Article Text

Download PDFPDF

Prescription of potentially inappropriate drugs for geriatric patients in long-term care: improvement through pharmacist's intervention
  1. Conxita Mestres1,
  2. Anna Agustí2,
  3. Laura Puerta3,
  4. Montse Barba4
  1. 1Department of Pharmacy, Grup Mutuam, Barcelona, Spain
  2. 2Department of Pharmacy, HSS Mutuam Girona, Girona, Spain
  3. 3Department of Pharmacy, HSS Mutuam Güell, Barcelona, Spain
  4. 4Department of Pharmacy, Primary Care, Barcelona, Spain
  1. Correspondence to Conxita Mestres, Pharmacy Director, Grup Mutuam, Mare de Deu de la Salut 49, Barcelona 08024, Spain; conxita.mestres{at}


Objectives To determine the prevalence of potentially inappropriate prescribing in our long-term care institutions with limited experience of pharmacists’ interventions. To apply preventative measures, based on pharmacists’ recommendations, to obtain better outcomes for our patients.

Methods Patients were aged >75 years. The study consisted of a retrospective assessment of the prescription of these drugs (July–December 2010: 600 patients), an educational and informative programme to physicians (2011), followed by application of pharmacists’ recommendations (March–November 2012: 1048 patients), based on Beers criteria.

Results In the retrospective period, at least one potentially inappropriate drug was prescribed for 19% of patients. Patients receiving temporary care for long-term disease (temporary long-term care) and continuing long-term care (LTC) were prescribed a higher proportion of potentially inappropriate drugs (46.5% and 36.0%, respectively). During the intervention period inappropriate drug prescription dropped to 14.5%. Physicians ‘acceptance of the recommendations made by pharmacists’ was 45.5%. Acceptance was lower in temporary LTC and subacute care. Fourteen different potentially inappropriate drugs were prescribed in both periods. Amiodarone and hydroxyzine are among the drugs which are more difficult to change.

Conclusions Revision of treatments by pharmacists and their subsequent recommendations to physicians improves the quality of treatments. Physicians are now more aware of the importance of pharmacists’ interventions and a new stage of wider interventions and recommendations has begun. An algorithm has been established that standardises pharmacist interventions and leads to continuing improvement.


Statistics from

Request Permissions

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.


Inappropriate prescribing is recognised as contributing to adverse drug-related events, a worsening of patients’ conditions and poor outcome.1 This leads to an increased use of the healthcare services and increased costs.2 In elderly people the risk is greater because they are more vulnerable, have a high prevalence of chronic illness and are often receiving a great number of drugs.

In recent years, with an increase in the older population, there has been increasing an interest in examining these problems. Different strategies have been developed to evaluate prescriptions in order to detect potentially inappropriate drugs and to reduce polymedication.

Abundant studies have been published that describe diverse methods with this aim—for example studies examining computerised alerts, medication reviews, pharmacists’ interventions, education of patients, etc.1 Moreover, different tools have been used to measure the impact of pharmacists’ interventions—namely, the Medication Appropriateness Index (MAI), improvement in the Assessing Care of Vulnerable Elders questionnaire (ACOVE) or Beers criteria.3–6 All show different results and degrees of success, such as reduction in the number of drugs administered, in the number of patients taking psychoactive drugs, or reduced drug costs and hospital visits, among others. However, it seems that some methods, such as educational interventions, computerised support systems, pharmacists’ interventions, regulatory policies and multidisciplinary teamwork, are more successful.1

The importance of pharmacists’ work in this area is well established. Several systematic reviews have shown that drug reviews, collaboration with other healthcare professionals and education (nurses, clinicians, physiotherapists, dieticians, etc) performed by pharmacists lead to more appropriate drug use and improved outcomes in patients.7

In our work in long-term care institutions, almost all those receiving care are elderly polymedicated people (around 85% are aged >65 years and 75% are aged >75 years), making it of utmost importance to provide appropriate drug treatment. However, we are faced with many problems. First, the drug treatments of a great number of patients require assessment, requiring a great amount of work. Additionally, we do not have a large number of pharmacists and the available information technology is not optimal.

Therefore, in view of these limitations, when we decided to examine the optimisation of drug treatments in our patients we adopted a simple approach in accordance with our technological and staffing levels. We chose to do pharmaceutical interventions, based on Beers criteria in collaboration with physicians. In this paper we explain our experience using this plan and the changes which resulted.


