Skip to main content
NCBI home page
Systematic Reviews logoLink to Systematic Reviews
. 2026 Apr 2;15:176. doi: 10.1186/s13643-026-03068-2

Prevalence and determinants of polypharmacy among cancer patients: a systematic review and meta-analysis

Mohamed M Tawengi 1,2,#, Jawaher Baraka 1,2,#, Rafal Al Shibly 1,#, Mohammed I Danjuma 1,2,3,4,✉
PMCID: PMC13200404  PMID: 41923255

Abstract

Background

Polypharmacy (PP) is a rising clinical challenge among patients with a cancer diagnosis. Uncertainty remains regarding its exact burden, exact prevalence estimates, and definitional themes in this vulnerable cohort of patients.

Methods

We searched PubMed, EMBASE, Scopus, the Cochrane Database of Systematic Reviews (CDSR), and Google Scholar, for studies published between 2000 and 2025 for eligible studies reporting on polypharmacy in cancer patients. These were critically appraised for eligibility and inclusion by two independent reviewers. Using quality and random effect models, pooled estimates of the prevalence of PP, prevalence by type of cancer, and geographical spread were determined. The prevalence rates of potentially inappropriate medications (PIMs) and drug-drug interactions (DDIs) were also estimated. Heterogeneity among the included studies was reported by corresponding I2 estimates.

Results

This meta-analytical review involved 20 studies comprising (n = 102,100) participants. The overall pooled prevalence of polypharmacy among patients with cancer was 29% (95% CI 10–52%) using the quality effects model, and 58% (95% CI 50–62%) using the random effects model. The overall heterogeneity among the included studies was significant (I2 = 100%, p < 0.001) for the random effects models.

Conclusion

From this meta-analysis of studies with diverse designs, we found a high pooled prevalence of polypharmacy among patient cohorts with cancer, with marked variability across studies. Given this level of heterogeneity, future prospective and systematic studies are needed to better characterize the determinants and consequences of polypharmacy to guide strategies that may improve patient outcomes in these cohorts.

Systematic review registration

PROSPERO, Number CRD42024576772.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13643-026-03068-2.

Keywords: Polypharmacy, Cancer, Epidemiology, Systematic review, Meta-analysis

Background

Polypharmacy is a prescribing practice associated with adverse outcomes [1]. Polypharmacy (PP) has generated and continues to generate increasing concerns among clinicians and policymakers owing to its negative consequences across different age groups [1]. Recent evidence has shown that the prevalence of PP is variable but is reported to exceed 20% in some studies [2, 3]. Among older patients diagnosed with cancer, the prevalence of PP is highly variable, ranging from 11 to 96% [4]. This variability in prevalence estimates extends to differences in study designs and patient populations, with indicating a rising burden amongst older patient population [2, 3]. In older patients with cancer for example, reported polypharmacy prevalence rates range widely from as low as 11% up to 96%; reflecting differences in how polypharmacy is defined (e.g., differing medication-count thresholds) among other factors. Because older cancer patients comprise a subset of the general geriatric population, polypharmacy is expected to be common in both groups. However, the extreme range observed in the cancer patient cohorts suggests that disease-specific factors (such as oncologic treatment regimens and multiple comorbidities) can further amplify medication use beyond baseline rates observed among older adults [2, 3]. This patient population is characterized by a high burden of comorbidities and geriatric syndromes, which increases overall medication exposure and, consequently, the likelihood of potentially inappropriate medication (PIM) use and drug–drug interactions (DDIs). Several studies have reported that a substantial proportion of identified interactions involve over-the-counter medications rather than anticancer therapies, with chemotherapy accounting for approximately 26% of PIMs in some cohorts [5]. Importantly, numerical polypharmacy and PIM use have been consistently associated with adverse clinical outcomes and higher rates of clinically significant DDIs; however, these relationships are derived from observational data and should not be interpreted as causal [5]. Among the range of factors suggested to explain these observations include increase in prescription rates over the past 20 years [6]driven by multiple comorbidities associated with aging populations [7], and the inevitable rise in the number of medications required to manage them. Additionally, this recent increase in prescription rates in the general population but more so in cancer patients may be partly a consequence of our increasing understanding of therapeutic interventions that increase the survival of various organ-specific morbidities. Higher medication counts, particularly when driven by evidence-based, mortality-reducing therapies (often referred to as appropriate polypharmacy), are not inherently harmful. Instead, the greater risk of medication-related harm arises from the complex combinations of these necessary drugs with additional treatments for comorbid conditions and over-the-counter agents [8]. This represent the reason why more studies are needed to elucidate the extent of polypharmacy clinical phenotypes and the implications of PP in patients with a cancer diagnosis [7]. Until key determinants of PP in cohorts of patients with a cancer diagnosis are fully understood, resourceful planning of intervention strategies necessary to address will continue to falter. Among these determinants is the lack of clarity regarding definitional thresholds, the exact case burden of PP in patients with a cancer diagnosis across populations, and differences study methodology among others. Outcomes from the few published studies attempting to address these often been conflicting, sometimes adding to the existing uncertainty [1]. For example, a recent systematic review in older cancer patients reported an association between polypharmacy, increased PIM use, and higher all-cause mortality, highlighting its potential impact on morbidity and mortality. In contrast, several other reviews in oncology populations have been less explicit regarding survival outcomes, largely due to heterogeneity in study design, outcome reporting, and residual confounding, and have therefore focused primarily on prevalence, prescribing quality, and medication-related risks rather than mortality [5, 9–17]. A consequence of these discordant outcomes the lack of consensus around key themes such as the exact definitional threshold of PP in this cohort of patients [1, 2].

