Drug related problems in admitted geriatric patients: The impact of clinical pharmacist interventions

Abstract

Background

In developing countries including Ethiopia, life expectancy is increasing. This is partly as a result of increased healthcare seeking behavior in the society and increased access to health service [1][2]. Related to this, population aging has resulted in an increased prevalence of chronic diseases and thus rise in hospitalizations and healthcare costs of older adults[3].

The geriatric population is at high risk of drug related problems (DRPs) due to the age-related pharmacokinetic and pharmacodynamic changes [4]. Furthermore, a higher incidence of drug related problems could result from age associated increased prevalence of multiple chronic diseases, which causes the use of complex therapeutic regimens [5]. According to pharmaceutical care network of Europe (PCNE), DRP is defined as, “an event or circumstance involving drug therapy that actually or potentially interferes with desired health outcomes”[6]. DRPs are associated with increased healthcare costs and hospital admissions, prolonged hospital stays, reduced quality of life, and increased mortality [7, 8]. Therefore, drug prescribing and use in older patients needs special considerations including avoidance of inappropriate drugs, rational utilization of indicated medications, side effects monitoring, prevention of drug-drug interactions, and evaluation of adherence and patient involvements [9].

The identification, resolution, and prevention of DRPs have been described as a core process of pharmaceutical care. Clinical pharmacists are suitably trained to carry out medication reviews in geriatric patients and have been found to improve the use of high-risk medications and improve the accuracy of medication regimens [10, 11]. In order to resolve DRPs, the cause should be identified and DRPs should be classified appropriately. For such purpose, the classification of DRPs is crucial. There are several classifications for DRPs. However, there is no single standardized classification in the world [12].  The PCNE classification system is commonly practiced and has better usability and internal consistency as it is updated and revised periodically.  It is very important for the documentation of DRPs in the pharmaceutical care process [13].

Ethiopia is the second from the top six countries, in which life expectancy and the number of geriatric population is increasing [14] however, there is limited attention for these older patients. Most of the health sciences training programs didn’t include gerontology training in their curriculums and there is no guideline that focuses on geriatric medicine. At the hospital level, geriatric wards are not established for these special populations.

As with other health care services, geriatrics care requires health care professional team work including clinical pharmacists. Experience from developed nations has shown involving clinical pharmacist in patient care resulted in a reduction of DRPs as well as associated costs [15, 16].Despite the good start up including patient-oriented pharmacy curriculum, clinical pharmacy service is still at the infant stage in Ethiopia. With poor involvement of clinical pharmacists on geriatric care, there is limited information on magnitude of DRPs, determinant factors and clinical pharmacist’s interventions among geriatric patients in hospital set up. Therefore, this study aimed to identify drug related problems and determinants in geriatric patients admitted to medical and surgical wards, and to evaluate the impact of clinical pharmacist interventions for treatment optimization.

Method

Study design and setting

A prospective interventional study was conducted at medical and surgical wards of Jimma University medical center (JUMC). JUMC is the only teaching and referral hospital with 500 beds in the southwestern part of the country located in Jimma town, Southwest Ethiopia, 352 km far from Addis Ababa.

Study population

All geriatric patients ≥ 60 years who were admitted for at least 24 hours in the medical and surgical wards of JUMC during the study period (01 April 2017 to 31 July 2017) were included. Geriatrics patients were defined as patients age 60 years and older. Although most developed world countries have accepted the chronological age of 65 years as a definition of 'elderly', this does not adapt well to the situation in Africa and it can be extended to 60 years and above [17, 18].  Accordingly, we used the cut point of 60 years old in our study. Geriatric patients re-admitted during the study period and patients who refused to participate in the study were excluded.

Data collection

Patients were interviewed using a standard questionnaire and their respective medical chart was reviewed using data abstraction format. Four MSc clinical pharmacists were involved in the identification, interventions and, documentation of DRPs. DRPs were identified by using the following literature resources such as Ethiopian guidelines, European or American standard guidelines such as, American diabetic association for diabetes, American heart association (AHA) guideline for heart failure, American urologic association guideline for benign prostatic hyperplasia, WHO guideline for surgical site infection prophylaxis and other relevant guidelines for the respective diseases identified. Moreover, adverse drug reactions (ADR) were assessed according to the Naranjoadverse drug reaction probability scale [19]. Micromedex was used to check drug-drug interactions. Pharmaceutical care network Europe (PCNE) DRPs classification system was used to classify and document DRPs[6]. For the identified DRPs, intervention was provided through discussion with individual prescriber immediately. Additionally, recommendations were forwarded during rounds and morning sessions orally and/or with written documents and the prescriber acceptance documented. DRPs which are not accepted were further discussed with a senior physician for confirmation. The acceptance of the recommended interventions was followed to check for their implementations and the status of the intervention was documented.

