Association of BCG, DTP, and measles containing vaccines with childhood mortality: systematic review

Objectives To evaluate the effects on non-specific and all cause mortality, in children under 5, of Bacillus Calmette-Guérin (BCG), diphtheria-tetanus-pertussis (DTP), and standard titre measles containing vaccines (MCV); to examine internal validity of the studies; and to examine any modifying effects of sex, age, vaccine sequence, and co-administration of vitamin A. Design Systematic review, including assessment of risk of bias, and meta-analyses of similar studies. Study eligibility criteria Clinical trials, cohort studies, and case-control studies of the effects on mortality of BCG, whole cell DTP, and standard titre MCV in children under 5. Data sources Searches of Medline, Embase, Global Index Medicus, and the WHO International Clinical Trials Registry Platform, supplemented by contact with experts in the field. To avoid overlap in children studied across the included articles, findings from non-overlapping birth cohorts were identified. Results Results from 34 birth cohorts were identified. Most evidence was from observational studies, with some from short term clinical trials. Most studies reported on all cause (rather than non-specific) mortality. Receipt of BCG vaccine was associated with a reduction in all cause mortality: the average relative risks were 0.70 (95% confidence interval 0.49 to 1.01) from five clinical trials and 0.47 (0.32 to 0.69) from nine observational studies at high risk of bias. Receipt of DTP (almost always with oral polio vaccine) was associated with a possible increase in all cause mortality on average (relative risk 1.38, 0.92 to 2.08) from 10 studies at high risk of bias; this effect seemed stronger in girls than in boys. Receipt of standard titre MCV was associated with a reduction in all cause mortality (relative risks 0.74 (0.51 to 1.07) from four clinical trials and 0.51 (0.42 to 0.63) from 18 observational studies at high risk of bias); this effect seemed stronger in girls than in boys. Seven observational studies, assessed as being at high risk of bias, have compared sequences of vaccines; results of a subset of these suggest that administering DTP with or after MCV may be associated with higher mortality than administering it before MCV. Conclusions Evidence suggests that receipt of BCG and MCV reduce overall mortality by more than would be expected through their effects on the diseases they prevent, and receipt of DTP may be associated with an increase in all cause mortality. Although efforts should be made to ensure that all children are immunised on schedule with BCG, DTP, and MCV, randomised trials are needed to compare the effects of different sequences.


Contact with experts in the field
The following groups of experts were consulted and provided information on published and unpublished relevant research in the field: Select comparison with least co-administration of other vaccines, particularly when vs. unvaccinated children 6. Select comparison involving children from the same area 7. Select estimate obtained using landmark (rather than retrospective) approach 8. Select estimate obtained from general population children rather than subgroups (e.g., hospitalized children) 9. Select comparison including the most comprehensive adjustment for potential confounders. 10. Select result for the shortest period of follow-up 11. Select result with the largest sample size 12. Select comparison with vaccination strategies according to the WHO recommendations (e.g., BCG at birth, MCV vaccine at 9 months) 13. Select estimate using the methodological approach claimed to be superior or more correct 14. Select result from more recent article We omitted studies in which all children in one of the comparison groups had two of the vaccines administered simultaneously. The sources of data used in the forest plots are described in the first column of the tables in Appendix 3.

Appendix 4: Excluded studies
Of the 877 full-text articles identified as potentially relevant for the review, 770 were excluded (See Figure 1). The reasons for exclusion for these studies were as follows.

Level (N excluded) Question Answer
Number  Allocation: "Families of comparable status in respect of housing, sanitation and certain other economic and social factors likely to affect the health of children were paired, and one member of each such pair was allotted at random to one of two groups, designated 'Group A' and 'Group B'. All children born into the families of Group A were vaccinated in one year, while all children born into families of Group B in the same year were taken as controls. In the following year this situation would be reversed and so on throughout the duration of the study." Allocation appears to be random in principle, but unlikely to be adequately concealed. No information on similarity of groups.

