Area level impacts on emergency hospital admissions of the integrated care and support pioneer programme in England: difference-in-differences analysis.

Objective: To examine whether any differential change in emergency admissions could be attributed to integrated care by comparing pioneer and non-pioneer populations from a pre-pioneer baseline period (April 2010 to March 2013) over two follow-up periods: to 2014/2015 and to 2015/2016.
Design: Difference-in-differences analysis of emergency hospital admissions from English Hospital Episode Statistics.
Setting: Local authorities in England classified as either pioneer or non-pioneer.
Participants: Emergency admissions to all NHS hospitals in England with local authority determined by area of residence of the patient.
Intervention: Wave 1 of the integrated care and support pioneer programme announced in November 2013.
Primary Outcome Measure: Change in hospital emergency admissions.
Results: The increase in the pioneer emergency admission rate from baseline to 2014/2015 was smaller at 1.93% and significantly different from that of the non-pioneers at 4.84% (p=0.0379). The increase in the pioneer emergency admission rate from baseline to 2015/2016 was again smaller than for the non-pioneers but the difference was not statistically significant (p=0.1879).
Conclusions: It is ambitious to expect unequivocal changes in a high level and indirect indicator of health and social care integration such as emergency hospital admissions to arise as a result of the changes in local health and social care provision across organisations brought about by the pioneers in their early years. We should treat any sign that the pioneers have had such an impact with caution. Nevertheless, there does seem to be an indication from the current analysis that there were some changes in hospital use associated with the first year of pioneer status that are worthy of further exploration.
Competing Interests: Competing interests: None declared.
(© Author(s) (or their employer(s)) 2019. Re-use permitted under CC BY. Published by BMJ.)