In the light of the deficits in our understanding of PP in patients with cancer diagnosis, we carried out this review to systematically explore the exact prevalence of PP, its various clinical phenotypes, and the downstream consequences such as PIMs and DDIs, in patients with cancer diagnosis across all adult age ranges. This is to ascertain the exact clinical burden of polypharmacy as well determine discernible global trends.

Materials and methods

Registration and methodology reporting

This systematic review and meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist and the Cochrane Handbook guidelines [3, 4]. The review protocol was registered on the Prospective Register of Systematic Reviews (PROSPERO) under the registration number: CRD42024576772.

Patient population

Studies reporting on patient populations with cancer diagnosis (regardless of the histological type) who had medication counts labeled as polypharmacy and reported as such. The main outcome is the pooled estimate of polypharmacy period prevalence among these studies.

Data source and search strategy

A comprehensive literature search was carried out using PubMed, EMBASE, Scopus, Cochrane Database of Systematic Reviews (CDSR), and Google Scholar (first 50 pages) electronic databases, and search engines from 2000-February 2025 for studies meeting the eligibility criteria for inclusion in the review. These databases were utilized considering their validity as updated resources for most current medical literature. We utilized the following terms to search for eligible studies that satisfied the review’s inclusion criteria: (polypharmacy[MeSH Terms] OR polypharmacy OR “multiple medications” OR “multiple medication” OR “multiple drug therapy” OR “multi-drug therapy” OR “multiple drug use” OR “concurrent medication” OR “concurrent drug” OR “complex medication regimen*” OR “medication burden” OR “drug burden” OR “medication count” OR “high medication use” OR “inappropriate medication*” OR “potentially inappropriate medication*” OR PIM OR PIMs OR “Beers criteria” OR “STOPP criteria” OR “medication appropriateness” OR deprescribing) AND ( neoplasms[MeSH Terms] OR cancer* OR malignant* OR “malignant tumor*” OR “malignant neoplasm*” OR carcinoma* OR adenocarcinoma* OR sarcoma* OR leukemia* OR lymphoma* OR myeloma* OR “hematologic malignancy” OR “solid tumor*” OR oncology*)**. We reviewed gray literature for similar articles that were not captured in the databases. The search was limited to “English language” and “Human species” as applicable to each database.

Study eligibility criteria

This systematic review and meta-analysis examined studies of patient cohorts with any cancer diagnosis who had medication counts labeled as polypharmacy and reported as such. We included studies satisfying the following eligibility criteria: (1) studies reporting numerical data on the prevalence of polypharmacy; (2) studies on patients with any cancer diagnosis; (3) studies with participants aged ≥ 18 years; and (4) studies published in the English language. We included studies with all definitional thresholds of polypharmacy (e.g., ≥ 5, ≥ 10 medications). However, our primary analysis uses PP defined as ≥ 5 medications, which is the most accepted definition. We have other definitions in the summary of included studies for the purposes of reporting additional iterational definitions explored by those studies. Case reports, case series, and reviews were excluded as were studies that included patients with polypharmacy but without a cancer diagnosis.