Risk-of-bias assessments for randomized and quasi-randomized trials of BCG vaccine
Blinding of participants: None. Blinding of vaccine administrators: None.
Co-interventions and departures from allocated intervention: No information.
Assessed at 12-month visit.
Blinding of outcome assessors: Probably not, but objective outcome.
Missing data: Outcome data reasonably complete. Analysis: Unadjusted comparison of numbers of deaths. Allocation: Randomization methods described elsewhere. 16 Blinding of participants: None. 16 Blinding of vaccine administrators: None. 16 No details. Blinding of outcome assessors: Probably not, but objective outcome.

Unclear risk of bias due to confounding (allocation may be unconcealed and no information about similarity of groups) Moderate risk of performance bias due to deviations from intended interventions (participants knew vaccination status) Low risk of detection bias in measurement of outcomes Low risk of attrition bias due to missing outcome data Unclear risk of bias in selection of the reported result Overall: Moderate risk of bias (participants knew vaccination status)
Missing data: Outcome data reasonably complete. Analysis: Cox proportional hazards model. Allocation: "Block randomization procedures have been described in detail elsewhere. Twins were allocated the same treatment to prevent potential confusion regarding who had been vaccinated." A cited reference refers to the factorial period of the trial only: 17 "Once consent was provided, the mother drew an envelope from a bag. Each bag was prepared by the study supervisor and contained 48 envelopes; each envelope contained a lot name. Within each bag were 12 envelopes with lots marked "BCG 6," 12 marked "BCG 7," 12 marked "no BCG 6," and 12 marked "no BCG 7." The numbers "6" and "7" indicated from which of two numbered bottles, "6" or "7," the child should receive treatment (that is, either 25000 IU vitamin A or placebo)." "The envelopes were closed and non-transparent, making it impossible to identify the allocation before the envelopes were opened." "There were few differences in anthropometric measurements, gestational age, or background factors between children who received BCG and controls. The BCG group had more twins and tended to have more mothers who had died before enrollment".

Low risk of bias due to confounding Low risk of performance bias due to deviations from intended interventions Low risk of detection bias in measurement of outcomes Low risk of attrition bias due to missing outcome data Unclear risk of bias in selection of the reported result Overall: Low risk of bias
Blinding of participants: "No placebo for BCG was given."; "a control vaccine may have a nonspecific impact on mortality. Furthermore, if we had used a placebo, control mothers might have believed that the child had received BCG and might therefore not have sought BCG vaccination. We therefore preferred not to use a placebo". Blinding of vaccine administrators: None.
Co-interventions and departures from allocated intervention: No information.
Comment: 11% of children in the control group received BCG during follow-up.
When a death was identified a standard verbal autopsy was conducted by a clinician about 3 months after the death. Blinding of outcome assessors: Probably not, but objective outcome.
Missing data: "Of the 2343 children enrolled in the trial between November 2004 and March 2008, 23 were excluded ( Figure 1). The remaining 2320 children were included in the main analysis" Analysis: Cox proportional hazards model with age as underlying time. Allocation: "a record was prepared for each person who failed to react to tuberculin PPD. All of these records for each school an adjacent area were then sorted b sex and year of birth. An alternate division of the records was then made within each sex and age group. Approximately one-half received the BCG vaccine while the remaining number served as controls. In a small number of instances, the person selected was absent from school and one of those listed for the control group was substituted and the absentee then served as a control."

