Assessing efficiency in the UK breast screening programme: does size of screening unit make a difference?

The UK breast screening programme (UK BSP) is organised into a large number of individual screening units. Decision makers need to ensure these units are producing efficiently, particularly as the programme is anticipated to expand. Data envelopment analysis (DEA) was applied to investigate: the relative efficiency of screening units; the impact of screening unit size on efficiency; and how individual units could improve. Sixty-four screening units were categorised into 33 large and 31 small. Data were collected using a national survey and routinely collected data. The overall median efficiency score was 91%, 39 units were inefficient. Variation in efficiency scores was wide. Large units had a median efficiency score of 100% and 12 units were inefficient. Smaller units had a median efficiency score of 95% and 19 were inefficient. This difference was not statistically significant (Mann-Whitney, P=0.076). Forty-two percent of large units and 21% of small units were operating at constant returns to scale (mean difference 0.20, 95% CI: 0.15-0.43). Although there is no systematic difference in efficiency by size of screening unit there are inefficiencies in both large and small units and there is scope for many individual units to improve their use of current resources. It will be necessary for decision-makers to examine the practices of individual screening units before considering options for how best to improve their resource use. DEA can help to identify feasible options.