Implementation of medication reviews in community pharmacies and their effect on potentially inappropriate drug use in elderly patients Martina Teichert, Susan Noyon Luijben, Anouk Wereldsma, Ton Schalk, Jacqueline Janssen, Michel Wensing & Peter de Smet International Journal of Clinical Pharmacy International Journal of Clinical Pharmacy and Pharmaceutical Care ISSN 2210-7703 Int J Clin Pharm DOI 10.1007/s11096-013-9794-8
1 23
Your article is protected by copyright and all rights are held exclusively by Springer Science +Business Media Dordrecht. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.
1 23
Author's personal copy Int J Clin Pharm DOI 10.1007/s11096-013-9794-8
RESEARCH ARTICLE
Implementation of medication reviews in community pharmacies and their effect on potentially inappropriate drug use in elderly patients Martina Teichert • Susan Noyon Luijben • Anouk Wereldsma • Ton Schalk • Jacqueline Janssen Michel Wensing • Peter de Smet
•
Received: 24 October 2012 / Accepted: 11 May 2013 Springer Science+Business Media Dordrecht 2013
Abstract Background In 2008 recommendations were launched to prevent medication-related hospital admissions in the Netherlands. Elderly patients using several drugs on a chronic basis were among the target group. Pharmacy-led medication reviews (MRs) were identified as having potential for improving patient safety. Objective This observational study evaluated the implementation success rate of performing all five steps of a complete MR for patients and changes in the presence of nine issues of potentially inappropriate medication (PIM) use. This change was compared between patients with a complete MR (intervention group, IG) and a reference group (RG) who attended the same pharmacy; all patients were eligible for MR, but only selected patients formed the IG. Setting Dutch community pharmacy. Method After appropriate training, the rate of IG with complete MRs was measured by pharmacists registering the various MR steps in the MR
Electronic supplementary material The online version of this article (doi:10.1007/s11096-013-9794-8) contains supplementary material, which is available to authorized users. M. Teichert M. Wensing P. de Smet Scientific Institute for Quality of Healthcare (IQ Healthcare), Radboud University Nijmegen Medical Centre, 6500 HB, PO Box 9101, Nijmegen, The Netherlands M. Teichert (&) T. Schalk P. de Smet Koninklijk Nederlandse Maatschappij ter bevordering der Pharmacie, 2514JL The Hague, The Netherlands e-mail:
[email protected] S. N. Luijben A. Wereldsma Achmea, 2314 AB Leiden, The Netherlands T. Schalk J. Janssen Stichting Farmaceutische Kengetallen, 2514 JL The Hague, The Netherlands
tool. Patients were eligible for a MR if aged C65 years with C5 drugs being used chronically. Main outcome measures The success rate of implementing MRs with five steps completed; the presence of nine PIMs for both study groups. Results In pharmacies with specifically trained pharmacists, 63 % of selected patients received a complete MR. This was 12 % higher than in pharmacies without trained pharmacists. PIMs reported at study start declined in the IG and at study end had decreased by an average of 19 % (with a range between 34 and 100 % per PIM); this decrease did not significantly differ from the RG. Conclusion Additional efforts are needed to improve the implementation of pharmacist-led MRs in order to realize its full potential in general practice, and for a substantial decrease of PIMs to occur in susceptible patients. These efforts should focus on training courses and additional support using computerized systems to share information with GPs and to register MR activities, together with sufficient financial reimbursement. Keywords Drug safety Elderly Implementation Medication review Potentially inappropriate medication The Netherlands
Impact of study findings on practice •
•
Twenty five percent of Dutch elderly patients, living in the community and using several drugs on a chronic basis, receive at least one potentially inappropriate medication (PIM). After complete medication reviews (MR), the total number of PIMs in Dutch elderly patients, using several drugs on a chronic basis, decreases by 19 %. Per specific PIM the decrease will vary between 34 and 100 %.
