Establishing a service improvement network to increase access to care and improve treatment outcomes in community mental health: a series of retrospective cohort studies

Abstract Background National reports using aggregate data have demonstrated that referrals to improving access to psychological therapies (IAPT) services increase year-on-year, but some groups struggle to access the services and outcomes are variable. To identify moderators of access and outcomes, individual patient data are required to control for confounding factors and mitigate ecological biases, and a collaborative effort is needed to carry through findings to local policy and practice. We created a network of services for these purposes. Methods As part of the network, a dataset was formed using routinely collected anonymised data from all individual patients referred to eight IAPT services across the north, central, and east London areas, UK (total population n=483 683; mean age 37·83 years [SD 14·40]; female n=311 790 [66·25%], male n=158 860 [33·75%]; Black and minority ethnic n=137 373 [35·40%], White n=250 740 [64·60%]). Cohort studies were done using pre-treatment clinical, sociodemographic, treatment-related, and care pathway data, to investigate the effects of service-level changes to care, and identify groups for whom there are inequalities in access or poorer treatment outcomes. Primary analyses involved regression modelling and propensity score matching. NHS ethical approval was not required for this study (confirmed by the Health Research Authority July 2020, reference number 81/81). The data were provided by the IAPT services for evaluation as part of a wider service improvement project conducted in accordance with the procedures of the host institution and the NHS Trusts which operate the IAPT services (project reference 00519-IAPT). Findings After adjustment for all available confounders, increases in the number of treatment sessions received per patient (odds ratio 1·09, 95% CI 1·08–1·10), reductions in non-attended sessions (0·85, 0·84–0·86), and matching treatments to diagnoses (based on National Institute for Health and Care Excellence guidelines) were associated with better odds of symptomatic remission (0·91, 0·88–0·95). Furthermore, people aged 65 years and older had better outcomes than younger patients (1·33, 1·24–1·43), but were under-represented, and people younger than 25 years not in employment, education, or training (NEETS) had worse outcomes (0·68, 0·63–0·74). Social gradients were observed: people living in the least deprived areas had considerably better outcomes than those in more deprived areas, despite similar levels of access to care and referrals (1·28, 1·07–1·53). Consequently, the services improved the system of appointment reminders to reduce non-attendance, offered training and consultations to staff on matching treatment to diagnoses, and have set up initiatives to encourage more referrals of older adults and improve access and outcomes for NEETS. Interpretation These studies demonstrate the value in appropriate data sharing aimed specifically at service improvement. Collaboration between services can better inform potential area-level modification to treatment pathways to improve access and care for adults with depression and anxiety. Funding Wellcome Trust (grant number 201292/Z/16/Z).