Introduction Patient-specific medication reviews might be useful for correcting prescribing omissions. The Dutch Healthcare Inspectorate (IGZ) demands that a medication review is performed by a pharmacist in cooperation with a physician for all residents of nursing homes (twice a year) and residential homes (once a year).
Objective This study aims to show the clinical practice of medication reviews for older people in residential and nursing homes by pharmacists. We asked the following research questions. (i) To what extent are medication reviews performed as required by the IGZ? (ii) How much time is needed to perform the required medication reviews? (iii) In the opinion of the pharmacist, is the medication review economically efficient and medically relevant?
Methods A web-based survey was sent to 87 hospital pharmacists and 270 community pharmacists.
Results In a ‘best case vs worst case’ scenario, 76% vs 42% of the IGZ-required medication reviews are performed. Considering a mean of 730.7 medication reviews required per year and the mean time spent on one medication review (29.3 min), a pharmacist would need to spend 2.5 months a year to meet the IGZ requirements. Almost every pharmacist considers medication reviews to be medically relevant, but only 47% consider them to be economically efficient.
Conclusions This survey shows that medication reviews performed in institutional care settings do not meet the IGZ requirements, which is not surprising considering the time needed for one medication review. Automation of medication review processes should be initiated to increase efficiency.
- PHARMACY MANAGEMENT (PERSONNEL)
- INDIVIDUALISED MEDICATION SURVEILLANCE
- CLINICAL PHARMACY
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Elderly patients often have multiple disorders and receive a variety of drugs. Polypharmacy is defined as the concurrent use of multiple drugs. Depending on the definition, this could be more than four, five or nine drugs.1 The Dutch Institute for Public Health and the Environment reported that 30–45% of the elderly (defined as 65 years or older) use five or more different drugs daily. This percentage is reported to be even higher for patients in residential homes and nursing homes.2
Polypharmacy and multimorbidity are strongly related. Polypharmacy increases the risk of adverse effects, while adverse effects induce prescribing cascades. The prescribing cascade is the misinterpretation of a drug side effect for a new medical condition, which leads to the prescribing of a new drug.3 In contrast, polypharmacy may induce suboptimal treatment because the probability of underprescription paradoxically increases with the number of drugs used.4 Other consequences of polypharmacy are an increased incidence of drug–drug interactions, drug–disease interactions, non-adherence, and the likelihood of inappropriate prescription.5 ,6 Using the Beers’ Criteria, potentially inappropriate medication is found with a high prevalence of 12% and 40% in community-dwelling elderly and nursing home residents, respectively.7 A recent study showed potentially inappropriate prescribing in 60% and potential prescribing omissions in 42% of the elderly.8 Inappropriate prescribing increases the incidence of related hospital admissions and should be prevented.9 Patient-specific medication reviews could be useful to correct inappropriate prescribing or prescribing omissions.
A medication review is defined as a structured evaluation of the medication of a specific patient, aimed at reaching agreement with the patient, optimising the impact of medicines, and minimising the number of medication-related problems while considering medical history and laboratory values.10
A regular revision of medication increases medication appropriateness while reducing medication-related problems. Although effects on primary end points such as hospital admissions have not been proven, it is likely that the frail population of nursing homes and residential homes will benefit from medication reviews.11
Nursing homes mainly have older residents with chronic somatic diseases or progressive dementia who are not able to perform activities of daily living. A nursing home physician, an officially acknowledged medical discipline, offers medical care to these residents.12 Elderly residents of residential homes are usually less dependent on care. A general practitioner treats elderly residents of these homes.
The Dutch Healthcare Inspectorate (IGZ) demands a medication review by pharmacists and physicians for all residents of nursing homes (twice a year) and residential homes (once a year). In this review process, the information given by the nursing staff and the patient him/herself should also be taken into account.13 It is unclear to what extent the guidelines of the IGZ regarding the performance of medication reviews are followed in daily practice. This study aims to show clinical practice of medication reviews for older people in residential and nursing homes by pharmacists. Our aim is to answer the following research questions. (i) To what extent are medication reviews performed as required by the IGZ? (ii) How much time is needed to perform the required medication reviews? (iii) Is the medication review economically efficient and medically relevant in the opinion of the pharmacist?
A self-administered web-based survey was developed by three hospital pharmacists and a community pharmacist, who all specialise in performing medication reviews.
The survey was piloted on three community pharmacists and two hospital pharmacists, and was adjusted according to their feedback. The data from the pilot study were not included in the present study. The questions were related to (i) pharmacist/pharmacy characteristics, (ii) cooperation between physicians and pharmacists, and (iii) medication reviews. The survey consisted of 50 questions: 35 multiple choice, 9 numerical, 3 rating scale, and 3 open-ended.
