1 Introduction
Older adults are at an increased risk of experiencing drug-related problems (DRPs), owing to a high prevalence of multimorbidity and subsequent polypharmacy [
1‐
5]. Up to 45% of community-dwelling older adults live with polypharmacy [
6], with several studies highlighting the high prevalence of DRPs in this population [
7,
8]. Drug-related problems include adverse drug events (ADEs), adverse drug reactions (ADRs), medication errors (MEs), potentially inappropriate prescribing and drug interactions, which are associated with adverse health outcomes [
9,
10].
Drug-related readmissions are a significant adverse outcome for patients discharged from hospital, where newly initiated medications and changes in medication regimes are common. Past reviews suggest that the rate of MEs occurring after discharge is up to 53% and almost 20% of patients are affected by ADEs [
11]. Reoccurring admissions due to DRPs can impact patient safety and place a large financial burden on both the patient and healthcare system, highlighting the need to understand the prevalence and nature of drug-related readmissions. Although the prevalence of drug-related admissions has been studied previously [
12‐
14], few studies have focused on drug-related readmissions, with existing studies reporting largely variable results. Past reviews have found that the prevalence of drug-related readmissions in the general adult population has ranged from 0.09 to 64% [
15,
16]. Further, the characteristics and risk factors associated with drug-related readmissions are understudied [
15,
16]. An understanding of these factors is necessary to inform the implementation of effective policy and clinical interventions to improve patient care, decrease medication-associated harms and reduce healthcare costs [
17,
18].
The inconsistency in results from prior research highlights the need for an updated review on the prevalence of drug-related readmissions, with a particular focus on older adults. Thus, this systematic review and meta-analysis primarily aim to assess the prevalence of drug-related readmissions in older adults. Secondary objectives that will be investigated include the preventability, implicated drug classes and potential risk factors associated with drug-related hospital readmissions in older adults.
2 Method
A systematic review was conducted on the prevalence of and risk factors for drug-related readmissions in older adults in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analysis) guidelines. “Drug-related” is a broad term that encompasses ADRs, ADEs and MEs (Table
1). For this review, any study that reported on a drug-related problem that resulted in a hospital readmission was included in the screening process. This systematic review was registered in PROSPERO (CRD42023409005).
Table 1
Definitions of key terms associated with drug-related harms
DRPs | “an event or circumstance involving drug therapy that actually or potentially interferes with desired health outcomes.” [ 19] |
ADEs | “An injury resulting from the use of a drug. Under this definition, the term ADE includes harm caused by the drug (adverse drug reactions and overdoses) and harm from the use of the drug (including dose reductions and discontinuations of drug therapy).” [ 20] |
ADRs | “Response to a drug which is noxious and unintended and which occurs at doses normally used in man for prophylaxis, diagnosis, or therapy of disease or for the modification of physiologic function.” [ 20] |
MEs | “Errors in the process of prescribing, dispensing, or administering the medications that may cause or lead to inappropriate medication use or patient harm while the medication is in the control of the health care professional, patient, or consumer” [ 15, 21, 22] |
2.1 Inclusion and Exclusion Criteria
Studies were included in this review if they were published peer-reviewed articles from January 2013 to April 2023 (to ensure recency and relevancy of included studies); that examined drug-related readmissions in participants aged 65 years and over; and contained data that could allow for the calculation of prevalence such as the number of participants, all-cause readmissions and drug-related readmissions. Studies were excluded if they focused on drug-related issues such as drug abuse, intentional overdose or any illicit drug use; were case reports, case series or qualitative studies where calculating prevalence was difficult; or were studies not available in the English language.
2.2 Search Strategy
To identify relevant studies for the review, a search of MEDLINE, Embase, CINAHL, and Scopus was conducted. A search strategy was developed based on three key searchable concepts including ‘DRPs’, ‘readmissions’ and ‘older adults’ (see Tables 1–4 of the Electronic Supplementary Material [ESM]). The studies were collated in Endnote (EndNote 20; Clarivate, Philadelphia, PA, USA) and uploaded to Covidence (Veritas Health Innovation, Melbourne, VIC, Australia), where title and abstract screening was conducted by three independent reviewers (NP, EL, IW) as per the inclusion and exclusion criteria. Any conflicts regarding inclusion and exclusion were resolved by discussion among all reviewers. Three independent reviewers (NP, EL, IW) conducted full-text screening of the remaining studies to determine the final eligible studies for the review. Conflict regarding inclusion of studies was resolved by discussion between all authors.