Outcome measure

We have two main objectives in the study. The first objective is to measure the prevalence and determinates of DRPs among geriatric population attending surgical and internal medicine wards of JUMC. The second objective of the study is measuring the impact of clinical pharmacy intervention on reducing the number of DRPs and the acceptance rate of clinical pharmacists’ intervention by prescribers.

Explanatory variables

DRPs were identified by evaluating the appropriateness of drug therapy in terms of indication, dosage, safety, and efficacy. For the identified DRPs, the clinical pharmacists provided interventions. Clinical pharmacist interventions were defined as any action by a clinical pharmacist that directly results in a change of patient management. The Pharmaceutical care network Europe DRPs classification system [6] used in this study have different parts that include: i)The problems(e.g. effect of drug treatment not optimal, unnecessary drug-treatment), ii) causes (e.g. , drug dose too low, wrong drug administered, inappropriate timing or dosing intervals), iii)planned interventions (e.g. intervention discussed with prescriber, patient (drug) counselling), iv) acceptance of the intervention proposals (intervention accepted or not accepted) and v)outcome of the intervention (problem solved, partially solved or unsolved). Different interventions were applied for some of the DRPs that resulted in the number of interventions greater than the number of DRPs.

The other variables assessed were (1) socio-demographics (including sex, age, financial status, educational level, residence, marital status, and social support) (2) disease related factors (diagnosis, co-morbidities) and (3) drug/therapy related factorsthat include (polypharmacy, class and type of drugs).Poly-pharmacy was considered when the patients were prescribed with five or more drugs together. Patients who had one or more additional disease condition in addition to the main disease were considered as having comorbidity.

Statistical analyses

EpiData version 4.0.2. for data entry and Statistical Package for Social Science (SPSS) version 20 (IBM SPSS version 20.0 Inc., Chicago, Illinois) were used for data analysis. Categorical variables were expressed in terms of frequencies and percentages. Continuous variables were presented using mean and standard deviation. The proportion of DRPs and prescriber’s acceptance rate of the intervention were reported using these descriptive statistics. Univariable logistic regression analysis and multivariable logistic regression analysis was performed to determine the potential determinant of DRPs.  Results were reported as odds ratios (ORs) with 95% CIs.  A p value of less than 0.05 was considered significant.

Results

Socio-demographic characteristics

A total of 200 patients were included in the final analysis. Of this, 135(67.5%) were male, the mean (SD) age was 67.3 (7.5) years, and the majority (68.5%) of the patients were from rural areas.  Two third of the patients did not have a formal education and only 36.5% of the patients had a monthly income of ≥ 800 Ethiopian Birr. Furthermore, 75% of them relied on the caregiver (Table 1).

Clinical and medication characteristics

The majority (65. 5%) of the patients were admitted to the medical ward. Participants had an average of 2.20 (SD1.57) clinical conditions and they took an average of 3.90 (SD2.11). The most common medical condition was heart failure (24%) followed by stroke (13%) and benign prostatic hyperplasia (11%). Polypharmacy was present in 35.5% of the patients (Table 2). 

Prevalence of drug related problems

A total of 380 DRPs were identified from 81.5% of the study participants. Every patient had an average of 1.90 (SD1.47) DRPs. The most commonly found DRPs belonged to the treatment effectiveness related (effect of drug treatment not optimal, untreated indication, and no effect of drug treatment) with 47.6%, followed by (unnecessary drug treatment, problem with the cost-effectiveness of treatment) 28.2%, and adverse drug event 24.2%. Regarding the number of DRPs, 24.5% had one, 26.5% of patients had two, and 30.5% had three and more DRPs (Table 3).  Some examples and description of DRPs were provided in table 4.

Causes of drug related problems

Four hundred sixty six causes of DRPs were identified. Of this, the most common causes of DRPs were inappropriate drug selection (54.1%) followed by inappropriate dose selection (14.6%) and drug use process (12.2%). Among the inappropriate drug selection causes, the presence of new indication for drug treatment (36.1%) was the most common cause followed by no indication for the prescribed drug (20.6%), and inappropriate drug according to guidelines (16.7%) (Table 5).  The most frequently involved class of drugs in DRPs were cardiovascular agents (38.1%) followed by antibiotics (21%), and hematological agents (19.7%) (Figure 1).

