Creation of family planning service quality score using factor analysis

Objectives: Frameworks for assessing service quality in reproductive health/family planning (RH/FP) have traditionally only used the client's perspective, and quantitative analyses of service quality in reproductive health clinics are limited and rarely combine multiple sources of information. We derived a quality index using survey data from facilities, clients and providers of FP services in two countries.

Methods: We used data collected by the Carolina Population Center Alternative Business Models study. It is a multi-stage cluster sample comprising two waves of data collection in Pakistan and Ethiopia in 2001 and 2004: resulting in a total sample of 23664 clients and 3702 health providers in 2505 facilities. Principal components and factor analysis were used to obtain distinct domains of quality, and Cronbach's alpha coefficient assessed the reliability of each scale. A contingent market evaluation scenario with FP experts yielded weights for each of the domains creating a composite index.

Results: The analysis disclosed five distinct domains of FP quality which follow with Cronbach's α shown respectively for Pakistan and Ethiopia 1)availability of FP methods(α=0.95, 0.90); 2) information given by the providers (α=0.72, 0.89); 3)technical competency in abortions (α=0.94, 0.96); 4) client satisfaction (α=0.75, 0.64); 5) service amenities (α=0.79, 0.82) and 6)overall technical competency (Ethiopia only, α=0.86). Domains included both binary and ordinal components with low uniqueness.

Conclusion: Data confirms frameworks posed by Bruce and others, and indicates that generalizable FP quality domains and an overall score are possible. Composite scores can be used in further analyses such as equity, cost and usage.