Low risk of bias due to confounding Low risk of performance bias due to deviations from intended interventions Low risk of detection bias in measurement of outcomes
Blinding of participants: Probably not. "At the same time that the vaccine was given, the control group received an intracutaneous injection of 0.1 cc physiological saline." However children receiving BCG would develop a reaction. Blinding of vaccine administrators: No information.
Co-interventions and departures from allocated intervention: No information. "Neither vaccinated nor controls were isolated before or after vaccination, nor was their mode of living modified." No details. Blinding of outcome assessors: Probably not, but objective outcome.
Missing data: Virtually all children followed for 6 years. There were variations after this due to World War II; attempts made to collect data up to 11 years later but numbers are much lower for later years. Analysis: Unadjusted comparison of counts of deaths. Allocation: "In each subgroup… the children were randomly drawn and alternately assigned to vaccinated or control groups. Thus, the randomized drawing was independent of the process of classifying into subgroups. The formation of the subgroup before alternation reduces the size of population required to ensure unbiased vaccinated and control groups. This alternation was done on a master chart in the main office by a physician who did not do the field work." No significant differences found between vaccine and control groups with the exception of birth weight when split into 13 weight intervals. When regrouped into 3 groups, no significant difference found.

25
Blinding of participants: "In the nonvaccinated, the same procedure was followed except that saline was used instead of the vaccine (placebo). However children receiving BCG would develop a reaction. Blinding of vaccine administrators: No information.
Co-interventions and departures from allocated intervention: BCG was administered at different times depending on whether child was removed at birth/ later and the level of TB risk. It ranged from birth to 3 months old. "There were no appreciable differences in the vaccinated and control groups in any of the above categories" [home visits, roentgenograms, examinations by physicians].
Visited weekly while in foster home. Every 6 months they returned to the clinic. Blinding of outcome assessors: Probably not, but objective outcome.
Missing data: Results table imply deaths known for all participants. However, "some 23 per cent were not followed for the specified time in both groups… There were no significant differences in the total of vaccinated and control subjects lost from the study. However, there were significant differences between the groups in regard to reason for loss. This was due to an excess of vaccinated lost because of delinquency or moving and an excess of controls lost for 'other' reasons". It is not clear whether these are counted in the denominator when presenting mortality data. Analysis: Unadjusted comparison of counts of deaths.

Moderate risk of bias due to confounding (alternation with no allocation concealment) Low risk of performance bias due to deviations from intended interventions Low risk of detection bias in measurement of outcomes
Moderate risk of attrition bias due to missing outcome data (unclear whether outcomes known for 23% lost to follow up were) Unclear risk of bias in selection of the reported result Overall: Moderate risk of bias (no allocation concealment and uncertainty around missing data)  Analysis by conditional logistic regression. DTP: "Most of the vaccinated children received either BCG followed by DTP or the vaccines simultaneously"; no information on proportions. High probability of differential DTP co-intervention (follow-up 6 months from first visit (which was in the first 6 months of life)). MCV: High probability of differential MCV cointervention (follow-up 6 months from first visit (which was in the first 6 months of life)).

High risk of bias in selection of participants into the study (deaths in first month excluded) High risk of bias due to confounding (despite matching, some key confounders were not addressed) Moderate risk of bias in classification of vaccination status High risk of bias due to deviations from intended interventions (high probability of differential co-intervention of DTP and MCV) Low risk of bias in measurement of outcomes
Other: No information about any other co-interventions.
Collected during 6 monthly-annual visits.
Cox proportional hazards model. Follow-up began at first visit (before 7 months). Censoring at 2 nd visit, 6 months after 1 st visit, outmigration or death. DTP: High probability of differential DTP cointervention (follow-up to 5 years). MCV: High probability of differential MCV cointervention (follow-up to 5 years). Other: No information.

Very high risk of bias in selection of participants into the study (first visit took place up to 7 months old, so early effects of BCG on mortality not considered) High risk of bias due to confounding (likely confounding, including by SES) High risk of bias in classification of vaccination status (children assumed unvaccinated when card not seen) High risk of bias due to deviations from intended interventions (likely co-interventions including DTP and MCV) Low risk of bias in measurement of outcomes
All-cause mortality, quarterly visits.
Time conditional HR. No description of any censoring.   DTP: High probability of differential co-intervention with DTP (followed-up to 6 months). MCV: Low probability of differential co-intervention with DTP (followed-up to 6 months).

