Accounting for model uncertainty in estimating global burden of disease
David M Vock, Elizabeth A Atchison, Julie M Legler, David RJ McClure, Jamie C Carlyle, Elysia N Jeavons & Anthony H Burton
Volume 89, Number 2, February 2011, 112-120
Table 2. Additional covariates considered for predicting the percentage of diarrhoea deaths caused by rotavirus in each WHO Member State
| Variable | Year data collected | Source |
|---|---|---|
| Life expectancy at birth in males (years) | 2004 | WHOSIS |
| Life expectancy at birth in females (years) | 2004 | WHOSIS |
| Under-5 mortality rate (per 1 000 live births) | 2004 | WHOSIS |
| Infant mortality rate (per 1 000 live births) | 2004 | WHOSIS |
| % DPT3 immunization coverage among 1-year-olds | 2004 | WHOSIS |
| Per capita GNI (in PPP international dollars) | 2004a | WHOSIS if availableb |
| Natural log of per capita GNI (in PPP international dollars) | 2004a | WHOSIS if availableb |
| % urban population | 2005 | WHOSIS |
DTP3, third dose of diphtheria–tetanus–pertussis vaccine; GNI, gross national income; log, logarithm; PPP, purchasing power parity; WHO, World Health Organization; WHOSIS, WHO Statistical Information System.
a The year closest to 2004 was used if 2004 estimates were unavailable.
b When WHOSIS data were not available, we used corresponding estimates from The world factbook.13 The data collected from the The world factbook were actually the gross domestic products (in PPP international dollars). Other sources reporting GNI (in PPP international dollars), such as The World Bank, contained the same missing values as the WHOSIS.14