Prevalence and Assessment of Potentially Inappropriate Medications among Lebanese Elderly inpatients


 Background: Potentially inappropriate medications (PIMs) prescribing in older increases the risk of hospitalization, drug-related problems, and other adverse health outcomes by two to three folds (Lu WH, 2015) (WHO, 2018). Hence several regulations have been implemented in geriatric clinical practice to improve drug appropriateness. This study aimed to assess the prevalence of PIMs identified by the Beers 2019 criteria among Lebanese elderly inpatients, and to identify the factors that may influence such inappropriateness.Methods: This observational study was conducted among elderly patients of 3 hospitals in Lebanon. Data were collected from patients using a questionnaire based on the 2019 Beers criteria. Descriptive statistics and multivariable logistic regression were used to describe and identify the predictors of PIMs use. Statistical analysis was carried out using the SPSS software version 22.0.Results: A total of 110 in-patients aging 65 years and older were analyzed. 80% of patients were taking drugs that are considered PIMs to be avoided; The most commonly prescribed PIM was proton pump inhibitors (PPIs) (60%). PIMs use was significantly associated with polypharmacy and other comorbid conditions (p-value <0.05). The prevalence of PIMs to be used with caution was 78.2%, those having drug-disease interactions were 22.7% and those having clinically important drug-drug interactions were 15.5%.Conclusion: A high prevalence of PIMs use in Lebanese hospitalized patients was observed, and was significantly associated with polypharmacy and other CNS conditions. This study supports the need for the implementation of clinical guidelines and staff to improve drug prescribing in older adults. Trial registration: retrospectively registered.


88
It is well documented that the worldwide proportion of older people aging 65 years or more is expected to nearly double 89 from 12% to 22% by 2050, causing an increased shift in the overall health cost and demand for services (WHO, 2018). A 90 major patient safety concern in elderly is the increased risk of adverse drug reactions, which is manifested as falls, 91 orthostatic hypotension, delirium, renal failure, gastrointestinal and intracranial bleeding (Lavan AH, 2016). Many factors 92 may contribute to such risks including age-related changes in pharmacokinetics and pharmacodynamics, increasing 93 burden of comorbidity, polypharmacy, use of potentially inappropriate medications (PIMs) and suboptimal monitoring of 94 drugs (Lavan AH, 2016). PIMs are defined as "drugs for which use among older adults should be avoided due to the high 95 risk of adverse reactions for this population and/or insufficient evidence of their benefits when safer and equally or more 96 effective therapeutic alternatives are available" (American Geriatrics Society, 2019). PIMs prescribing in older adults is a 97 common cause of increased risk of hospitalization, drug-related problems, and other adverse health outcomes (Lu WH, 98 2015) (Alhawassi T, 2019). To prevent PIMs use, different criteria have been set by experts including the Beers criteria, a 99 compendium of medications worth discussing with health professionals because they may not be the safest or most 100 appropriate options for older adults (American Geriatrics Society, 2019). Recently, it has been adopted by the American 101 Geriatric Society (AGS) to provide guidance regarding medication selection, educate clinicians and patients, and reduce 102 adverse drug events (American Geriatrics Society, 2019). Many trials have been conducted in various countries to study 103 the use of PIMs in different settings including ambulatory primary care, community, hospital, and nursing homes. In 104 Lebanon, no studies have studied the prevalence of PIMs identified by the updated 2019 version of the Beers criteria in 105 hospitalized elderly patients. Thus, the aim of this study was to assess the prevalence of PIMs identified by the American 106 Geriatrics Society Beers Criteria among Lebanese hospitalized elderly patients, and to explore risks that may influence 107 such inappropriateness. 108 109

