we analysed in greater detail potential selection bias among the vaccinees.

Subjects and methods

The data for the present analysis originated from the Matlab rural field station of the International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR, B). Since 1966, ICDDR, B has maintained a system of continuous surveillance of vital events in Matlab. The demographic data that have resulted are unique as far as developing countries are concerned in terms of both their accuracy and size of the population under surveillance (190 000 in 1985) [8].

An experimental maternal and child health/family planning programme was set up in roughly half of the study area in late 1977 [9]. In March 1982, measles vaccine (Schwarz strain) was offered to all children aged 9 months or older in two of the four sub-areas in Matlab (blocks A and C), according to a pre-set plan of phased introduction throughout the whole intervention area [10]. A detailed record-keeping system was maintained on the vaccination status and date of vaccination of all children aged [is less than]5 years in the intervention area. The coverage rates for measles vaccination for children aged 9 months to 2 years in these two blocks were 60-70% over this period (Fig. 1). In November 1985, measles vaccination was extended to the remaining two blocks (B and D)> the coverage levels rose sharply in these latter blocks and equalled or surpassed those in blocks A and C within a few months.

Since measles vaccination was phased in gradually, this facilitates evaluation of its impact on mortality. In the present study, the analysis cohort of vaccinees consisted of all children in blocks A and C who were immunized between the ages of 9 and 60 months from March 1982 to October 1985 (9133 children). These children were compared with a randomly matched group of nonvaccinated children in blocks B and D, according to the following criteria: the latter children had no record of having been vaccinated for measles during the study period (records for only 8 children in blocks B and D showed that they had received measles vaccination during the study period)> each nonvaccinee was born in the same year and month as the corresponding matched vaccinee> and each nonvaccinee must have survived at least up to the date of vaccination of their matched vaccinee. The baseline child mortality rates in the two areas from which vaccinees and nonvaccinees were drawn were similar (Table 1). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.