Medication safety at the interface: evaluating risks associated with discharge prescriptions from mental health hospitals.

ABSTRACT

WHAT IS KNOWN AND OBJECTIVE: When compared to general hospitals, relatively little is known about the quality and safety of discharge prescriptions from specialist mental health settings. We aimed to investigate the quality and safety of discharge prescriptions written at mental health hospitals. METHODS: This study was undertaken on acute adult and later life inpatient units at three National Health Service (NHS) mental health trusts. Trained pharmacy teams prospectively reviewed all newly written discharge prescriptions over a 6-week period, recording the number of prescribing errors, clerical errors and errors involving lack of communication about medicines stopped during hospital admission. All prescribing errors were reviewed and validated by a multidisciplinary panel. Main outcome measures were the prevalence (95% CI) of prescribing errors, clerical errors and errors involving a lack of details about medicines stopped. Risk factors for prescribing and clerical errors were examined via logistic regression and results presented as odds ratios (OR) with corresponding 95% CI. RESULTS AND DISCUSSION: Of 274 discharge prescriptions, 259 contained a total of 1456 individually prescribed items. One in five [20·8% (95%CI 15·9-25·8%)] eligible discharge prescriptions and one in twenty [5·1% (95%CI 4·0-6·2%)] prescribed or omitted items were affected by at least one prescribing error. One or more clerical errors were found in 71·9% (95%CI 66·5-77·3%) of discharge prescriptions, and more than two-thirds [68·8% (95%CI 56·6-78·8%)] of eligible discharge prescriptions erroneously lacked information on medicines discontinued during hospital admission. Logistic regression analyses revealed that middle-grade [whole discharge prescription level OR 3·28 (3·03-3·56)] and senior [whole discharge OR 1·43 (1·04-1·96)] prescribers as well as electronic discharge prescription pro formas [whole discharge OR 2·43 (2·08-2·83)] were all associated with significantly higher risks of prescribing errors than junior prescribers and handwritten discharges, respectively. Similar findings were reported at the individually prescribed item level. Middle-grade prescribers were also more likely to make both non-psychotropic and psychotropic prescribing errors than their junior colleagues [individual item OR 4·24 (2·14-8·40) and OR 1·70 (1·16-2·48), respectively]. WHAT IS NEW AND CONCLUSION: Discharge prescriptions issued by mental health NHS hospitals are affected by high levels of prescribing, clerical and communication errors. Important targets for intervention have been identified to improve medication safety problems at care transfer.