Study selection

All the retrieved studies were imported into Rayyan Computing Research Institute (QCRI 2016) software. Two independent reviewers (JB and MT) screened the titles, abstracts, and full texts of the records to ascertain eligibility for inclusion in the review. Disagreements between reviewers were resolved by consensus, failing which this was resolved by the third reviewer (MD). A final list of selected studies was then drawn and included in the review.

Data extraction

A final data collection sheet was created after a trial on randomly selected studies. Two independent reviewers (JB and MT) then extracted the following variables from the studies: author, year of publication, study design, site, country of publication, population, sample size, primary cancer diagnosis (solid tumors only, hematological malignancies only, and combined), proportion of cancer patients with polypharmacy/PIMs/DDIs/adverse drug reactions (ADRs), iteration of polypharmacy definition (where available), duration of study, duration of follow-up, mortality, and Charlson comorbidity index (CCI).

Quality assessment

The risk of bias of the included studies was assessed using The Methodological Standards for Epidemiological Research (MASTER) scale. A comprehensive description and operability of this tool is given elsewhere [5]. Briefly, it assesses the risk of bias of clinical research studies across analytical designs through a unified framework [5]. The tool has 36 unique methodological safeguards that are categorized into seven methodological standards to be fulfilled in the MASTER scale [5]. Two reviewers (JB and MT) independently assessed the methodological quality. Disagreements were resolved through consensus or by adjudication by a third reviewer (MD). The results of the risk of bias assessment are available from the corresponding author upon reasonable request.

Statistical analysis

Continuous variables are presented as means (± standard deviation [SDs]) or medians (interquartile range [IQRs]) as appropriate, whereas categorical variables are presented as numbers (percentages). We quantified the pooled prevalence estimates of polypharmacy (utilizing the random effects model [REM], fixed effects model [FEM], and quality effects model [QEM]) among patients with cancer. The traditional methods of pooling of the effect sizes in meta-analytical synthesis are the fixed and random effects models [6]. The fixed effects models utilizing inverse variance methods are limited by the assumption that all studies estimate a single, common effect size, which is often unrealistic in clinical research because of obvious inherent variability in study populations, the scope of interventions, and, most importantly, study designs [6]. Owing to this “no between-study variation,” the FEM has limited generalizability since it assumes that its results cannot readily be generalized beyond the included studies. Conversely, the REMs, although it attempts to correct FEM flaws, is limited by its attempt at weighing. It assumes two sources of variability in effects, one from study-level differences and the other from sampling error [6]. This results in several problems, including overestimation of heterogeneity (resulting in erroneous assumption that all variability between studies is random, potentially obscuring the impact of systematic differences in study quality or methodological rigor); marked sensitivity to study size, often resulting in smaller studies being allocated disproportionately greater weights than a fixed-effect model, and assignment of weights on the basis solely of within-study and between-study variance, ignoring the methodological quality or risk of bias in individual studies amongst other limitations. For the primary outcome analyses of this review, we have chosen the quality effect model, due to its ability to correct for the most important source of individual study heterogeneity (differing methodological quality and risk of bias) by introducing a safeguard called a quality score (with a quality index [Qi] derived from this). This ensures that all studies are assessed based on fixed sets of predefined safeguards against risk of bias. It takes “quality rank” values between 0 and 1 (with 0 and 1 representing studies with no safeguards, and all safeguards, respectively). This resulting redistribution of weights due to Qi would then help reduce estimator variance, which is a liability with the REM. These scores were then incorporated into the quality effects model. When significant heterogeneity was reported, we performed subgroup analyses to investigate the effect of the source of primary data and the risk of bias scores of the reviewed studies on the final point prevalence estimates. We assessed the heterogeneity between studies with theI2 statistic and the τ2statistic [7]. We considered theI2 thresholds of 25%, 50%, and 75% to represent low, moderate, and high heterogeneity between-study variances, respectively. When τ2was reported to be zero, this was indicative of no heterogeneity [8]. We utilized funnel and Doi plots to visualize small-study effects and publication bias [9]. Finally, we performed sensitivity analyses excluding each study to ascertain its effect on the final point prevalence estimate. All the statistical analyses were carried out with Meta XL, version 5.3 (EpiGear International, Queensland, Australia).

