Strategies to improve insulin prescribing in UK hospitals: results from a national survey

Introduction : Insulin prescribing errors in hospital are a common and costly problem and can result in patient harm. Recommended strategies to improve inpatient insulin prescribing include the use of dedicated insulin prescription forms, electronic prescribing (EP), clinician education, self‐administration policies and access to specialist diabetes pharmacists. Despite efforts to improve insulin safety, results from national audits show considerable variation in insulin errors across hospitals1. The extent to which hospitals currently adhere to insulin prescribing recommendations is currently unknown, as are details of how prescribing systems are used to minimise insulin prescribing errors.

Aims : To determine the systems currently used to prescribe subcutaneous insulin for in‐patients with diabetes and describe the current use of strategies to improve the quality of in‐patient insulin prescribing.

Methods : We sent a self‐administered, cross‐sectional postal questionnaire to chief pharmacists in 175 acute NHS hospital trusts in the United Kingdom in January 2019, followed by an online version to help increase response rate. The questionnaire was iterated with academics and specialist diabetes pharmacists to enhance face validity. Questions concerned the use and functionality of electronic and paper systems used to prescribe subcutaneous insulin, and strategies designed to reduce insulin prescribing errors. We met with our hospital's diabetes patient public involvement group and multidisciplinary inpatient group throughout the study, who supported the study objectives, contributed to its design and interpreted results from their unique perspectives. The main outcome for analysis was a binary variable (yes/no) indicating the use of strategies and system functions relating to inpatient subcutaneous insulin prescribing. Data were descriptively analysed in SPSS (IBM v24). Chi‐squared and Fishers Exact tests were used to determine associations between sets of two categorical data (e.g. hospital type and EP), and t‐tests were used to determine differences between categorical and continuous data (e.g. diabetes pharmacist and number of strategies used) The survey was validated using exploratory factor analysis. Consistency reliability was estimated with Cronbach's alpha.

Results : Ninety‐five hospital trusts responded (54%) across all four countries of the UK (55 postal, 40 online). Inpatient EP of insulin was reported in 40% of hospitals, most of which were English teaching hospitals. Out of 57 (60%) hospitals not using EP for insulin, 38 (67%) used dedicated insulin prescribing charts. The functionality of both electronic and paper‐based systems to enable the safe prescribing of insulin for inpatients varied. The availability of specialist diabetes pharmacists to support the safe prescribing of insulin was low (29%) but was positively associated with the use of a greater number of insulin prescribing strategies (P = 0.002). The use of specific strategies to improve insulin prescribing quality varied greatly between respondent hospitals, and there was no single strategy that all hospitals were using. Sixty‐three percent of hospitals had insulin self‐administration policies, less than half included mandatory insulin training (46%), and only 31% used insulin passports.

Conclusions : There is potential to optimise the functionality of prescribing systems and use of strategies to improve the safe prescribing of insulin in hospitals in the United Kingdom. The wide variation in the use of insulin error reduction strategies may be improved by the availability of specialist diabetes pharmacists who can support their design and implementation. The anonymity of respondents limited our ability to clarify any incomplete or ambiguous answers, but this was not considered to be problematic. Although it may have been useful to incorporate hospital insulin error data, we felt that it would be methodologically inappropriate in this study. Further exploration of the implementation, use, and impact of insulin prescribing interventions on insulin errors could be achieved with the use of appropriate qualitative or evaluation approaches in future studies.

Reference : 1. NaDIA HQIP. National Diabetes Inpatient Audit England and Wales 2017. 2018. https://files.digital.nhs.uk/pdf/s/7/nadia‐17‐rep.pdf (accessed 15 Nov 2019).