ORCID account: 0000-0003-1763-7334
Total number of words: 4742
Number of words abstract: 243
Number of Tables: 3
Number of Figures: 0
Number of references: 21
Abbreviations: DRP = drug related problem, NA = not applicable, SD = standard deviation.
3.2 Prevalence of DRPs
A total of 548 DRPs were identified, with a mean (standard deviation, SD) of 1.5 (1.3) DRPs per patient. The mean time for performing a medication review was 10 minutes per patient. Demographic and drug-related characteristics are shown in Table 1. The majority of included patients had at least 1 DRP. The most common problem was treatment safety (44%), followed by treatment effectiveness (31%), and unnecessary treatment (20%). Causes of DRPs are presented in Table 2. From the total of 548 DRPs, the most common causes were inappropriate drug selection (72%), and dose selection (18%). Within the group of inappropriate drug selection, the most common cause was the absence of an indication (25%), followed by no treatment for the indication (14%), and drug selection according to the guidelines but contra-indicated (13%). In the group of inappropriate drug selection, the majority of DRPs were caused by drug doses being too high (9.3%), and by too frequent dosing regimens (4.6%).
† This category includes, among other causes, considering a fixed dose combination and allergy registration in the electronic patient management system.
Abbreviation: TDM = therapeutic drug monitoring
3.3 Treatment domains of DRPs
Table 3 represents the distribution of the most common DRP causes per ATC group. The most frequently involved class of drugs differed per DRP cause. Within the group of inappropriate drugs according to guidelines (C1.1), the majority of drugs affected the nervous system (n = 41; 66%) including interventions such as stopping of tramadol and oxazepam. For the group of contra-indicated drugs (C1.2), the most common ATC group involved the cardiovascular system (n = 35; 49%). In this group, the majority of patients experienced falls while using multiple antihypertensive agents, or electrolyte imbalance during diuretic use. For drugs without an indication (C1.3), the majority of drugs affected the alimentary tract and metabolism (n = 82, 59%) which mainly involved redundant use of a proton-pump inhibitor. In the DRP group with no treatment for indication (C1.6), most drugs affected the alimentary tract and metabolism (n = 28, 37%). Here, recommended interventions included for instance the absence of vitamin D supplementation and use of laxatives during opioid treatment. Within the group of too many drugs for an indication (C1.7), the most common ATC group involved the cardiovascular system (n = 32; 91%) due to falls or low blood pressure while using multiple antihypertensive agents. For the group of inappropriate dose selection (C3), the majority of drugs affected the nervous system (n = 33; 34%). The majority of these patients received a total dose paracetamol of 4000 mg/day for chronic use instead of 3000 mg/day. In the DRP group with inadequate treatment duration (C4), the most common ATC group involved the genitourinary system (n = 13; 42%) caused by long-term tamsulosin use without previous stop attempts.
† This category includes, among other causes, the absence of therapeutic drug monitoring or the use of controlled-release tablets instead of immediate release tablets.
3.4 Acceptance rates
After ED visit, 169 patients with at least one DRP were admitted to the hospital. Recommended interventions were accepted by the clinical physician for 93 patients (55%) and recommendations were documented in the discharge letter to primary care for 61 patients (36%) upon discharge.
Of 86 patients being discharged directly from the ED with at least one DRP, recommended interventions were included in the discharge letter to the general practitioner or elderly care specialist for 47 patients (55%). The acceptance rate of these recommendations was 32% (15 out of 47). Five patients passed away during the follow-up period.
-
Discussion
In this ’real-world’ study, 548 DRPs were identified in older patients living with frailty visiting the ED with a mean of 1.5 DRP per patient. Of identified patients, 86 patients (24%) had no DRPs, 104 patients (29%) had 1 DRP, 95 patients (27%) had 2 DRPs and 71 patients (20%) had at least 3 DRPs. The majority of DPRs were caused by inappropriate drug selection (72%) and dose selection (18%) with a diffuse distribution per ATC group. The acceptance rate or recommended interventions among clinical physicians was 55% compared to 32% among general practitioner and elderly care specialists.
The prevalence of DRPs in this study was 76%. Similar percentages have been described in literature.
In older patients admitted to the medical and surgical wards, DRPs were identified in 82% of patients [18]. The same percentage of 82% was found in patients ≥ 65 years visiting the ED [19]. Furthermore, the mean number of DRPs per patient was comparable [19]. These results show that by practical implementation of a medication review program at the ED, comparable DRP rates are found as in interventional trials.
