BACKGROUND: Effective coverage (EC) is a measure of health service performance. The use of EC, however, is limited by the frequent lack of information about one of its constitutive parameters: the quality of health services, understood as the fraction between observed and an optimum health gain from health interventions. We proposed a novel method that overcomes this issue.

METHODS: The procedure calculated the fraction of avoidable disability (or health-state utilities) in people receiving treatment, over the disability attributable to a disease. This was done using a regression model that predicts disability at individual level, where exposure to the disease without treatment, and with treatment (G2) were key variables. We compared total predicted disability in three scenarios: actual, worst (‘0’ coverage), and plausible (assuming quantile 90 of predicted health gain distribution was the optimum performance in G2). We demonstrated this procedure using data from 4359 individuals aged 50+ at Wave 2 (2004-2005) of the English Longitudinal Study of Ageing (ELSA). Hypertension was used as an example disease. People reporting treatment were considered covered. Disability was measured as an index score, ranging from 0-100.

RESULTS: The prevalence of hypertension was 25.3% [23.9-26.5] and accounted for 1.9% [1.9-1,9] of total disability in the population. The coverage of treatment was 80.5% [78.2-82.9]. In people with hypertension, given current coverage, 48.8% [47.5-50.1] of the predicted disability in the worst scenario would have been avoided (i.e. relative-EC), and 2.0% [2.0-2.1] in the whole population (i.e. absolute-EC). The average health gain in people with treatment was 54.0% [52.9-55.0] (i.e. health benefit), while the avoidable disability against the optimum health gain was 68.6% [67.1-70.2] (i.e. quality). Effective coverage given the optimum health gain was 54.9% [53.6-56.1].

CONCLUSION: We developed a pragmatic way to estimate EC, which overcame the issue of the lack of information about health service quality.