Effects of pharmacist prescribing on patient outcomes in the hospital setting: a systematic review

The objective of the review was to synthesize the best available evidence on the safety and effectiveness of pharmacist prescribing on patient outcomes in patients who present to hospital.

Pharmacist prescribing is legal in many countries. Different models of prescribing include dependent, collaborative and independent. Existing reviews of pharmacist prescribing focus on studies in the community setting, or both community and hospital settings. Other reviews focus on descriptions of current practice or perspectives of clinicians and patients on the practice of pharmacist prescribing. A systematic review on the effects of pharmacist prescribing on patient outcomes in the hospital has not been previously undertaken and is important as this practice can help ease the burden on the healthcare system.

Studies with controlled experimental designs comparing pharmacist prescribing to medical prescribing in the hospital setting were included in the review. Primary outcomes of interest included clinical outcomes such as therapeutic failure or benefit, adverse effects, and morbidity or mortality. Secondary outcomes included error rates in prescriptions, medication omissions on the medication chart, time or proportion of International Normalized Ratios in therapeutic range, time to reach therapeutic range, and patient satisfaction.

A comprehensive three-step search strategy was utilized. The search was conducted in January 2017 in eight major databases from database inception. Only studies in English were included. The recommended Joanna Briggs Institute approach to critical appraisal, study selection and data extraction was used. Narrative synthesis was performed due to heterogeneity of the studies included in the review.

The 15 included studies related to dependent and collaborative prescribing models. In four studies that measured clinical outcomes, there was no difference in blood pressure management between pharmacists and doctors while patients of pharmacist prescribers had better cholesterol levels (mean difference in low density lipoprotein of 0.4 mmol/L in one study and 1.1 mmol/L in another; mean difference in total cholesterol of 1.0 mmol/L) and blood sugar levels (mean difference of fasting blood sugar levels of 15 mg/dL, mean difference of glycosylated hemoglobin of 2.6%). In two studies, pharmacists were better at adhering to warfarin dosing nomograms than doctors (average of 100% versus 62% compliance). In six studies, when prescribing warfarin according to dosing nomograms, equivalent numbers or more patients were maintained in therapeutic range by pharmacist prescribers compared to doctors. The incidence of adverse effects related to anticoagulant prescribing was similar across arms but all six studies were underpowered to detect this outcome. Three studies found that pharmacist prescribers made less prescribing errors (20 to 25 times less errors) and omissions (three to 116 times less omissions) than doctors when prescribing patients' usual medications on admission to hospital or in the preoperative setting. Two studies reported that patients were as satisfied with the services provided by pharmacist prescribers as with doctors.

This review provides low to moderate evidence that pharmacists can prescribe to the same standards as doctors. Pharmacists are better at adhering to dosing guidelines when prescribing by protocol and make significantly less prescribing errors when charting patients' usual medications on admission to hospital.