Study selection

A total of 404 citations were retrieved from the initial literature search; of these 20 studies were subsequently included in the systematic review and meta-analysis following exhaustive eligibility assessment (Fig. 1) [10–16, 18–29]. The most common reason for exclusion was the absence of sufficient data to estimate the prevalence of polypharmacy.

Fig. 1.

Fig. 1

Open in a new tab

PRISMA flow diagram of the study selection process

Characteristics of the included studies

Overall, a total of (n = 102,100) participants were included across the 20 studies in this systematic review and meta-analysis, out of which (n = 29,333) had polypharmacy. The included studies were conducted across Europe, Asia, the Middle East, and North America. Studies from Japan accounted for four publications [23, 24, 28, 30], while South Korea contributed three studies [25, 31]. Two studies each originated from Ireland [15, 27], Spain [13, 29] France [11, 21], and the USA [12, 14, 26]. Single studies were reported from Denmark [22] the Netherlands [16] Australia [18], Qatar [20], and Croatia [10]. Most included studies employed observational designs, predominantly retrospective cohort or cross-sectional analyses [10, 14–16, 18, 19, 21, 22, 26, 27, 31]. A smaller subset were conducted as prospective cohort studies [12, 25, 29]. No randomized controlled trials were identified.

Thirteen studies were conducted in outpatient or ambulatory oncology settings [10, 13–16, 19, 22–24, 26–29]. Five studies recruited patients exclusively from hospital inpatient settings [12, 18, 21, 25, 31], while two studies included participants from mixed clinical settings [11, 20]. Reported follow-up durations varied widely, ranging from short-term assessments (e.g., [28]) to long-term longitudinal follow-up [22].

Eleven studies included patients with both solid and hematologic cancers [10, 13–15, 19, 22, 25–27, 29, 31] whereas nine focused exclusively on solid tumors [11, 12, 16, 18, 20, 21, 23, 24, 28].

Quality of the included studies

The overall methodological quality of the included studies was moderate to good, mainly because of issues related to the sample size calculation and the appropriateness of the outcome measures (SUPP 3).

Prevalence of polypharmacy in patients with cancer diagnosis

Figure 2 presents a forest plot summarizing the pooled prevalence of polypharmacy among cancer patients across 20 included studies using the quality-effects model (QEM). The overall pooled prevalence was 29% (95% CI 10–52%), indicating that nearly one-third of patients with cancer were exposed to polypharmacy. Substantial between-study heterogeneity was observed (I2 = 100%, Q = 8719.15, p < 0.001), reflecting signficant variability in prevalence estimates across studies. Reported estimates ranged from approximately 20 to 80%, with most clustered between 40 and 70%. This wide dispersion likely stems from differences in cancer type, study design, age, comorbidity burden, healthcare setting, and numerical thresholds used to define polypharmacy.

Fig. 2.

Fig. 2

Open in a new tab

A forest plot of the prevalence of polypharmacy among cancer patients across reviewed studies by quality effects model approach. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals. The diamond at the bottom of the plot represents the overall pooled estimate, and the horizontal lines indicate the 95% confidence intervals for each study. No continuity correction was applied, as the double-arcsine transformation used in the QEM accommodates studies with extreme proportions, ensuring stable variance estimation [32]

The overall pooled prevalence of polypharmacy among patients with cancer using the random effects model was 58% (95% CI: 50–67%), as shown in Fig. 3. The overall heterogeneity among the included studies was significant (I2 = 100%, p < 0.001). Supplementary material (SUPP 1 and SUPP 2, respectively) additionally summarizes estimates due to fixed effects and inverse heterogeneity models.

Fig. 3.