123
Author's personal copy Int J Clin Pharm
Introduction Drugs are essential for the prevention and treatment of disease. However, they can also cause harm because of adverse effects and medication errors. In the Netherlands, studies have shown that about 5 % of unplanned hospitalizations are drug-related [1, 2]. Similar findings have been reported in other countries [3, 4]. A striking conclusion from these studies was that the majority of all the potential drug-related hospital admissions could be assigned to a limited number of frequently used drugs; these included anticoagulants, Non-Steroidal Anti-Inflammatory Drugs (NSAIDs), opioids and blood glucose lowering drugs [5]. As a consequence, specific recommendations have been developed to reduce potentially inappropriate medication (PIM) use, and to avoid drug-related hospital admissions. Community pharmacists in the Netherlands have responsibility for the effective and safe use of drugs in their patients. Because patients in the Netherlands usually visit a single pharmacy [6], pharmacists can maintain complete medication histories of their patients. These are used for medication surveillance and to identify PIMs. Several randomized controlled trials (RCTs) and observational studies of pharmacist-led medication reviews (MRs) have shown that pharmacists are able to identify drug-related problems and initiate changes in drug therapy [7–15]. The most promising strategy to improve drug safety is implementation of MRs that use information from pharmacists and general practitioners (GPs) with consideration of the individual’s needs [16].
Aim of the study This study evaluated the success rate of implementing pharmacist-led MRs in Dutch community pharmacies and the effect of MRs on specific PIMs present at study start, compared between an intervention and reference cohort.
Method Design We conducted a comparative cohort study in community pharmacies in the Netherlands between May 2009 and December 2011. All Dutch community pharmacies were invited to participate. The pharmacists conducted MRs, with their collaborating GP, for elderly patients taking multiple drugs on a chronic (long-term) basis (see patient inclusion criteria for details). The structure of MRs was
123
based on earlier research [16] and consisted of five steps as per Supplementary Table 1. Participating pharmacists were offered training courses, a web-based MR tool, and reimbursement for complete MRs. The MR tool was used to register MR activities, to record the amount of time spent undertaking MRs, and to identify and measure PIMs—all on a monthly basis. PIMs were reported from dispensing data collected by the Dutch Foundation of Pharmaceutical Statistics (SFK). Pharmacists selected their patients using the MR tool; patients who met the inclusion criteria (see below) formed the intervention group (IG) if they had a MR identified by a pharmacist who had attended the training courses. Patients from the same pharmacies that met the patient inclusion criteria, but were not selected, were used as the reference group (RG). During the study, the presence of PIMs in the IG and RG groups was tracked using monthly SFK dispensing data. The only pharmacies that could contribute to the study were those who supplied dispensing data to the SFK database during the study period, and that used the MR tool for registering the various MR steps. Because pharmacists could start doing MRs at any time between May 2009 and June 2011, the recording of PIM data for each pharmacy began in the month that the pharmacists selected their patients for MR; for both the IG and RG groups the last PIM data were collected in December 2011.
Patient inclusion criteria According to risk factors identified in previous research [1], patients were eligible for MRs if they were 65 years or older and were taking at least five different drugs on a chronic (long-term) basis, of which at least one drug had to belong to the Anatomical Therapeutic Chemical (ATC) Classification of drugs from the World Health Organization (WHO), wherein ‘A’ stands for ‘alimentary tract and metabolism’ and ‘B’, ‘blood and blood forming organs’ [17]. These two drug classes include drugs used for the treatment of diabetes and as anticoagulants, which have shown to be associated with potentially preventable hospital admissions [16]. When the patients were selected, the pharmacists were asked to pay close attention to any PIMs that were reported by the web-based MR tool. Primary outcome measures 1. 2.
The success rate of implementing MRs with five steps completed; The changes in the number of PIMs present in patients at study start, compared between an intervention and reference cohort.
