Pharmacist prescribers

Last month's research roundup provided you with an overview of three articles all with a relationship to prescribing in alcohol dependence (Robertson, 2022). This month, we will be reviewing articles looking at pharmacist prescribing. The first article reviews a collaborative pharmacist prescribing model in an emergency department. The others look at prescribing by pharmacists in primary care from two perspectives, relieving the workload of GPs and using advanced practice pharmacists.

This original piece of evaluation research, published online in May 2022 in the Journal of Research in Social and Administrative Pharmacy sought to assess the safety and accuracy of inpatient medication charts within a pharmacist collaborative prescribing model as an intervention, in direct comparison with the usual medical model, defined as a control, in an emergency department (ED) (Ogilvie et al, 2022). This was set up as a randomised controlled trial with the control being prescribed by medical practitioners as opposed to the intervention group having the prescription charts completed by pharmacists. Although not meeting all CONSORT (Moher et al, 2010) criteria for a randomised controlled trial (RCT) protocol, it did follow enough guidelines to be considered quasi-RCT. The trial was conducted between October 2016 and September 2017 in a medium-sized Australian hospital ED. A secondary outcome of interest was the evaluation of venous thromboembolism (VTE) prophylaxis prescribing and any differences in this between pharmacist prescribers and the control group of doctors. Adult patients presenting at the ED were recruited for the study within a referral window for medical admissions between Monday to Friday, 7:30 am–2 pm. Randomisation to either intervention or control was done a four-block randomisation method until the necessary sample size, defined by using a power calculation as 250 individual drug prescriptions, was achieved. The patient was blinded to the intervention but for practical and legal reasons, the prescriber could not be. The medication charts were reviewed and audited using validated audit tools. This was done retrospectively and independently, to reduce researcher bias. A total of 94 patients were recruited, (48 to control and 46 for intervention) with 769 medication orders overall. Exclusions for accidental prescribing by a medical prescriber on a pharmacist-initiated patient or a patient discharged/transferred lead to a final cohort size of 38 in the control arm (with 357 medication orders) and 35 in the intervention group (with 412 medication orders).