The survey was sent to a total of 357 pharmacies: 87 hospital pharmacies and 270 community pharmacies. The community pharmacies consisted of 30 independent pharmacies and 240 pharmacies belonging to the Mediq-group, which is the market leader of chain pharmacies in the Netherlands. The hospital pharmacies and 30 independent community pharmacies were approached personally, while the chain pharmacies were approached with the aid of the Mediq-group. The survey was closed to answers from pharmacists 6 months after the first invitation had been sent. We received 68 completed surveys. For each pharmacy, only one pharmacist filled in the survey, thus every pharmacist represents one pharmacy organisation. All answers were reported anonymously and could not be linked to a pharmacy. Numerical variables were presented as mean (SD), and n (%) was used for categorical variables. Although some variables were expected to be positively skewed, means (SD) were used instead of medians (interquartile range) because the mean is more informative in this situation. Descriptive statistics were computed using Microsoft excel 2010.
The medication review can be separated into at least three and as many as six phases: (1) gathering patients’ information; (2) pharmacotherapeutic anamnesis; (3) pharmacotherapeutic analysis; (4) creation and determination of the pharmaceutical care plan; (5) determination of the pharmaceutical care plan with the patient; (6) the follow-up.14 ,15 For the purpose of this survey, we translated these phases into three steps: preparing (phases 1–3), discussing (phase 4) and finishing (phases 5 and 6) a medication review. The survey also examined the practice of Dutch pharmacists of performing medication surveillance and the use of a clinical decision support system (CDSS). A system that can provide algorithm-based medication safety alerts electronically is referred to as a CDSS.16 In regular practice, pharmacy information systems use the ‘G-standard’ to generate medication safety alerts; the G-standard is a nationwide drug database issued by the Royal Dutch Association for the Advancement of Pharmacy (KNMP). These medication safety alerts are based on drug–drug and drug–disease interactions which are required to be checked by a pharmacist daily.17
A newer method for generating additional medication safety alerts involves clinical rules. Clinical rules are algorithms using multiple sources of information such as laboratory values and medication to generate medication safety alerts. If electronic, the clinical rules can be used in a separate monitoring system, but also integrated into the pharmacy information system.16 ,18
The survey was filled in by 68 pharmacists, a response rate of 19%. Of the respondents, 60% were practising in a community pharmacy and 40% in a hospital pharmacy.
Characterisation of the pharmacists
Of the 68 pharmacists, 88.2% offered pharmaceutical services to one type or both types of institution: nursing homes and residential homes. The other pharmacists did not offer pharmaceutical services to these institutions. Table 1 shows that pharmacists offered pharmaceutical services to a mean of 1.0 (SD 1.6) residential home and 2.2 (SD 3.3) nursing homes. The mean number of patients for the pharmacists that offered services was 151.1 (SD 134.0) in residential homes and 555.0 (SD 514.6) in nursing homes. Medication surveillance using the G-standard was used by 98.5% of the pharmacists.
Cooperation between physicians and pharmacists
Table 2 shows that 90% of the pharmacists have recurrent consultations with physicians. In 90% of the reported cases, these consultations were pharmacotherapy audit meetings (not shown in table 2), which usually occurred monthly (44%) or quarterly (45%). Pharmacotherapy audit meetings occur between pharmacists and physicians (general practitioners or nursing home physicians) to come to an agreement on pharmacotherapy, taking into account national guidelines and the costs of medicines.19 Access to patient information in regular practice (ie, the indications related to the prescribed drugs and laboratory values) varied greatly between hospital and community pharmacists. Most of the hospital pharmacists (78%) had access to laboratory values compared with 29% of the community pharmacists. The hospital pharmacists with routine access to laboratory values reported that 86% had access to all values and 14% to a limited set of laboratory values. A limited set could be renal function only, or also liver function, albumin, electrolytes, cholesterol or the international normalised ratio. Of the community pharmacists with routine access to laboratory values, 25% had access to all values and 75% to a limited set, often restricted to kidney function only (67%). Routine access to the indications for the prescribed drugs was available in 37% of the hospital pharmacists and 7% of the community pharmacists.
Table 3 shows that most hospital (82%) as well as community (88%) pharmacists performed medication reviews (85% of all pharmacists). Access to laboratory values in regular practice was different from when performing medication reviews; 17% of pharmacists acquired access to laboratory values when performing medication reviews, and thus 36% of the pharmacists remained without access. When differentiated, 18% of hospital pharmacists and 47% of community pharmacists still lacked access to laboratory values when performing medication reviews. When pharmacists perform a medication review, 67% reported having access to the indication related to the prescribed drugs.
