Developing and implementing a model for changing physicians’ prescribing habits – the role of clinical pharmacy in leading the change

Summary Background and objective:  Budgetary constraints led the Israeli Hillel Yaffe Medical Center management to implement policies for reducing expenditure while maintaining the quality of care. For this purpose, the pharmacy services management developed and implemented a three‐tier intervention feedback model for changing physicians’ prescribing habits, and achieving cost‐effective changes in antibiotic utilization. Methods:  A prospective drug utilization evaluation was conducted to profile antibiotic utilization. The results established a base from which a three‐tier feedback, evidence‐based intervention model was built. This model corresponds to the following three hierarchical levels: Level 1 activities involved management actions that influenced all levels of staff and concentrated mainly on the creation of guidelines. Level 2 activities involved the reorganization of the restricted antibiotics prescription authorization system, through the co‐operation of the clinical pharmacy unit and the hospital infection control specialist. Level 3 focussed on clinical pharmacist activities on the wards. The model was implemented and assessed in the hospital from June 2002 until December 2004. Results and discussion:  Implementation of the model resulted in a cumulative decrease of 6473 i.v. antibiotics daily defined doses (DDDs) and a parallel increase in total oral antibiotic DDDs ( Table 1). These changes were especially notable with high‐bioavailability antibiotics and co‐amoxiclav, where over 2.5 years there was a reduction of 2472 and 4752 i.v. DDDs, respectively ( P  < 0·000). The successful implementation of the model resulted in a reduction of 375 000 NIS (∼66 190 €) in pharmacy antibiotic costs, equivalent to 10 i.v. DDDs or 570 NIS (∼102 €) saved per clinical pharmacist working day. 1 Changes in antibiotic utilization pattern over the years 2001–2004Drug Daily defined doses (DDD) 2001 DDD 2002 DDD 2003 DDD 2004 Percent change 2001/2004 Cumulative decrease DDD Statistical significancePO* cefuroxime axetil   7876   6906 2191    663 −91·5% −13 868 <0·00 PO* co‐amoxiclav 10 896 13 283 16 082 16 682 53·1% +13 359 <0·00 I.v. quinolones**   2774   2415 1837   1598 −41·7% −2472 <0·00 PO quinolones** 17 628 21 710 21 846 21 823 23·7% +12 494 0·35 I.v. co‐amoxiclav 1 g   4880   4841 3807   1240 −74·5% −4752 <0·00 I.v. clindamycin + metronidazole   4166   3792 3604   3936 −5·5% −1166 NA Total DDD i.v. antibiotics 47 298 43 700 44 482 47 237 −6473 0·18 Total savings per clinical pharmacist working day***9·8 i.v. treatment days*PO = oral **Quinolones = ciprofloxacin, levofloxacin, moxifloxacin, ofloxacin. ***Calculated for a period of 2·5 years.Conclusions:  Our study demonstrates the successful implementation of a three‐tier model for changing physicians’ antibiotic prescribing.