110
This cross-sectional observational study was conducted in three main Hospitals in Bekaa/ Lebanon within 8-111 weeks duration. Data was collected from patients' profiles using the hospital's systems, and directly from patients 112 or caregivers. Subjects were assessed by officially licensed Lebanese pharmacists who were trained to present and 113 collect data in order to maintain its uniformity. Ethical approval was obtained from the School of Pharmacy at the 114 Lebanese International University, and verbal informed consent was taken from the patients/ caregivers and from 115 the 3 hospitals' administrations. A written informed consent was also obtained. 116 The questionnaire used was based on the 2019 Beers criteria, and was divided into three main parts: Demographic 117 properties, Comorbidities/ disease states, and medications taken. Medications and co-morbidities were then classified into 118 categories based on Beers' criteria division also, for example, medications were classified as: (1) medications to be 119 avoided, (2) medications to be used with caution, (3) medications having drug-disease or drug syndrome interaction that may exacerbate the disease or syndrome, and (4) medications having clinically important drug-drug interactions that 121 should be avoided. PIMs use also was classified into two categories: PIMs use, and no PIMs use. Data were coded and 122 entered for analysis. Descriptive testing was used to summarize and describe subjects' related information; frequencies 123 and percentages were used to describe demographic properties of patients and primary outcomes. Fisher exact test was 124 used to determine statistical significance in the bivariate analysis between PIMs use and subjects' characteristics and 125 comorbidities, while logistic regression was used in the multivariable analysis. The 95% confidence intervals (CI) were 126 based on likelihood. All reported p-values were 2-sided and considered significant if they were less than <0.05. Data were 127 entered and analyzed using software IBM Statistical Package for Social Science (SPSS) version 20. 128 129

130
During the eight-week study, 110 subjects were enrolled in the study by convenient sampling. Average age was 36.9 ± 131 13.5 (range between 18 and 85 years), and the majority were females. Subjects above 65 years were divided according to 132 age group, with 40 subjects aging between 65 and 69 years, 17 aging between 70 and 74 years, 13 aging between 75 and 133 79 years, and 38 aging more than or equal to 80 years. The majority of the study population had chronic conditions, 134 particularly hypertension and diabetes [ Figure 1]. The prevalence of polypharmacy was high among subjects (73.6%) with 135 significant associations. Detailed participants' socioeconomic data are presented in [Table 1]. 136 The prevalence of PIMs in the tested population was 105 out of 110 (95.5%) -one subject may have more than 137 one type of PIM-, and was divided as follow: PIMs to be avoided (80%), followed by PIMs to be used with 138 caution (78.2%), PIMs having drug-disease or drug syndrome interaction that may exacerbate the disease or 139 syndrome (22.7%) and PIMs having clinically important drug-drug interactions that should be avoided (15.5%) 140 [ Table 2]. The most commonly prescribed PIMs to be avoided were gastrointestinal agents (particularly PPIs) 141 followed by CNS acting drugs [ Table 3]. The most commonly prescribed PIMs to be used with caution were 142 antipsychotics (48.2%), followed by diuretics (30.9%), while aspirin and SSRIs were in equal proportions (15.5%) 143 [ Table 4]. In regards to drug-disease interactions, the most commonly encountered drug-disease interactions were: 144 1) Falls with SSRIs (4.5%), antipsychotics (3.6%), BZDs (3.6%), 2) Dementia with BZDs (4.5%), antipsychotics 145 (4.5%), 3) Delirium with antipsychotics (3.6%). Two main drug-drug interactions were noticed: the use of 3 or 146 more CNS active drugs with a proportion of 14.5%, followed by the concomitant use of 2 anticholinergics with a 147 proportion of 1.8%. 148 The prevalence of PIMs to be avoided was associated with different variables like marital status and 149 polypharmacy, and conditions like hypertension, diabetes mellitus, falls and bipolar. Detailed percentages and P 150 values are presented in [Table 5]. In the multivariate analysis, older adults with polypharmacy, and conditions like 151 dementia, falls, insomnia, and schizophrenia were associated with PIMs. For example, 88.9% of patients who 152 were taking more than or equal 5 drugs have a PIM to be avoided, while 11.1% don't (P value= 0.001) [ Table 6]. 153 Another example is in the case of falls, where 94.7% of patients suffering from this condition have PIMs to be 154 used with caution, while 5.3% don't (P value = 0.05). Adjusted odds ratios and 95% confidence intervals (CI) for 155 factors associated with PIMs use are presented in [Table 6]. 156