Fig. 3

Open in a new tab

A forest plot of the prevalence of polypharmacy among cancer patients across reviewed studies by the random effects model approach. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals

Prevalence estimates by type of cancer

Subgroup analyses of polypharmacy prevalence by tumor type using the quality-effects model is shown in Fig. 4. Distinct differences were observed across cancer subtypes. In studies restricted to solid tumors, the pooled prevalence was 27% (95% CI 10–47%; I2 = 100%), indicating substantial heterogeneity within this group. In contrast, studies of hematologic malignancies demonstrated a markedly higher prevalence of 68% (95% CI 53–81%; I2 = 82%), consistent with the intensive, multi-agent therapeutic regimens typical in hematologic oncology. Mixed-cohort studies that included both solid and hematologic cancers reported the highest prevalence at 72% (95% CI 60–83%; I2 = 97%). When all subgroups were combined, the overall pooled prevalence was 29% (95% CI 10–52%; I2 = 100%, p < 0.001), confirming pronounced between-study variability. These patterns highlight the differing treatment complexity across tumor types, which likely contributes to the wide heterogeneity observed in polypharmacy prevalence.

Fig. 4.

Fig. 4

Open in a new tab

A forest plot of the prevalence of polypharmacy among cancer patients across reviewed studies using the quality effects model by type of Cancer. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals

Prevalence estimates by geographical region

Figure 5 illustrates the pooled prevalence of polypharmacy among cancer patients stratified by geographical region using the quality-effects model. The overall pooled prevalence was 29% (95% CI 10–52%), again indicating high between-study variability (I2 = 100%, p < 0.001). Marked regional differences were observed. Studies conducted in Europe demonstrated the highest pooled prevalence at 63% (95% CI 38–85%; I2 = 99%), followed by studies from the Americas with 32% (95% CI 1–79%; I2 = 100%), while those from Asia reported the lowest pooled prevalence of 22% (95% CI 1–61%; I2 = 98%). These findings reveal wide regional variability, likely reflecting differences in prescribing culture, cancer treatment protocols, medication availability, and healthcare system organization. Despite the stratification, heterogeneity within each region remained considerable (I2 > 98%), highlighting the influence of study-level factors and the need for region-specific strategies to optimize pharmacotherapy in oncology practice.

Fig. 5.

Fig. 5

Open in a new tab

A forest plot of the prevalence of polypharmacy among cancer patients across reviewed studies using the quality effects model by geographical region. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals

Prevalence of various polypharmacy phenogroups (potentially inappropriate medications)

Supplementary Fig. 4 depicts the pooled prevalence of potentially inappropriate medications (PIMs) among cancer patients using the quality-effects model. The overall pooled prevalence was 39% (95% CI 29–50%), indicating that roughly two out of five patients with cancer were prescribed at least one inappropriate medication. Between-study heterogeneity remained substantial (I2 97%, Q = 365.97, p< 0.001) reflecting methodological and population differences across studies. Individual study estimates ranged from 29 to 73% with the study byLund 2018 contributing the largest weight (70.6%) and reporting a prevalence of 37% (95% CI 36–38%), closely aligning with the pooled estimate. The double-arcsine transformation used in the quality-effects model stabilized variances, avoiding the need for continuity correction even when proportions approached boundary values.

Bias diagnostics and influence analyses of PIMs outcomes

Given the high weight ofLund 2018 (70.6%) and the major asymmetry in the Doi plot (LFK = 7.24) additional bias and influence analyses were undertaken. Leave-one-out analysis showed that exclusion ofLund 2018 increased the pooled PIM prevalence from 39% to approximately 44%, with no reversal of effect direction. Heterogeneity remained substantial (I2 > 95%), indicating that the dominance of this large study primarily affected the precision rather than the direction of the pooled estimate. Trim-and-fill procedures suggested the presence of missing studies on the left of the funnel plot, imputing three studies and yielding an adjusted pooled prevalence of approximately 33–35%, slightly lower than the observed estimate. A Vevea–Hedges selection model produced similar adjusted values (34–38%), confirming the presence of small-study effects but demonstrating that PIMs prevalence remained consistently high after correction.

Sensitivity analysis restricted to studies using the Beers criteria showed a somewhat lower pooled prevalence (≈ 0.31–0.36) and reduced heterogeneity, indicating that definitional inconsistency contributes to the observed dispersion. Collectively, these analyses confirm that although small-study effects and weight domination are present, the overall conclusion that PIM use is common among patients with cancer remains robust.