High risk of bias in selection of participants into the study (appears that follow-up could begin after DTP vaccination) High risk of bias due to confounding (no adjustment for child's health, potential adjustment for post BCG variable (DTP)) High risk of bias in classification of vaccination status (unvaccinated group may include vaccinated children, and bias towards null from landmark approach) High risk of bias due to deviations from intended interventions (high rate of subsequent DTP vaccination) Low risk of bias in measurement of outcomes
Other: Nothing reported for any of the co-interventions.
Cox proportional hazards model. All survival analyses started from age at first examination or 8 days of age, whichever came latest. Children were censored at moving or 6 months of age, whichever came first. Also did separate analyses censoring at the start of the war.

Moderate risk of bias in selection of participants into the study High risk of bias due to confounding (likely residual confounding despite adjustment) High risk of bias in classification of vaccination status (likely that unvaccinated group included assumed unvaccinated children) High risk of bias due to deviations from intended interventions (high degree of co-intervention with DTP) Low risk of bias in measurement of outcomes Moderate risk of bias due to missing outcome data Moderate risk of bias in selection of the reported result Overall: High risk of bias (co-intervention with DTP; likely residual confounding)
Also from this paper Boys vs. girls: Boys RR=0.11 (0.02, 0.57); Girls RR=0.24 (0.06, 0.91); footnote to Table 2 Age:   All-cause mortality likely collected at 3 monthly visits at home.

High risk of bias in selection of participants into the study (follow-up begins after BCG vaccinations for some children) High risk of bias due to confounding (no adjustment for SES and child's health, potential adjustment for post BCG variable (DTP)) High risk of bias in classification of vaccination status (assumed no card meant unvaccinated) High risk of bias due to deviations from intended interventions (high degree of co-intervention with DTP and moderate risk of co-intervention with MCV) Low risk of bias in measurement of outcomes
Unadjusted analysis based on mortality rates in first 12 months, follow-up censored at receipt of DTP. But result is not more extreme than an adjusted analysis that also includes children who had BCG after DTP [MRR=0.41, Table 5].   Table 4 Vitamin A supplementation: DTP: Low probability of differential DTP (follow-up to 6 weeks). MCV: Low probability of differential MCV (follow-up to 6 weeks). Other: No information is provided about any of the cointerventions.

High risk of bias in selection of participants into the study (follow-up begins at 1 week, after some BCG vaccinations) High risk of bias due to confounding (no adjustment for SES or child's health, likely confounding SES) High risk of bias in classification of vaccination status (retrospective collection of vaccination data) Moderate risk of bias due to deviations from intended interventions
All-cause mortality Visits / computerized system.
Unadjusted mortality ratios. Children included in each group from receipt of vaccine until receipt of another vaccine or any other exit criteria. Exit criteria were 9 months of age, end of study or migration. DTP: High probability of differential DTP co-intervention (follow-up to 8 months; censored in analysis). MCV: Low probability of differential MCV co-intervention (follow-up to 8 months). Other: No information provided for any pre-defined cointerventions.

Moderate risk of bias in selection of participants into the study Very high risk of bias due to confounding (unadjusted analysis, with importantly different ages in vaccinated and unvaccinated periods) Moderate risk of bias in classification of vaccination status Moderate risk of bias due to deviations from intended interventions Low risk of bias in measurement of outcomes
No information, likely collected at monthly visits.
Cox proportional hazards model. Estimates refer to BCG as most recent vaccine. Children were censored at 8 months. Absent children censored in analysis until they were again examined. Small differences in MR for BCG between retrospective and landmark approaches. DTP: High probability of differential DTP cointervention (follow-up to 6 months). MCV: Low probability of differential MCV cointervention (follow-up to 6 months).