160
It is well known that PIMs use among older patients is associated with negative health consequences and can impact 161 patients' quality of life, with increased hospitalization risk, drug related problems and cost burden on the healthcare 162 system (Alhawassi T, 2019). Clinical studies show there is high prevalence of PIMs among elderly, ranging between 163 53.5% to 58.4%, which is consistent with the results of this study (Zhang X, 2017) (Ní Chróinín D, 2016) (Endres H, 164 2016). The prevalence of PIMs in this sample of the Lebanese population was apparently higher than that in other 165 countries, especially for the PIMs to be avoided (80%), and the PIMs to be used with caution (78.2%). Reasons behind 166 these results may intersect, where the absence of specific in-patient criteria to assess the use of medications for elderly 167 patients in the assessed Lebanese hospitals could be the main one. The absence of clinical pharmacists in the assessed 168 hospitals may have contributed to these results, where the comprehensive medication reviews clinical pharmacists do, has 169 been shown to improve overall patient health and alleviate health care burdens (Nemes S, 2009). Previous studies 170 strongly support the clear role for pharmacists in working directly or collaboratively to improve medication use and 171 management in older population (Lee J, 2015). Pharmacist integration into a multidisciplinary team has been shown to 172 have a positive impact in the management of a variety of different disease states. Different assessment tools or versions 173 may also contribute to such difference. Also, a marked difference in the prevalence of PIMs was shown when applying 174 two versions of beers criteria (2003 and 2012) on the same population, as results varied by a 9% increase (48% versus 175 59% respectively), which could be another cause for the results variation (Baldoni ADO, 2014). 176 Being the most prevalent PIM to be avoided (60%), proton pump inhibitors (PPIs) are among the most used medications 177 in hospitals, while antipsychotics (48%) and diuretics (31%) are among the most common PIMs to be used with caution. 178 Long term use of PPIs without a specific clinical indication is considered a PIM in older adults, increasing the risk of C. The most important independent predictor for having a PIM in the tested population was found to be polypharmacy, 192 especially when the majority of patients suffer from concomitant chronic conditions. It is an area of concern for elderly 193 To our knowledge, this is the first study in a developing country like Lebanon that investigates the prevalence of PIMs for 213 in-patients, and particularly in Bekaa area, were no university hospitals are found. Thus, it was essential to assess the 214 prevalence of such inappropriateness in areas were medications' safety and efficacy protocols are uncommon, under-215 estimated in practice, or even absent. A validated and updated criterion adopted by the American Geriatric Society (AGS) 216 was used in assessment, which gives the study more strength and validity. The primary limitation of this study is the small 217 sample size. Data was collected during the Covid-19 pandemic, where approaching patients and files was low. Small 218 sample sizes usually lead to the overestimation of odds ratio and to the reduction in the statistical power of a study (Deziel 219 C, 2020). Also, only hospitalized patients were included which is not representative of community-based aged patients 220 who were recently discharged. The last limitation is related to the nature of the study; it is cross sectional which doesn't 221 allow us to determine the presence or absence of a temporal link between PIMs use and its determinants. High prevalence of potentially inappropriate medications in the Lebanese hospitalized elderly population was shown to be 226 evident. There was a strong association between potentially inappropriate medications use and polypharmacy, which is 227 considered a serious challenge in the geriatric healthcare sector. These findings underscore the need to improve the 228 medication prescription quality for the elderly by implementing clinical guidelines, reducing the number of unnecessary 229 prescribed medications when possible, and assisting physicians in delivering appropriate medications. Pharmacists, 230 working in a multidisciplinary team can play a significant role in reducing PIMs use among elderly, and thus their absence 231 may affect the overall clinical outcomes, safety, and efficacy of the prescribed drugs. This role can be illustrated in reviewing the patients' prescriptions, detecting PIMs and recommending appropriate drug use, and creating changes in 233 prescribing practices in accordance with guidelines from the literature. Ethics approval and consent to participate 239 This study design and methods were carried out in accordance with relevant guidelines and regulations under Ethics 240 approval and consent to participate. Ethical approval was obtained from the School of Pharmacy at the Lebanese 241 International University and the hospitals' administrations. Written informed consent was also obtained.

243
Consent of publication 244 Not applicable 245 246 Availability of Data and Material 247 Data sharing not applicable to this article as no datasets were generated or analyzed during the current study 248 249 Conflicts of interest 250 The authors declare that they have no competing interests 251 252 Funding 253 None 254 255 Authors' contributions 256 The authors confirm contribution to the paper as follows: study conception and design: