Afghanistan Mortality Survey 2010 - Measure DHS

The third panel estimates life expectancy for all Afghanistan assuming that the mortality rates inthe South zone are 25 percent higher than those in the rest of Afghanistan, lowering the life expectancy byabout 2 years. However, there is no specific basis to this assumption.The fourth panel combines the modifications done in panels two and three and results in lifeexpectancies for all Afghanistan that are 2.3 to 2.7 years less than that reported for females and 1.1 to 1.9years less for males. Therefore for all Afghanistan life expectancies could vary between 61.5-64.2 yearsfor females and between 61.7-63.6 years for males, values which are far higher than from previouslymodeled estimates.6.5 CONCLUSIONThe levels of adult mortality are obtained using three different sources of information within theAMS 2010: a household roster of deaths during the preceding five years, deaths to siblings of respondentsto the Woman’s Questionnaire, and the survival at the time of the survey of parents of members of thehousehold. All the estimates are in broad agreement with each other given the variation due to samplingerror. The death roster gives values a little higher than the sibling history for the recent past, with aprobability of dying of 86 per 1000 between ages 15-49 for women and 90 for men and between ages 15-59 years of 151 for women and 161 for men. Life table calculations of mortality give life expectancies ofapproximately 64 years for each sex.However, there are indications that the mortality levels may be somewhat understated. The basicdata show substantial digit preference for both numerators (deaths) and denominators (exposure). Sexratios are too high for both sibling and death roster exposure and death rates. These problems are muchmore extensive in the South zone, probably due to security situations that inhibited full interviewing incertain locations and a desire by respondents to protect their families. The Brass Growth Balance methodadds evidence to these conclusions as do comparisons with external sources. However, the omission ofdeaths and exposure tend to offset one another so that the effect on mortality rates is thereby reduced. It iscertainly evident from the data that adult mortality levels have fallen substantially in the last decade andare below those of even the recent past. Estimates of life expectancies based on reported deaths, adjustedunder-5 mortality rates and potential undercounting of deaths results in life expectancies for allAfghanistan that are 2.3 to 2.7 years less than that reported for females and 1.1 to 1.9 years less for males.Therefore, for all Afghanistan life expectancies could vary between 61.5-64.2 years for females andbetween 61.7-63.6 years for males.124 | Adult Mortality