Moderate risk of bias in selection of participants into the study High risk of bias due to confounding (no adjustment for SES or child's health) High risk of bias in classification of vaccination status (assumptions about non-vaccination) High risk of bias due to deviations from intended interventions (high probability of co-intervention with DTP and MCV) Low risk of bias in measurement of outcomes Moderate risk of bias due to missing outcome data Moderate risk of bias in selection of the reported result Overall: High risk of bias (no adjustment for SES or child's health; assumptions about non-vaccination)
Other: No information about other co-interventions.
Reported during monthly demographic surveillance. Determined by verbal autopsy.
Cox proportional hazards model with propensity score adjustment. All vaccine included in model. Censoring at migration or end of the study period. DTP: High probability of differential DTP cointervention (follow-up to 24 months; censored at receipt of DTP). MCV: High probability of differential MCV cointervention (follow-up to 24 months; censored at receipt of MCV). Other: No information provided for any of the co-interventions.

High risk of bias in selection of participants into the study (children had to survive to 29 days to be included) High risk of bias due to confounding (no adjustment for SES or child's health, adjustment for future DTP) Moderate risk of bias in classification of vaccination status High risk of bias due to deviations from intended interventions (high probability of co-intervention with DTP) Low risk of bias in measurement of outcomes
No information. Cox proportional hazards model. Censored at 24 months of age, registration of next vaccine, death or migration.   Table 3 Observational comparison. Based on children born between 1986 and 1999. 37,894 children were followed between 6 weeks and 9 months of age. Selected for inclusion in review: Children who received BCG only (N=670 by subtraction; top row of Figure 2) versus children who received BCG followed by DTP1 (N=5740   Table 2 Case-control study. Cases: All children who died between January 1986 and October 1987 and before the age of 3 years. Controls: selected from register of all children in the area, matched for age, sex and village of residence. "Some 74 cases and 230 controls were available for analysis" with 1-4 matched controls per case. "Children who were born and died within the interval of about 3 months between surveys were rarely reported. 2 infants who were known to have died within one month of birth were excluded. All other children had at least attained 4 months of age." Selected for inclusion in review: 91 children who were unvaccinated and 41 children who had 1 DTP dose.

Very high risk of bias in selection of participants into the study (included children were 6-35 months, so early effects of DTP on mortality not considered) High risk of bias due to confounding (likely confounding) Moderate risk of bias in classification of vaccination status Very high risk of bias due to deviations from intended interventions (high proportion of co-administration with BCG and with MCV) Low risk of bias in measurement of outcomes Moderate risk of bias due to missing outcome data Moderate risk of bias in selection of the reported result Overall: Very high risk of bias (selection of children into the study; high proportion of co-administration with BCG; high proportion of co-administration with MCV)
Ghana

Very high risk of bias in selection of participants into the study (follow-up begins after DTP vaccinations, and restriction to sample determined by subsequent MCV) High risk of bias due to confounding (no adjustment for SES or child's health) High risk of bias in classification of vaccination status (children without vaccination card assumed unvaccinated) Moderate risk of bias due to deviations from intended interventions Moderate risk of bias in measurement of outcomes Moderate risk of bias due to missing outcome data Moderate risk of bias in selection of the reported result Overall: Very high risk of bias (follow-up begins after DTP vaccinations, and restriction to sample determined by subsequent MCV)
Also from this paper Boys vs. girls: Boys RR=0.21 (0.01, 3.34); Girls RR=2.33 (0.14, 39.16); OPV: Almost always coadministered with DTP. MCV: Moderate probability of differential co-intervention with MCV (children potentially followed up to maximum of 12 months, and 791 children received MCV between 7 and 11 months). Other: No information provided for any other cointerventions.
Information on mortality was obtained at subsequent visits (meaning children had to be visited twice to be included in the study). Cox proportional hazards model. "There was no loss to follow-up because it was always possible to get information on all children from relatives living in the same compound". Analysis repeated excluding children considered unvaccinated because they had no card. Observed during the hospital stay.

