u5mr_cohort {u5mr} | R Documentation |
u5mr_cohort()
uses the maternal age cohort-derived methods (MAC) through summary
birth history information and maternal age (or time since first birth) to
calculate the under-five mortality estimates.
u5mr_cohort(
data,
women = "women",
child_born = "child_born",
child_dead = "child_dead",
agegrp = "agegrp",
iso3 = "KHM",
svy_year = 2010
)
data |
preprocessed data |
women |
total number of women |
child_born |
children ever born |
child_dead |
children dead |
agegrp |
age grouping or time since first birth |
iso3 |
the |
svy_year |
end of the survey |
In this cohort-derived method, under-five mortality and reference time are estimated through summary birth history information from the mothers and her age or time since her first birth.
In case sample weights are available for the mothers, final variables should be multiplied by these weights before summarizing.
Computational Procedure
Two formulas were used to quantify MAC method:
5q0
component
logit(5q0ijk) = \beta
0i + Uij +
\beta
1i x logit(CDijk /
CEBijk) + \beta
2i x CEBijk +
\beta
3i x PR1 + \beta
4i x PR2 +
\epsilon
ijk
where
i = 5-year maternal age group \epsilon
(15-19, 20-24, ... , 45-49)
j = country
k = survey
CDi = total dead children from maternal age group i
CEBi = total children ever born from maternal age group i
PR1 = ratio of parity among maternal age group 15-19 and parity among maternal age
group 20-24
PR2 = ratio of parity among maternal age group 20-24 and parity among maternal age
group 25-29
Uij = country-specific random effects
All coefficients vary by maternal age group, as indicated by i
and the random
effects also vary by country, as indicated by j
.
Reference time component
reftimeijk = \beta
0i +
\beta
1i x (CDijk /
CEBijk) +
\beta
2i x CEBijk +
\beta
3i x PR1 + \beta
4i x PR2 +
\epsilon
ijk
data.frame
iso3
- total number of women
svy_year
- total number of children ever born
agegrp
- five-year age groups
ref_time
- time between survey year and interpolated year
year
- reference year
q5
- under-five mortality
Rajaratnam JK, Tran LN, Lopez AD, Murray CJL (2010) Measuring Under-Five Mortality: Validation of New Low-Cost Methods. PLOS Medicine 7(4): e1000253. (doi: 10.1371/journal.pmed.100025310.1371/journal.pmed.1000253)
## Example using fake survey data from Cambodia
data(cambodia)
camb <- u5mr_cohort(cambodia, women = "women", child_born = "child_born",
child_dead = "child_dead", agegrp = "agegrp", iso3 = "KHM", svy_year = 1234)
with(camb,
plot(year, q5 * 1000, type = "b", pch = 19,
col = "black", xlab = "Year", ylab = "U5MR per 1000 live births",
main = paste0("Under-five mortality, q(5) in Bangladesh, estimated\n",
"using the maternal age cohort-derived method")))