2.3 Data Extraction and Quality Assessment
Data extraction was done independently by three reviewers (NP, EL, IW). Extracted data included the author, country, study design, setting, sample size, mean age, definition of DRP, type of DRP that resulted in readmission, number of drug-related readmissions, total number of readmissions, drug classes, clinical presentations, preventability, reported risk factors and method of causality assessment. Risk factors associated with drug-related readmissions were extracted and presented adjusted odds ratios or hazard ratios (HRs) with 95% confidence intervals, as reported by the study. Quality assessment of included studies was conducted independently by two authors (NP, EL) based on the checklist for prevalence studies developed by The Joanna Briggs Institute. Any discrepancies were resolved by discussion with all authors.
2.4 Statistical Analysis
The prevalence of drug-related readmissions was calculated by the number of drug-related readmissions divided by the total number of participants in the study. Only statistically significant findings were included where a 95% confidence interval that did not include the null value of 1 was considered statistically significant. Preventability was calculated as the number of drug-related readmissions deemed preventable divided by the number of drug-related readmissions. A meta-analysis was conducted using Meta-XL 5.3 (
http://www.epigear.com) to calculate the pooled prevalence of drug-related readmissions across all studies, and
\(I^{2}\) statistics were used to assess degree of heterogeneity. Subgroup analyses were performed to examine heterogeneity between studies involving ADR-related readmissions in terms of study design, sample size, study objective, setting, causality assessment, identification method and time to readmission.
4 Discussion
This systematic review and meta-analysis found the pooled prevalence of drug-related readmissions in older people was 9%, with a range from 3.3 to 22.2%, while the prevalence of ADR-related readmissions in older people was 6%, with a range from 3.3 to 8.7%. Overall, 15–22% of drug-related readmissions were deemed to be preventable. Several risk factors associated with an increased risk of readmission were identified including comorbidities such as cancer, liver disease, peptic ulcer disease and ischaemic heart disease, polypharmacy and specific drug classes including antibiotics, anticoagulants, psychotropics and chemotherapy agents.
A total pooled prevalence of almost one in ten was found for drug-related readmissions, and importantly, a significant proportion of cases were deemed preventable. Importantly, 15–22% of cases were deemed preventable. Past reviews focusing on the entire adult population reveal a large variation in the prevalence of drug-related readmissions, with ranges from 3 to 64% and from 0.09 to 64%, respectively [
15,
16]. One systematic review found a median prevalence of 21% (interquartile range 14–23) for drug-related readmissions, which is higher that the prevalence determined by this study. Although the preventability of drug-related readmissions has not been extensively studied, a systematic review [
15] found a preventability rate of 69% in DRP-related readmissions, which is higher than the prevalence rate found in this study. Because of significant heterogeneity between studies including variations in target populations and methods of calculating prevalence, it is difficult to compare the prevalence found in this review to past studies. However, it must be noted that a large variation in study designs does not allow for an accurate prevalence rate to be calculated. Despite these differences, it is evident that a significant proportion of older adults experience preventable drug-related hospital readmissions. Exploring risk factors associated with drug-related readmissions can provide insight into certain populations groups at a greater risk of readmissions than others to allow for targeted interventions to be implemented.
Several comorbidities were found to be associated with an increased risk of drug-related readmissions. Cancer was found to increase the risk of readmission for ADR-related cases [
23]. Chemotherapy agents are known to be associated with a high level of toxicity and ADRs. Age-related physiological changes, multiple comorbidities and drug–drug interactions may further increase the risk of experiencing an ADR, thus the increased risk of readmission is not an unexpected finding [
30]. However, multiple comorbidities were linked to DRP-related readmissions with liver disease presenting the highest risk [
24,
31]. Older adults suffer from a poorer prognosis of liver conditions because of age-related changes in liver cells [
32]. An increased risk of hepatotoxicity from the use of drugs predominantly metabolised by the liver can explain the increased risk of DRPs. Furthermore, changes in pharmacotherapy to account for decreased metabolic ability of the liver can increase the risk of ADEs. Interestingly dementia, urinary tract infections and injuries, intoxications and certain other complications or external factors were identified as protective factors. However, this may be because of dementia being classified as “caused by the progression of the disease” rather than as a DRP. Other protective factors identified are relatively unrelated to pharmacotherapy compared with other identified comorbidities [
24]. These protective factors may act as a confounder or possibly compete with the risk of readmissions where patients readmitted for these causes are less likely to have the opportunity for a drug-related readmission. A systematic review has identified the Charlson Comorbidity Index score as a prominent risk factor of drug-related readmissions, although specific conditions were not explored further. Conflicting results have been found in past papers regarding age and sex; however, this review found no association between these factors and readmissions [
15].