High risk of bias in selection of participants into the study (follow-up begins after DTP vaccinations) High risk of bias due to confounding (likely confounding not adjusted for) High risk of bias in classification of vaccination status (assumed no card meant unvaccinated) High risk of bias due to deviations from intended interventions (likely MCV co-intervention) Low risk of bias in measurement of outcomes
Mantel-Haenszel method, stratified by age and number of doses. MCV: Low probability of differential MCV cointervention ("A total of 13 children received MCV at 4.5-5 months as part of a trial of early MCV"; children censored at MCV). Other: Sample comes from a trial also of vitamin A, so this was balanced. More than two thirds of the children in the no DTP group received DTP during follow-up.

Very high risk of bias in selection of participants into the study (restriction to hospitalized sample, follow-up begins after DTP vaccinations, and restriction to sample determined by subsequent MCV) High risk of bias due to confounding (limited adjustment for SES or child's health) High risk of bias in classification of vaccination status (children without vaccination card excluded) Moderate risk of bias due to deviations from intended interventions Low risk of bias in measurement of outcomes
All-cause mortality collected at 3 monthly visits. Verbal autopsy was conducted by a clinician 3 months after death.
Cox proportional hazards model. Censoring at death, MCV, migration or the 6 month visit, whichever came first.

High risk of bias in selection of participants into the study (follow-up begins after DTP vaccinations) High risk of bias due to confounding (likely confounding from SES and child's health) Moderate risk of bias in classification of vaccination status High risk of bias due to deviations from intended interventions (substantial DTP vaccination in the no DTP group) Low risk of bias in measurement of outcomes Moderate risk of bias due to missing outcome data Moderate risk of bias in selection of the reported result Overall: High risk of bias (likely confounding; DTP co-intervention in the no-DTP group)
Also from this paper Boys vs. girls: Boys RR=2.48 (0.61, 10); Girls RR=7.18 (1.53, 33.7); Table 3 India All-cause mortality at 3monthly visits at home.
Unadjusted analysis based on mortality rates in first 12 months.  All-cause mortality from visits/ computerized system.

Moderate risk of bias in selection of participants into the study High risk of bias due to confounding (no adjustment for SES or child's health, likely confounding by SES) High risk of bias in classification of vaccination status (retrospective collection of vaccination data) Moderate risk of bias due to deviations from intended interventions Low risk of bias in measurement of outcomes
Unadjusted mortality ratios. Children included in each group from receipt of vaccine until receipt of another vaccine or any other exit criteria. Exit criteria were 3 years of age, end of study or migration. Sensitivity analyses conducted using 2 week lag period.      Allocation: "The mothers or guardians were asked to select an envelope with the allocation number that defined to which group the child belonged. We randomised children to one of three equal sized groups by using block randomisation with 24 envelopes per bag. Block randomisation was organised within the whole study area and not by health centre. The data manager, who was not involved in the recruitment of children, prepared bags with 24 numbered envelopes indicating the randomisation group. These numbers could not be seen by the doctor informing and obtaining consent from the mothers or guardians. A new bag was opened only when the previous one was empty." "No major differences existed in the demographic, socioeconomic, and health related background factors between children included in the early two dose measles vaccine arm (group A) and the two groups with measles vaccine at 9 months of age (groups B +C) (table 1)."

