Exploring the association of the discharge medicines review with patient hospital readmissions through national routine data linkage in Wales: a retrospective cohort study

Abstract

Objective: To evaluate the association of the discharge medicines review (DMR) community pharmacy service with hospital readmissions through linking National Health Service data sets.

Design: Retrospective cohort study.

Setting: All hospitals and 703 community pharmacies across Wales.

Participants: Inpatients meeting the referral criteria for a community pharmacy DMR.

Interventions: Information related to the patient's medication and hospital stay is provided to the community pharmacists on discharge from hospital, who undertake a two-part service involving medicines reconciliation and a medicine use review. To investigate the association of this DMR service with hospital readmission, a data linking process was undertaken across six national databases.

Primary outcome: Rate of hospital readmission within 90 days for patients with and without a DMR part 1 started.

Secondary outcome: Strength of association of age decile, sex, deprivation decile, diagnostic grouping and DMR type (started or not started) with reduction in readmission within 90 days.

Results: 1923 patients were referred for a DMR over a 13-month period (February 2017-April 2018). Provision of DMR was found to be the most significant attributing factor to reducing likelihood of 90-day readmission using χ² testing and classification methods. Cox regression survival analysis demonstrated that those receiving the intervention had a lower hospital readmission rate at 40 days (p<0.000, HR: 0.59739, CI 0.5043 to 0.7076).

Conclusions: DMR after a hospital discharge is associated with a reduction in risk of hospital readmission within 40 days. Linking data across disparate national data records is feasible but requires a complex processual architecture. There is a significant value for integrated informatics to improve continuity and coherency of care, and also to facilitate service optimisation, evaluation and evidenced-based practice.

Keywords: health informatics; health policy; health services administration & management; information technology; organisation of health services.

Figure 1

Information flow diagram depicting the architecture required to link all required patient data to investigate hospital readmission outcome data for those offered a DMR service. DMR, discharge medicines review; NDR, National Data Resource; PEDW, Patient Episode Database for Wales; SQL, Structured Query Language.

Figure 2

Survival analysis looking at the probability of readmission postdischarge for patients who had started a discharge medicines review (DMR) service compared with those who had not, over time.