The acceptance rate of recommendations among general practitioners for older patients at the ED varies in clinical trials from 27–66% [8, 20]. In admitted patients with at least one recommendation, the acceptance rate among clinical physicians varies from 55–66% [19, 21]. In our study, the acceptance rate of recommendations among general practitioners and elderly care specialists was low (32%). This could have several explanations. First, general practitioners received recommendations via discharge letter. To our surprise, only 55% of recommendations were included in discharge letters, meaning that almost half of the DRPs were not brought to the attention of the general practitioner or elderly care specialist. This may be due to a preselection of relevant DRPs in the discharge letter by the ANP or geriatrician, or because the recommendations simply got lost in follow-up. Second, general practitioners and elderly care specialists were not informed about the geriatric assessment including a medication review, while clinical physicians were. Third, general practitioners may know of treatment indications that are unknown in the clinical setting, such as nervousness, for which medications need to be continued.
A strength of this study is selection of the study population, as older patients living with frailty are prone to experiencing DRPs. Furthermore, results of this study show the prevalence and type of DRPs in a “real-life” setting of conducting medication reviews at the ED. Another strength of this study was the identification of acceptance rate in admitted patients and in patients being discharged from the ED, as most studies do not determine the acceptance rate for both patient groups.
This study also has some limitations. As described, the acceptance rate for discharged patients was low. Moreover, the geriatric assessment suffered from restrictions during the Covid-19 pandemic, as general practitioners were under high pressure and workload. Additionally, the impact of medication reviews at the ED was not investigated related to clinical outcome measures.
-
Practical recommendations
We provide the following suggestions for implementing a medication review program at the ED or other departments by which we hope to stimulate other healthcare professionals to start similar projects.
Before implementation, inform the surrounding healthcare network about the medication review program. This includes clinical physicians, general practitioners, elderly care specialists and pharmacists at a public pharmacy. This way, recommendations will have more support and this could potentially improve acceptance rates. Our project clearly benefited from the close collaboration with the geriatric department, resulting in higher acceptance rates in admitted patients.
Select older patients living with frailty for medication reviews rather than older patients (≥ 70 years) with polypharmacy (≥ 5 drugs) as described in guidelines. Performing medication reviews is time-consuming and by adequate patient selection, the efficacy of the review process may be improved. Furthermore, vulnerable patient populations are more prone to experiencing DRPs.
Include medication reviews in daily routine, as such that it is a time-efficient process done by a group of pharmacists routinely. In our experience, conducting medication reviews could be easily implemented in routine care and was well received amongst physicians.
Educate treating physicians so they know what type of DRPs are often seen at your hospital. In our study, we identified drug classes that often cause DRPs at the emergency department and presented those to physicians regularly. As a result, at the end of the study period, common DRPs were already identified by the physician before the pharmacists’ consultation.
If possible, include pharmacist-led patient interviews in the medication review process to identify medication related problems and patients’ experiences and wishes. This may further eliminate harmful medication use.
Optimize transfer of (information about) medication recommendations in the discharge letter by including accepted and yet to be considered recommendations. Even more, we believe medication changes should be an integral part of the discharge letter, regardless of the reason for ED visit.
Conclusion
Our results show that pharmacy-led medication reviews in daily routine lead to better assessment of DRPs at the ED. Of included geriatric patients at the emergency department, 76% had at least one DRP, with a mean of 1.5 DRP per patient. The acceptance rate among clinical physicians in admitted patients was higher (55%) than the acceptance rate among general practitioners/ elderly care specialists in discharged patients (32%). Results of this study reflect the real-life setting of conducting medication reviews at the ED for older patients living with frailty.
-
List of abbreviations
ANP = advanced nurse practitioner
ATC = anatomical therapeutic chemical
DRP = drug related problem
ED = emergency department
PCNE = The Pharmaceutical Care Network Europe
SD = standard deviation
START STOPP criteria = Screening Tool to Alert to Right Treatment and Screening Tool of Older Persons' Prescriptions criteria
WHO = world Health Organization
Tables
Table 1. Demographic and drug-related patient characteristics (n = 356). Values are reported in number (n) and %, unless otherwise specified.
† This category includes others infections, dyspnea (due to COVID-19) and unbearable pain.
Abbreviations: DRP = drug related problem, NA = not applicable, SD = standard deviation.
Table 2. Causes of DRPs according to the Pharmaceutical Care Network Europe (PCNE) classification system. In total, 548 DRPs were identified during the study period.
† This category includes, among other causes, considering a fixed dose combination and allergy registration in the electronic patient management system.
Abbreviation: TDM = therapeutic drug monitoring
Table 3. Distribution of causes of drug-related problems (DRPs) per anatomical therapeutic chemical (ATC) group. Numbers in bold are the highest values in each column.
† This category includes, among other causes, the absence of therapeutic drug monitoring or the use of controlled-release tablets instead of immediate release tablets.
Supplementary files
Supplementary Table 1. Trigger tool to identify adverse events [11].
Abbreviations: ACE = angiotensine converting enzyme, AII = angiotensine II, INR = international normalized ratio, NSAIDs = non-steroidal anti-inflammatory drugs