Moderate risk of bias in selection of participants into the study High risk of bias due to confounding (no adjustment for child's health) High risk of bias in classification of vaccination status (children with no information included as unvaccinated, and further bias towards null from landmark approach) High risk of bias due to deviations from intended interventions (high probability of MCV) Low risk of bias in measurement of outcomes
Blinding of participants: No, "No placebo was given." Blinding of vaccine administrators: No, "However, the health workers, doctors, and nurses in the paediatric ward and at the health centres do not inspect the vaccination cards of sick children to guide the treatment, and they were not aware of the purpose of the study." Co-interventions and departures from allocated intervention: None apparent.
Visit to health centre at 9 plus 3 monthly home visits. Blinding of outcome assessors: Probably no, although the outcome is objective. "However, the health workers, doctors, and nurses in the paediatric ward and at the health centres do not inspect the vaccination cards of sick children to guide the treatment and they were not aware of the purpose of the study." Missing data: "The study was not perfectly balanced with respect to drop-out. As we tested specific strains of measles vaccine, children who had received measles vaccine elsewhere before they attended the 9 month vaccination session were censored in the study because the strain and quality of their measles vaccine would not be known. We excluded more children at the 9 month vaccination session in groups B and C than in group A (fig 1)." 231 children were excluded from the main analysis (80 children had had measles infection before enrolment at 4.5 months of age (five died), 131 children were enrolled within 25 days of the third DTP vaccination as opposed to 4 weeks after (seven died), 18 children were enrolled twice (none died); 2 children who had the wrong age recorded; they were both one year older than originally assumed (neither died). "A further 17 children received the wrong type of measles vaccine at 9 months of age and they have been censored in the analysis from the date of this vaccination (none died)." Analysis: Cox proportional hazards model with age as underlying time. Allocation: "By a system of random numbers, these children were divided into two equal groups".

Low risk of bias due to confounding Moderate risk of performance bias due to deviations from intended interventions (participants knew vaccination status) Low risk of detection bias in measurement of outcomes
Blinding of participants: Probably yes, since placebo was apparently used. "The other group was a control one, receiving gamma globulin and inert material identical with that used in the culture of the virus". Blinding of vaccine administrators: Unclear. Another study reported in the paper is described as "This was not a blind study, since the investigators knew which children had received measles vaccine", suggesting that this might not be the case for the current study. Co-interventions and departures from allocated intervention: None apparent.
Not clear. "At the end of the trial period at the Ilesha Hospital 1962 record cards were Available"; "Among the children in this trial there were 17 known deaths". It is not clear whether these two sources are linked. The paper reports that "the follow-up had to be made in a crowded and overworked clinic". Blinding of outcome assessors: Probably no, although the outcome is objective.  Analysis by conditional logistic regression.

Moderate risk of bias in selection of participants into the study High risk of bias due to confounding (despite matching, some key confounders were not addressed) Moderate risk of bias in classification of vaccination status Insufficient information for bias due to deviations from intended interventions Low risk of bias in measurement of outcomes
Moderate risk of bias due to missing outcome data Moderate risk of bias in selection of the reported result Overall: High risk of bias (key confounders were not addressed) Also from this paper Age: Table 2 Burundi Followed-up for 6 months. 67 children received a non-measles vaccine after 9 months of age and had missed an opportunity for MCV.       Death rates during one year of followup in children aged 6-11 months at the beginning of 1981. All cause mortality and measles deaths. "Health problems are discussed with the village committee, which has been given the responsibility of registering deaths." "Inquiries have been made about children who do not turn up for a re-examination in order to find out if they have died, moved or left the area temporarily."

High risk of bias in selection of participants into the study (enrolment after vaccination likely to have happened) High risk of bias due to confounding (crudely adjusted analysis) Moderate risk of bias in classification of vaccination status Insufficient information to assess risk of bias due to deviations from intended interventions Low risk of bias in measurement of outcomes
Unadjusted counts provided for vaccinated and unvaccinated children (also rates standardized to one year of follow-up, but not for both comparator groups).

Insufficient information to assess risk of bias in selection of participants into the study High risk of bias due to confounding (unadjusted analysis, previous vaccines not considered) Insufficient information to assess risk of bias in classification of vaccination status (cannot tell how vaccination status was defined) Insufficient information to assess risk of bias due to deviations from intended interventions Low risk of bias in measurement of outcomes
Moderate risk of bias due to missing outcome data Moderate risk of bias in selection of the reported result Overall: High risk of bias (unadjusted analysis)