Author's personal copy Int J Clin Pharm
Implementation strategy for complete MRs Training course At the start of the project, pharmacists were invited for two days of training. Pharmacists who attended all of the training were regarded as ‘trained pharmacists’. Professional trainers explained how to analyse patient medication histories efficiently, detect and prioritize problems, and formulate improvements. The training focused on PIMs from the study of hospital admissions related to medication (HARM study [1]), drugs that have the propensity for causing harm, risks for potentially avoidable hospital admissions, and recommendations from guidelines that could improve medication safety. Finally, the use of the web-based MR tool for registering the MR steps at patient level was demonstrated; this registration was used for measuring the success rate of complete MRs. During the training, specific attention was paid to improving communication skills; these helped pharmacists when dealing with patients, e.g., for gathering information about their drug use, co-morbidities, and individual problems associated with their medication regimens. Training about communicating with GPs focused on practical advice to motivate GPs to cooperate with completing MRs, especially when sharing information to improve pharmaceutical care plans. Furthermore, pharmacists had the opportunity to send MR test cases from their patients to specialized trainers, together with their recommendations, in order to receive critical comments and feedback about their work. Web-based MR tool The MR tool was accessed via the internet on the SFKwebsite using a specific login code for each pharmacy. The tool was used in two ways: first, to identify the presence of PIMs and measure these on a monthly basis; second, to register MR activities and the time spent per patient (in order to measure the implementation success rate). SFK data included detailed information about the drugs dispensed, e.g., the codes from the ATC system [17], and the prescribed daily dose. Patients were identified using an anonymous code, gender and year of birth [18]. Specifically developed algorithms were used to identify nine PIMs from the HARM study from the dispensing data, as per Supplementary Table 2: three PIMs focused on the underuse of protective agents (e.g. NSAID without gastrointestinal protection, opioids without laxatives, corticosteroids without bisphosphonates); three PIMs addressed overuse (proton pump inhibitors continuation after cessation of NSAIDs, long-term use of benzodiazepines, bisphosphonate use after cessation of corticosteroids); and three PIMs reported sub-optimal drug use (glibenclamide
use in elderly patients, concomitant use of three drugs that could elevate serum potassium levels, NSAIDs together with ‘RAS inhibitors’). The presence of PIMs was reported for each patient within the MR tool, together with background information from relevant guidelines and literature. For NSAID use in patients older than 70 years for instance, the Dutch guideline was cited for co-administration of proton pump inhibitors or misoprostol. The pattern of drug use during the previous three years was plotted graphically for each patient as ‘‘Medication Analysis Profiles’’; these only included prescribed medicines and not OTC drugs because the latter were not registered in Dutch community pharmacies. Reimbursement of MRs Pharmacists who completed MRs for at least 20 eligible patients who were registered in the MR tool during study period received a fee of 45 Euro per patient, as well as reimbursement for the use of the MR tool. Reimbursement was offered to community pharmacists by a nationwide health insurance company, Achmea, for MRs that were performed for their clients. Data analysis Pharmacy characteristics were obtained from information in national quality criteria [19]. The MR implementation success rate was calculated as the percentage of patients with completed MRs out of all the selected patients in the IG. Information about the MR steps used was derived from pharmacist activity registrations and time spent with patients recorded in the MR tool. The mean numbers of PIMs in the IG and RG were calculated at study commencement and completion; changes in the presence of PIMs from study start to follow-up were compared at follow-up using a logistic regression model, which adjusted for patients’ age, sex, and number of drugs used chronically at the point of recruitment into the study. P-values below 0.05 were taken as significant. Data were analysed with SPSS 19.0 (IBM Corp., Chicago, USA).