Use of high-risk medications, often to treat comorbidities such as those identified in this review, has also been associated with an increased risk of readmission. Anticoagulants, antibiotics, psychotropics and chemotherapy agents were the most common drug classes identified by this review [
23,
25]. Past reviews identified these drug classes as risk factors in past reviews along with other drug classes such as opioids, antihypertensives, diuretics and corticosteroids. In particular, anticoagulants were found to be most associated with preventable drug-related readmissions [
15,
16]. These drug classes align with the APINCHS [
33] (Antimicrobials, Potassium, Insulin, Narcotics and other sedatives, Chemotherapeutic agents, Heparin and other anticoagulants and Systems) acronym classifying high-risk medications [
34]. In particular, anticoagulant and psychotropic use is often seen in longer term care of older patients, increasing the likelihood of ADEs [
34‐
36]. Gastrointestinal bleeds and falls were the most common clinical presentations identified in this review, which are potentially preventable ADRs of anticoagulants and psychotropics.
Along with specific medication classes, taking an increasing number of medications was linked to an increased risk of readmission. Polypharmacy is a well-established issue, particularly in the older population, which is associated with an increased risk of medication regimen complexity, ADEs, MEs and drug–drug interactions, which can lead to an increased burden on individual health and the healthcare system [
37]. Importantly, DRPs that can occur from polypharmacy can be prevented through rational prescribing and deprescribing initiatives [
38‐
40].
Risk factors including the use of high-risk medications and polypharmacy can be managed through the implementation of targeted interventions addressing these issues. One study [
24] suggests the causes of potentially preventable readmissions were inadequate treatment (undertreatment, low dose or a lack of treatment, wrong or inappropriate treatment), a lack of monitoring or follow-up and a lack of investigations or diagnostics. Past studies suggest that communication issues and a lack of medication management are barriers to effective transitions in care that can possibly increase the risk of hospitalisation [
41‐
43]. Pharmacist-led interventions during hospital stay, discharge and post-discharge may reduce potentially inappropriate medication use and ADEs. Interventions that have been studied in the past include improved discharge planning, medication reconciliation and patient education, a medication care plan on discharge and follow-up monitoring by a pharmacist [
44‐
47].
4.1 Strengths and Limitations
This review had several strengths, particularly the extensive search strategy and extraction of data. The extensive search strategy included multiple databases that minimised the chances of relevant studies being missed by the search. The screening and extraction process involved three independent reviewers that minimised the risk of bias. However, there were some limitations in this review. The heterogeneity between the studies included in the review and the number of studies included suggests the calculated prevalence should be interpreted with caution. The search was limited to the English language and excluded unpublished articles, which increased the likelihood of excluding relevant studies. Some studies did not report on the risk factors and/or drug classes associated with drug-related readmissions, which may have led to underestimation or underreporting of information. The studies included in the review present data from European countries and therefore the results from this study may not be generalisable to healthcare systems in other countries. Furthermore, ICD-10 codes were used by a few studies to identify cases of drug-related readmissions; however, this method of identification has limited sensitivity and specificity, which may have missed eligible cases in the identification process [
48].
4.2 Implications for Practice, Policy and Future Research
Despite its high impact on health burden and healthcare costs, there is currently a lack of studies exploring drug-related readmissions in older adults and current studies are difficult to compare. A recent systematic review suggests further investigations using a more standardised approach to define drug-related readmissions and preventability to help gain a deeper understanding of possible causes and risk factors associated with readmissions [
15]. This review has also emphasised the role of comorbidities and polypharmacy in contributing to drug-related readmissions; however, further investigations are required to identify other possible risk factors, including sociodemographic and environmental factors, which can contribute to readmissions. Further research is also required to investigate other common DRPs including ADEs and MEs. An increased understanding of risk factors and causes of preventable readmissions will help inform the development of targeted interventions in healthcare settings to optimise the quality use of medicines in older adults.