Configuring balanced scorecards for measuring health system performance: evidence from 5 years' evaluation in Afghanistan.

BACKGROUND: In 2004, Afghanistan pioneered a balanced scorecard (BSC) performance system to manage the delivery of primary health care services. This study examines the trends of 29 key performance indicators over a 5-year period between 2004 and 2008. METHODS AND FINDINGS: Independent evaluations of performance in six domains were conducted annually through 5,500 patient observations and exit interviews and 1,500 provider interviews in >600 facilities selected by stratified random sampling in each province. Generalized estimating equation (GEE) models were used to assess trends in BSC parameters. There was a progressive improvement in the national median scores scaled from 0-100 between 2004 and 2008 in all six domains: patient and community satisfaction of services (65.3-84.5, p<0.0001); provider satisfaction (65.4-79.2, p<0.01); capacity for service provision (47.4-76.4, p<0.0001); quality of services (40.5-67.4, p<0.0001); and overall vision for pro-poor and pro-female health services (52.0-52.6). The financial domain also showed improvement until 2007 (84.4-95.7, p<0.01), after which user fees were eliminated. By 2008, all provinces achieved the upper benchmark of national median set in 2004. CONCLUSIONS: The BSC has been successfully employed to assess and improve health service capacity and service delivery using performance benchmarking during the 5-year period. However, scorecard reconfigurations are needed to integrate effectiveness and efficiency measures and accommodate changes in health systems policy and strategy architecture to ensure its continued relevance and effectiveness as a comprehensive health system performance measure. The process of BSC design and implementation can serve as a valuable prototype for health policy planners managing performance in similar health care contexts. Please see later in the article for the Editors' Summary.