Results A total of 278 pharmacies volunteered to participate in the project (Fig. 1). They scored higher on all aspects of their professional quality profile compared with Dutch community pharmacy benchmarks (Table 1). Pharmacists from 116 pharmacies attended the project’s training course; pharmacists from 56 pharmacies (48 %) registered MR steps, compared with 46 of 162 pharmacies (28 %) that did not participate in the training course
123
Author's personal copy Int J Clin Pharm
Fig. 1 Inclusion of pharmacies and patients
Table 1 National quality profile aspects of participating pharmacies compared with all community pharmacies
a
Measured at first edition of the quality profile in April 2011
Pharmacies volunteering for MR N = 278 (%) Participation in out-of-hours services
274 (99.3)
1,728 (93.0)
Participation in reporting of drug-related incidents to a national databasea
212 (76.8)
1,119 (60.2)
Participation in pharmacotherapy meetings
270 (97.8)
1,613 (86.8)
Availability of a certificated quality management system
254 (92.0)
1,345 (72.4)
Completion of a patient satisfaction study within the previous year
260 (94.2)
1,460 (78.6)
Pharmacists’ registration for having followed regular trainings in the previous year Consultancy room present
269 (97.5)
1,667 (89.7)
271 (98.2)
1,700 (91.5)
(Table 2). The group of trained pharmacists performed all five MR steps for 660 patients (63 %) from the initial group of 1,052 patients, compared with 312 patients (51 %) from the group of pharmacies with pharmacists who had received no training. In the IG of 660 patients with a completed MR from the pharmacies with trained pharmacists, 58 % were women with a median age of 77 years, 26 % of whom had at least one PIM identified at the start of the study (Table 3). These characteristics did not differ markedly from the 26,584 patients in the RG. In the IG, each patient had a mean of 8.5 chronic drugs, whereas patients in the RG were each taking a mean of 7.9 drugs on a long-term basis. Compared with RG patients, IG patients used more drugs of ATC class A (IG, 90.5 %; RG, 79.3 %) and ATC class B (IG, 82 %; RG, 68.9 %).
123
All Dutch community pharmacies, N = 1,858 (%)
On average, IG patients from the 56 pharmacies that employed trained pharmacists had a mean number of 0.31 PIMs at baseline compared with a mean number of PIMs 0.29 in RG patients. At study end, the mean number of PIMs decreased in IG patients by 0.06 (19 %), and by 0.07 (24 %) in RG patients. Between-group differences were not significant (p = 0.56). Table 4 includes details about the number of patients at the beginning of the study (from both groups) for each of the nine PIMs. The relative percentages of patients did not differ between the groups per PIM, except for users of glibenclamide older than 70 years: 0.8 % in the IG compared with 0.3 % in the RG (p = 0.03). p values for the other PIMs between the groups were above 0.05 and are not shown. Furthermore, for every PIM detected earlier in the study the number of patients for whom this PIM was
Author's personal copy Int J Clin Pharm Table 2 Implementation success rate for completed medication reviews
Pharmaceutical patient history
Median minutes spent per MR step (range standard deviation)
With training course: number of patients per MR step (% of total patients N = 1,052, selected by 56 pharmacies)
Without training course: number of patients per MR step (% of total patients N = 609, selected by 46 pharmacies)
20 (1–120; 17.0)
984 (94)
571 (94)
Pharmacotherapy review
15 (2–300; 15.9)
989 (94)
576 (95)
Definition pharmaceutical care plan
15 (1–195; 13.0)
960 (91)
477 (78)
Implementation pharmaceutical care plan
10 (1–85; 7.5)
854 (84)
390 (64)
8 (1–60; 6.9)
695 (66)
322 (53)
660 (63)
312 (51)
Evaluation of the pharmaceutical care plan Complete MR
Table 3 Characteristics of patients in the intervention (IG) and reference group (RG) Patients form IG (% of N = 660)
Patients from RG (% of N = 26,584)
p value for differences between groups#
Female sex
383 (58.0)
16,201 (60.9)
Median age
77, range 65–96
78, range 65–105
Median number of chronic medication
8.5, range 5–19
7.9 (5–19)