The method most frequently used to support medication reviews involved the use of the pharmacy information system. This was used by 86% of the pharmacists (data not in table). The use of clinical rules as a support technique for medication reviews was reported by 33% of the pharmacists. The mean estimated time spent on a medication review for one patient was 29.3 min (SD 22.0). There was a difference between the mean time spent by community pharmacists (36.6 (SD 24.2)) and hospital pharmacists (20.2 (SD 15.0)). Preparing and finishing the medication review is the main reason for the difference between community and hospital pharmacist. Community pharmacists spent a mean time of 17.7 (SD 11.8) and 15.7 (SD 11.2) min preparing and finishing the medication review, respectively, while hospital pharmacists only spent 9.4 (SD 10.1) and 5.7 (SD 4.8) min.
Eighty-eight per cent of the pharmacists offered services to a total of 27 488 patients in any of the above-mentioned institutions, resulting in a mean of 404.2 (SD 506.0) patients for all pharmacists who completed the survey. Eighty-five per cent of all the pharmacists reported performing medication reviews. However, only 89.7% of the pharmacists reported reviewing a total of 18 824 patients yearly. The remaining 10.3% did not report the number of medication reviews performed. Therefore, the estimated number of medication reviews would be 20 996 with a mean of 308.8 (SD 635.3) yearly reviewed patients for all the pharmacists claiming to perform medication reviews.
According to 98% of the pharmacists, medication reviews are medically relevant. The mean estimated number of interventions suggested by pharmacists in a medication review was 2.0 (SD 1.2), whereas the estimated mean number of interventions accepted by physicians was 1.2 (SD 0.9). Of all the pharmacists, 47% considered medication reviews to be economically efficient, while 20% were compensated for performing medication reviews. Almost all pharmacists (98%) believed that the pharmacist is the right person to perform medication reviews. Almost 13% of the pharmacists thought that an electronic system could replace the medication review.
This survey reports on 60 pharmacies that provided care to 27 488 patients in nursing homes and homes for the elderly. In 2009, approximately 155 000 patients were cared for in such institutions in the Netherlands.20 This survey gives insight into the clinical practice of medication reviews in institutional care settings by pharmacists. To what extent medication reviews are performed according to IGZ requirements can be calculated by taking into account that the IGZ requires patients in residential homes to be reviewed yearly and those in nursing homes to be reviewed twice a year. In this survey, 60 pharmacists offered services to a mean of 88.2 (in total 5290) residential home patients and a mean of 370.0 (in total 22 198) nursing home patients. This results in a total of 49 688 yearly IGZ-required medication reviews with a mean of 730.7 per pharmacy. In contrast, the reported annual number of reviewed patients per pharmacy was a mean of 308.8. This suggests that only 42% of the IGZ-required medication reviews are performed.
However, the survey reported the number of annually reviewed patients per pharmacy and not the number of medication reviews, meaning that these patients are not necessarily only reviewed once a year. Therefore the calculated 42% IGZ-required medication reviews performed is a ‘worst case’ scenario. A comparison of the reported number of patients per pharmacy in a residential or nursing home (404.3 ) with the reported mean number of patients reviewed (308.8) shows the ‘best case’ scenario, which is that 76% of IGZ-required medication reviews are performed.
Although these two percentages vary, it demonstrates that medication reviews are not performed according to IGZ requirements. However, it should be considered that pharmacists select high-risk patients—that is, patients with more than five different medications. Patient selection using a minimum number of medications was one of the inclusion criteria in the Preventing Hospital Admissions by Reviewing Medication (PHARM) Study.14
Time needed to perform medication reviews
The time needed to perform the required yearly medication reviews can be calculated using the mean pharmacy size and the amount of time needed by one pharmacist to perform these medication reviews. The 730.7 annually required medication reviews, multiplied by the reported 29.3 min, would take a pharmacist 2.5 months of full-time work a year. A recent report published by the KNMP determined that a regular medication review for patients in an institution such as a nursing home would require 68 min.15 Taking this into account, one pharmacist would require nearly 6 months of full-time work annually to perform the IGZ-required medication reviews. The difference between the KNMP report and our results should be interpreted with caution and can perhaps be attributed to our approach of simplifying the medication review to three phases. Nevertheless, these results do show that performing IGZ-required medication reviews is very time-consuming, and the amount of time spent by the physician is not even included in this calculation. It is questionable if this amount of time spent on manually performed medication reviews is achievable and acceptable.
The impact on a pharmacy organisation with regard to time needed to perform the required medication reviews will also depend on the number of pharmacists in the organisation. Taking into account the mean number of community and hospital pharmacists (1.8 vs 7.9), the IGZ requirements would have a larger impact on a community pharmacy than on a hospital pharmacy.