Clinical pharmacist interventions

For identified DRPs, a total of 670 interventions were provided at different levels. Most of the interventions (41%) were provided at the prescriber level, followed by 39.1% at drug level, and 16.1% at patient/caregiver level. At the prescriber level, interventions proposed and discussed with the prescriber were the commonest (82.7%) form of intervention.  Out of the interventions carried out at the drug level, the drug stopped and a new drug started was the most common with 30.5% each of them. The prescriber’s acceptance rate was calculated considering the interventions provided at the prescriber level. Accordingly, out of 300 interventions performed at the prescriber level, 275(91.7%) of them were accepted. After the implementation of the interventions, out of the total DRPs, 65.8% of the problems were solved while 27.6% of the problems were not solved (Table 6).

Determinants of drug related problems

In the multivariable logistic regression model, number of disease condition (AOR = 1.588, 95%CI = 1.029-2.450) and polypharmacy (AOR= 3.350, 95% CI= 1.212-9.260) have statically significant association with DRPs.  Patients who took five or more medications (polypharmacy) are 3.350 times more likely to have DRPs than those who took less than five medications. Moreover, as clinical condition increases by one unit the likelihood of developing DRPs increases by 58.8% (Table 7).

Discussion

Drug related problems are becoming a major public health concern. Geriatric patients are particularly highly vulnerable to DRPs caused by multiple factors such as polypharmacy and inappropriate prescribing [20]. Identification and prevention of DRPs occurrence in this population is crucial. In our study, DRPs were present in 81.5% of the geriatric patients. This is in line with a study conducted by Ramanath K et al in internal medicine rural tertiary care hospital of India, which showed an 83.4% prevalence of DRPs [21].

Our study identified an average of 1.90 ± 1.47 DRPs per participant. This is somewhat lower than a study conducted by Chan et al in Thailand, which showed an average of 2.2 ± 1.6 DRPs per participant [22]. This can be explained by the setting difference. Our study was conducted in hospitalized patients where frequents patients visit by senior physicians and residents were applicable whereas the study in Thailand was conducted on outpatients where patients were less likely to meet senior physicians frequently. More than half (54.1%) of the DRPs resulted from inappropriate drug selection. Among the causes of inappropriate drug selection, having a new indication for drug treatment accounted for a higher proportion (36.1%).  This is in line with an interventional study done in Belgium in geriatric wards, which reported the most frequent DRPs was an underuse of medications [23].

After the identification and characterization of DRPs, interventions were proposed by the clinical pharmacists. Interventions were provided at prescriber level, at patient /caregiver level, and at drug level. The physician’s acceptance rate of the clinical pharmacist’s recommendations was determined based on the intervention acceptance rate at the prescriber level. Accordingly, the prescriber acceptance rate was high (91.7%). This is in line with a study conducted in Belgium (87.8%) [23]. This finding has important implications for participating clinical pharmacists as part of the multidisciplinary team, can facilitate the identification of DRPs among geriatric patients and inform physicians to resolve the problems. 

Another purpose of our study was to identify the determinants of DRPS. We found that patients with polypharmacy were more likely to develop DRPs than without polypharmacy. In accordance with the present results, previous studies [23, 24] have demonstrated that polypharmacy was a major determinant for the occurrence of DRPs. The observed association of polypharmacy with DRPs in geriatric patients could be as a result of increased health-care costs for the multiple medications, drug-interactions, disability/cognitive impairment, noncompliance to medications, increased risk of adverse drug events, falls and fractures, malnutrition, and functional status decline [3, 25]. We also found that patients having one or more co-morbidities had more DRPs. This finding is in agreement with Tegegneet al [26], the finding which showed patients with comorbidities were at high risk of DRPS.

After the implementation of the clinical pharmacists’ interventions, the status of DRPs was determined. The interventions solved around two-third (65.8%) of the existed drug related problems. However, around 27.6% of the DRPs remained unsolved as a result of lack of coordination of prescribers, no possibility to solve the DRPs and lack of coordination of the patients. These results are consistent with those of other studies which were able to solve 58.9-68.3% of identified drug related problems [23].

Conclusion

Drug related problems were substantially high among geriatric inpatients. Patients with polypharmacy and co-morbidities had a much higher chance of developing DRPs.Hence, special attention is needed to prevent the occurrence of DRPs in these patients. Moreover, clinical pharmacists’ intervention was found to reduce DRPs in geriatric inpatients. The prescriber acceptance rate of clinical pharmacists’ intervention was also substantially high.