Global estimates of country health indicators: useful, unnecessary, inevitable?

Abstract

Background: The MDG era relied on global health estimates to fill data gaps and ensure temporal and cross-country comparability in reporting progress. Monitoring the Sustainable Development Goals will present new challenges, requiring enhanced capacities to generate, analyse, interpret and use country produced data. Objective: To summarize the development of global health estimates and discuss their utility and limitations from global and country perspectives. Design: Descriptive paper based on findings of intercountry workshops, reviews of literatureon and synthesis of experiences. Results: Producers of global health estimates focus on the technical soundness of estimation methods and comparability of the results across countries and over time. By contrast, country users are more concerned about the extent of their involvement in the estimation process and hesitate to buy into estimates derived using methods their technical staff cannot explain and that differ from national data sources. Quantitative summaries of uncertainty may be of limited practical use in policy discussions where decisions need to be made about what to do next. Conclusions: Greater transparency and involvement of country partners in the development of global estimates will help improve ownership, strengthen country capacities for data production and use, and reduce reliance on externally produced estimates.

Monitoring universal health coverage within the Sustainable Development Goals: development and baseline data for an index of essential health services

“…Monitoring the health-related SDGs is intended to be a country-led process, and in a time of increasingly sophisticated model-based estimates of levels and trends in health indicators, 6,41,42 more effort should be made to ensure methods are understandable, replicable, and adaptable for use by technical experts in countries. 43 More complex methods might be useful reference points for testing less complicated methods but some approaches, like amenable mortality, 39,44 are less meaningful for countries without high-quality death registration data because they depend on measurement (not prediction) of cause-specific mortality rates.…”

Mortality of neurological disorders in Tanzania: analysis of baseline data from sample vital registration with verbal autopsy (SAVVY)

“…There were no differences in CSMF among males (4.76%) and females (5.27%) (MRF = 0.90, 95% CI: 0.70-1.16). Compared to adults aged 50 years and above, mortality of ND was estimated to be lower among children aged 0-4 (MRF = 0.71, 95% CI: 0.52-0.97), and comparable in children aged [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] (MRF = 1.07, 95% CI: 0.64-1.72) and adults aged 20-49 (MRF = 1.07, 95% CI: 0.64-1.72). Mortality was lower by 23.0% in urban (4.35%) than rural (5.63%) areas (MRF = 0.77, 95% CI: 0.40-0.99).…”

“…Of all ND-related deaths, epilepsy accounted for 18.4%; 4.2% among children < 5 years, 40.9% among children aged [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19]36.9% among adults aged 20-49, and 11.8% among adults older than 50 years (Table 1). Unspecified convulsive epilepsy (98.0%) and other generalized epilepsy and unspecified epileptic syndromes (2.0%) were the category reported for all deaths.…”

“…Model-based estimates from GBD not only are free from coverage limitations but also considerable reliance on geospatial data and experts' opinions [12]. Global estimates in some countries and cases deviate from the true epidemiological profile and thus are deemed unreliable for local planning by policy makers [14].…”

Bridging the data gaps: do we have the right balance between country data and global estimates?

“…An introductory paper considers the usefulness of the expanding volume of global estimates [9]. Three country-specific papers consider the role of estimates in the specific contexts of Chile, Bangladesh and Thailand, respectively [10–12].…”