Assessing the safety of prescribing for people with mental illness through the development and implementation of prescribing safety indicators

Abstract

Background: Measuring the safety of prescribing is vital to understanding and improving patient care. As a result, several sets of prescribing safety indicators have been developed for use across primary and secondary care settings. Despite the fact that prescribing errors and medication related harm may be common in patients with mental illness, there has been limited research focusing on the development and application of prescribing safety indicators specifically for this vulnerable population. Also, while most patients with mental illness are managed entirely in primary care, there is a lack of data exploring potential prescribing safety issues in this setting for this population. Aim: The aim of this PhD is to assess the safety of prescribing for people with mental illness through the development and implementation of a suite of prescribing safety indicators related to mental health conditions and medications, and to use the findings to set an agenda for future research, policy and practice to support safety improvement efforts. Methodology: The prescribing safety indicators development process first involved a comprehensive systematic review to identify potential indicators from existing studies. This was followed by a two-stage e-Delphi consensus building study with a panel of 31 mental health experts from across the UK who rated their agreement with the potential indicators identified by the systematic review, suggested new indicators and rated the likelihood of occurrence and the severity of the most likely outcome of each indicator. Finally, high risk prescribing safety indicators relevant to primary health care were selected, operationalised and applied to the Clinical Practice Research Datalink (CPRD) GOLD. 361 general practices with over 3 million patients were included in a cross-sectional analysis up to September 2019 to examine the prevalence of, variations in, and risk factors for the indicators. In addition, 323 general practices with 4.5 million patients were included in a longitudinal analysis between 2009 and 2019 to examine the change in indicator prevalence over time. To examine variation in indicator rates between practices the intraclass correlation coefficient (ICC) and median odds ratio (MOR) were estimated using two-level logistic regression models. The relationship between patient and practice characteristics and the risk of triggering two composite indicators were assessed using odds ratio derived from multilevel logistic regression models. χ2 tests were used to examine the change in indicators prevalence over time. Findings: A total of 1386 mental health indicators were identified from 70 studies in the systematic review. After refinement, 101 potential prescribing safety indicators were sent to the e-Delphi expert panel, where 42 prescribing safety indicators were considered to be high or extreme risk for patient care. These indicators covered a broad range of prescribing and medication monitoring problems as well as different mental health related drug classes. Of these, 18 potentially hazardous prescribing and 4 inadequate medication monitoring indicators were operationalised and applied to the CRPD. A total of 9.4% of patients at risk (151,469 out of 1,611,129) received at least one potentially hazardous prescription in the third quarter of 2019, and between practices this varied from 3.2% to 24.1% (ICC 0.03, MOR 1.22). A total of 90.2% of patients at risk (38,671 out of 42,879) were exposed to at least one inadequate medication monitoring episode in the same quarter, with between practice variation of 33.3% to 100% (ICC 0.27, MOR 2.84). Patients aged 35-44, females, those receiving more than 10 repeat prescriptions and those living in the most deprived areas were at greatest risk of triggering a prescribing indicator. Patients aged less than 25, females and those with one or no repeat pre