Group Mutuam provides healthcare to different types of geriatric institutions: primary care, nursing homes and long-term care (LTC). Our study was conducted in two long-term care institutions: HSS Mutuam Güell in Barcelona (165 beds) and HSS Mutuam Girona in Girona (97 beds). The patients included in the study are those admitted to the units of temporary LTC (rehabilitation from hospital stay, or surgery), terminal medical condition, continuing LTC (dementia, permanent disability and need for supervision) and subacute care (SC).

The study comprised two stages and patients in the both periods of the study were aged >75 years. First, a retrospective study was carried out to determine the prevalence of inappropriate drug prescription. This was followed by a prospective study to evaluate whether the recommendations of pharmacists would lead to improvements in prescriptions. Identification of potentially inappropriate drugs was based on Beers criteria; a set of explicit criteria for determining inappropriate medication use in elderly.5 ,6 These criteria, based on consensus opinion of experts in geriatric pharmacology, comprise a list of drugs or drug classes and dosages that are known to cause harmful effects in geriatric patients. We focused on 36 drugs on the Beers criteria list; we excluded drugs not available in Spain and drugs whose long-term use could not be tracked (eg, non-steroidal anti-inflammatory drugs such as diclofenac or ibuprofen) (table 1). A total of 600 patients of both institutions were included in the retrospective study, conducted during the second half of 2010. Data of a potentially inappropriate drug prescribed to these patients were obtained from the electronic clinical records and prescription forms (Aegerus V.2.8.6).

Table 1

Inappropriate drugs included in the study

During 2011, we implemented, led by the pharmacy service, a programme to improve the prescription of inappropriate medication according to Beers criteria. The programme included education and explanation of the aim of our project to physicians, and an informative fact sheet was also distributed. From March to November 2012, we performed weekly interventions (owing to staff shortage we had to reduce checks to only once a week). Pharmacists carried out searches, using data extraction from our electronic prescription software, to determine which patients had potentially inappropriate drug prescriptions. When inappropriate prescriptions were found the pharmacists reviewed the complete pharmacological treatment and related documents (clinical record, home treatment, etc) and sent their recommendations/suggested interventions to the physician (via encrypted email). They used a template including patients’ identification, drug(s) considered inappropriate and their recommendations, which were usually instructions to physicians about how to stop the drug or change it for another more appropriate. Physicians could answer using the template via email, but oral answers by phone or direct meeting were also accepted. Pharmacists recorded all the recommendations/interventions and physicians’ replies using a Microsoft Excel 2010 template, from which results were extracted later.


In the retrospective study we evaluated potentially inappropriate prescriptions of 600 patients and in the intervention phase 1048 patients, whose characteristics are described in table 2. Both groups are of similar age and similar male/female proportion, and the patients are classified similarly.

Table 2

Characteristics of the study population

In the retrospective study 114 patients (19%) in our institution were found to have received at least one potentially inappropriate drug. Table 3 shows the data for patients of each hospital unit. For patients receiving temporary and continuing LTC potentially inappropriate drugs were frequently prescribed: 65% of women and 35% of men.

Table 3

Characteristics of the patients (according to their hospital ward), with potentially inappropriate medication in the retrospective study

Drugs potentially inappropriately prescribed are shown in table 4. Amiodarone, fluoxetine, hydroxyzine and diazepam were prescribed with high frequency.

Table 4

Potentially inappropriate drugs prescribed during the study periods*

In March 2012, we started the prospective study; pharmacists made weekly checks of potentially inappropriate drugs, prescribed by physicians in the two institutions. During the intervention programme, 145 patients received a potentially inappropriate drug (14.5%). We made interventions and recommendations to clinicians to change or discontinue a drug in 99 patients (65.1%) and these were accepted in 45 cases (degree of acceptance 45.5%). The reasons for non-acceptance were the drug had been prescribed by a specialist and the patient's disease is controlled 69%, antibiotic treatment with few alternatives 4%, no change but the dose was decreased 4% and no answer was received for 23%. In table 5 we show the patients with potentially inappropriate drugs and the degree of acceptance of pharmacists’ recommendations by physicians depending on the ward in which they were hospitalised.

Table 5

Characteristics of the patients with potentially inappropriate medication in the intervention study, recommendations made and degree of acceptance

The rates of accepted recommendations were 23% for diazepam, 23% for clonazepam, 13% for hydroxyzine, 10% for clorazepate, 6% for pentoxifylline and oxybutynin and 3% for fluoxetine, flunitrazepam and amitriptyline.