Relevance of the medication review
Pharmacists report that medication reviews are medically relevant, meaning that they are thought to improve the quality of pharmacotherapy. The results show a mean of 2.0 proposed interventions and 1.2 accepted interventions for every medication review. This results in a mean rate of acceptance by physicians of 60%. Other known physicians’ acceptance rates vary from 39% to 91.6%. It has been suggested that physicians’ acceptance rate can be correlated with the ‘cooperation between physicians and pharmacists’, which seemed to be good (90%) in our results.21 The literature remains ambiguous concerning the effects of manually performed medication reviews in institutions. Medication reviews in nursing home patients do not appear to reduce hospitalisation or mortality, although none of the studies included had the power to detect differences. Another study in care homes also did not show any effect on hospitalisation, mortality and adverse drug events, but did show an increase in medication appropriateness.11
Automation of the medication review process
In the KNMP report, the manual acquisition of patient information is an important time factor in the medication reviewing process.15 In our survey, community pharmacists needed to acquire laboratory values more often than hospital pharmacists. This might be the reason why the time spent preparing a medication review was longer for community pharmacists than for hospital pharmacists.
One-third of the pharmacists in our study were unable to acquire laboratory data and the indication for the prescribed drugs when performing medication reviews. It has been shown that the absence of laboratory values and indications can influence the outcome of the medication reviews.22 Fortunately, the limited access to indications and laboratory values has recently changed. A new Dutch law states that the reason for prescribing 39 medicines should be stated on the medication order, and, when requested, laboratory values should be given.23 This information is usually retrieved manually, or a ‘viewer’ application is used.15 Automated retrieval of patient information has not yet been incorporated into daily practice. It seems plausible that, when access to indications and laboratory values is automated, the time needed to perform a medication review will decrease.
A recent study has proven the low efficiency of manually performed medication reviews performed by physicians or pharmacists compared with an expert group, suggesting the evident need for a standardised system able to support medication reviews.24 The literature has proven the benefits of CDSS showing improved patient-important outcomes (eg, adverse drug events) and positive effects on preventive care reminder systems.25 ,26
The use of patient information in a CDSS designed to replace the medication review would also require automated access to patient information. This survey reports the scepticism of almost 90% of the pharmacists about full automation of medication reviews. There is a lack of consensus in the literature also regarding the possibility of a fully automated medication review.27 Besides the necessary laboratory values and indication for the prescribed medication, there is no possibility to electronically collect the required information from the patient or nursing staff. This might be the reason for the pharmacists’ scepticism about fully automated medication reviews. Information from the patient or nursing staff is still collected in meetings. A solution to this problem might be the use of self-reported information about an individual's symptoms and perceptions (patient-reported outcomes). Currently, the use of patient-reported outcomes is under investigation for clinical assessments and may be used in the future to facilitate medication reviews.28
A limitation of this survey was the low response rate of 19% (68 pharmacists). The low response rate may be the result of the survey-requesting letter. We encouraged pharmacists to complete the survey in relation to nursing and residential homes and the performance of medication reviews. It is possible that many pharmacists who do not perform medication reviews or offer pharmaceutical services to nursing homes or residential homes did not complete the survey.
In this survey, we asked the opinions and estimates of the pharmacist for the corresponding pharmacy rather than measuring it. Inherent to such a survey, the answers can vary somewhat, and specific results should be interpreted cautiously. Also, this survey was merely piloted on five pharmacists but not validated. Nevertheless, we believe it has provided the correct information to answer our research questions.
This quantitative, explorative survey shows that medication reviews in institutional care settings for older people are not performed as required by the IGZ. We demonstrate that the required medication reviews demand large amounts of time; however, pharmacists do believe that medication reviews are medically relevant. To be able to perform all the IGZ-required medication reviews, more efficient methods are required. A CDSS is able to support medication reviews in a constant and regular way while also reducing the time needed. Automation of the medication review process should be initiated to increase its efficiency.
What is already known on this subject
A medication review is a structured evaluation of the medication of a specific patient which increases medication appropriateness while reducing medication-related problems.
Medication reviews in Dutch institutional care settings are required by the Dutch Healthcare Inspectorate.
What this study adds
This study gives insight into the clinical practice of medication reviews in institutionale care settings for older people in the Netherlands.
Special thanks to the pharmacists who completed the survey.
Correction notice This paper has been amended since it was published Online First. The corresponding author was wrongfully described as a ‘Dr’.
Contributors All authors contributed equally.
Funding SCREEN-study supported by a grant from ZonMw (the Netherlands Organisation for Health Research and Development, Grant number: 113101001).
Competing interests None.
Provenance and peer review Commissioned; externally peer reviewed.
Data sharing statement All data are published in the article. The data are available from the first author.