Implementation of Pharmacist-Initiated Orders: “Pharmacist Suggests”

Abstract

INTRODUCTION

The Kingston General Hospital is a 452-bed teaching hospital with 20 full-time pharmacists providing direct patient care and support services to a wide variety of medical and surgical programs. Drug distribution is accomplished through a centralized unit-dose system. Pharmacists and pharmacy technicians use the pharmacy computer system (RxTFCTM, BDM Information Systems Ltd., Saskatoon, Saskatchewan) to maintain patient medication profiles. Pharmacists have the knowledge and skills to ensure optimal drug therapy. However, delays in implementing drug therapy may be caused by the lack of a mechanism for notifying physicians of pharmacists’ recommendations. Before the implementation of “Pharmacist Suggests” orders at Kingston General Hospital, pharmacists primarily used medication memoranda (medication memos) to communicate nonurgent drug order problems and drug therapy recommendations to physicians. Drug order problems included errors, orders for nonformulary or restricted drugs, and drug alerts (e.g., drug allergy, duplication, or drug interaction). Drug therapy recommendations included recommendations to modify drug therapy or perform additional drug monitoring. The medication memos (Appendix 1) were generated from RxTFCTM in the main dispensary and placed in the physician’s orders section of the patient chart as a permanent record. Physicians were required to review any medication memos and write new drug orders if needed. Approximately 700 memos were generated by pharmacists each month, for which the average resolution rate was 86%. Informal feedback from pharmacists, physicians, and nurses indicated that the recommendations in the memos were not addressed in a consistent and timely fashion. A previous study1 at the same hospital showed that the mean resolution time (± standard deviation) for medication memos was 1.92 ± 1.19 days (range 0 to 13 days) for “clarification” discrepancies (i.e., drug name, dose, route, frequency, duplication, or allergy) and 2.46 ± 2.58 days (range 0 to 15 days) for resolution of nonformulary medication issues. In attempts to have memos and drug-related problems resolved more quickly and efficiently, a decision was taken to implement “Pharmacist Suggests” orders. A “Pharmacist Suggests” order was defined as a conditional order written by a pharmacist in the patient chart according to specific order criteria. Physician cosignature was required for these orders to be processed. The purpose of this report is to describe the approval process and the institution’s experience with the implementation of “Pharmacist Suggests” orders.