In tables 3 and 5, we can see that even though that there was a decrease (46.5 vs 40.1%) in inappropriate medication in patients receiving temporary LTC, those patients continued to receive habitually more of these drugs. In continuing LTC there is a slight drop but for patients in a terminal medical condition the potential for inappropriate drugs increases; for SC we detected an increase, but due to the low number of patients in this group, we do not find it significant. The inappropriate drugs prescribed are almost the same; there is a drop in the prescription of fluoxetine, diazepam and nitrofurantoin. Meperidine, clomipramine, cyproheptadine and pentoxifylline were no longer used.

In this study we have focused on potentially inappropriate drugs, therefore it was not surprising to find that the mean number of drugs per patient did not change in the two periods (8.3±4.07 in the retrospective period vs 8.8±3.6 in the prospective period).


The application of a programme from the pharmacy service, based on Beers criteria, combined with education of health professionals and interventions by pharmacists led to a decrease in the potentially inappropriate drugs prescribed to our patients (19% vs 14.5%). Additionally, in the intervention period only one patient received more than one potentially inappropriate drug in comparison with 11 patients in the retrospective study. The initial value in 2010 (19%) is similar to that found by other authors in nursing homes and communities (18.5% and 21.3%).8 ,9 However, there is a wide range of prevalence (from 41.1% to 5.8%) and even values of 53.4% in different European countries.10 ,11

We think that our improved results were due to a combination of the pharmacists’ interventions and education.

The low degree of acceptance of the pharmacists’ recommendations/interventions (only 45.5%) may be explained by the fact that this was the first attempt by the pharmacy service in our institutions to undertake a programme involving interventions. Physicians accepted fewer recommendations for patients in units of shorter stay (temporary LTC and subacute), where they will have less time to assess the effects of a change of drug in the patient before discharge, or when the drug was prescribed by a physician specialist. The difficulty in altering prescribing practice depended on the medication. Of the potentially inappropriate drugs, amiodarone, hydroxyzine, clonazepam and nitrofurantoin stand out as drugs difficult to change. The case of amiodarone is particularly important. The use of antiarrhythmic drugs in older people is complex, because there is no drug that can be recommended, or which has a better profile. We discussed this problem with the physicians, and concluded that we will no longer include it in our list of potentially inappropriate drugs, but follow recommendations published in the directive to evaluate the antiarrhythmic treatment in a patient.12 ,13 Lack of change in the prescribing practice for clonazepam and nitrofurantoin had logical explanations. In the first case clonazepam was used for treatment of chronic disease and the patients were controlled for different neurological conditions. Nitrofurantoin was prescribed because the antibiogram results for urinary infections suggested only parenteral antimicrobial agents as an alternative. More difficult to understand is the lack of change in the prescription of hydroxyzine; the physicians claim that the anxiolytic effect on pruritus with this drug are the reason for not changing it. We will examine this problem in order to find acceptable alternatives for physicians.

In summary, physicians do not change prescriptions for the above drugs for the following reasons: no better alternative for amiodarone, sedative effect of hydroxyzine, patient is controlled with clonazepam and there is no better alternative than nitrofurantoin for treatment of urinary infection.

At the end of the study, we presented the results to all the healthcare professionals. We think that this continuing programme has made them more aware of this problem; from the perspective of the pharmacy service, we found that our physicians were more prepared to collaborate in drug review, leading us to start a more global approach. This has consisted in the development of a structured method (algorithm) of reviewing prescriptions in order to detect medication-related problems, based on the characteristics or our patients, number of drugs, diagnostics, medication appropriateness, interactions and contraindications, etc.

During 2013, after discontinuation of the intensive programme that focused on Beers criteria, but with all the other measures, prescription of potentially inappropriate drugs continued to drop, reaching 13.4% of the patients. This shows that, even though this study was limited to a small number of drugs, it has allowed us to initiate more complex activities in pharmaceutical care, such as assessment of medication at the patient admission and more extended drug treatment reviews. These improvements are due to greater involvement of physicians with the pharmacist work.

Key messages

  • What is already known on this subject

  • Potentially inappropriate drug prescribing in older patients is common.

  • Different strategies have been used to identify and prevent potentially inappropriate prescribing in older people, resulting in different outcomes for patients.

  • The role of the pharmacist is paramount in optimising pharmacotherapy in older people.

  • What this study adds

  • How to improve potentially inappropriate drug prescribing in long-term care institutions with limited resources and limited experience of pharmacists’ interventions.


We acknowledge the help of Mrs Núria Galedano, Mrs Salud Gonzalez and Mr Xavier Trelis in the extraction of data.



  • Contributors Conceived idea: MB, LP and AA. Data collection: LP and AA. Analysis of data, discussion and results: CM, LP and AA. Writing of the paper and preparation for publication: CM.

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.