Prevalence of downstream consequences of polypharmacy (drug-drug interactions)

Supplementary Fig. 5 shows the pooled prevalence of drug–drug interactions (DDIs) among patients with cancer, analyzed using the quality-effects model. The overall pooled prevalence of DDIs was 51% (95% CI 37–64%), indicating that approximately one in two cancer patients exposed to polypharmacy experienced at least one potential interaction. Between-study heterogeneity remained substantial (I2 = 97%, Q = 255.80, p < 0.001), reflecting possible differences in DDI detection methods, cancer types, and study populations. Reported prevalence estimates ranged from 25 to 70%. The study by Escudero-Vilaplana 2020, which contributed the largest weight (28.5%), reported a prevalence of 51% (95% CI 48–54%), closely aligning with the pooled estimate.

Publication bias and small study effects

We assessed small-study effects and publication bias using both funnel and Doi plots (SUPP 6 and 7, respectively), with effect sizes transformed via the double arcsine method to stabilize variance across studies reporting extreme prevalence values. The funnel plot showed clear asymmetry, with an overrepresentation of smaller studies on the right-hand side, indicating that studies with higher prevalence estimates are more likely to be published or included. The Doi plot further corroborated this asymmetry, displaying substantial distortion away from the center axis. The LFK index was 7.24, which far exceeds the threshold of ± 2.0 and indicates major asymmetry. This finding is consistent with the presence of strong small-study effects or publication bias, suggesting that the pooled prevalence estimates may be influenced by selective reporting, methodological heterogeneity, or other unmeasured factors. Given this major asymmetry and the high between-study heterogeneity (I2 = 100%), the pooled prevalence estimates should be interpreted with caution.

Discussion

To our knowledge, this systematic review and meta-analysis is the first exhaustive pooled exploration of the prevalence of polypharmacy and its geographical trends as well as factors closely associated with its variability among patients with cancer diagnosis. We reviewed the longitudinal data of 102,100 patients from 12 countries on 3 continents. The overall pooled prevalence of PP among patients with cancer was 29%, with a proportionately higher prevalence in Europe than in America and Asia. Despite the apparent disparity in the definition of PP among the included studies, our period prevalence estimate provides the first attempt at exploring the burden of this growing prescribing practice associated with adverse outcomes in this cohort of patients.

The prevalence estimates reported in this review are substantially greater than those reported in the general population, including cohorts at high risk of experiencing PP, such as elderly individuals [33], patients with chronic liver disease [34], and people living with HIV [35]. However, the estimates were comparable to those of other populations that are also known to be more prone to PP, such as heart failure patients [36]. This remarkably high prevalence is concerning, particularly in cancer patients, as they represent a more challenging and therapeutically vulnerable population, principally due to the higher number of comorbidities and geriatric syndromes, which also increase the risk of PIMs use and DDIs [26]. In cancer patient population, studies has shown that PP is associated with PIM use and an increase in all-cause mortality, highlighting its significant impact on morbidity and mortality [17, 37]. When a PIM prevalence estimate of 39% is considered alongside the pooled prevalence of PP (29%, 95% CI: 10–52%), it will suggests that a substantial proportion of the medications prescribed to cancer patients may be inappropriate, potentially exposing them to preventable medication-related harm. The observation that the prevalence of potentially inappropriate medications exceeds that of overall polypharmacy likely reflects the intensity and clinical complexity of cancer care, as well as the limitations of applying generic prescribing tools—such as the STOPP and Beers criteria—to oncology populations. Although these criteria are well validated and widely used, emerging evidence from disease-specific contexts suggests that their applicability may be limited in certain complex conditions, including cancer and heart failure, where guideline-driven and context-dependent prescribing is common. The pooled prevalence of polypharmacy observed among patients with cancer in this review is higher than estimates reported in community-dwelling, non-cancer adult populations, where prevalence typically ranges between 20 and 40%. This difference likely reflects the complex clinical profile of oncology populations, in whom pre-existing multimorbidity is compounded by cancer-specific therapies and supportive medications. At diagnosis, many patients already require pharmacological management for chronic conditions such as cardiovascular disease, diabetes, or chronic respiratory illness. The subsequent addition of anticancer therapies—including chemotherapy, targeted agents, immunotherapy, and hormonal treatments—further increases medication burden. Supportive medications used to prevent or manage treatment-related toxicities (e.g., antiemetics, corticosteroids, proton pump inhibitors, and bone-modifying agents) contribute additional layers of prescribing, collectively resulting in higher polypharmacy prevalence compared with non-cancer populations.

Secondly, the trajectory of cancer care involves frequent transitions between curative, palliative, and supportive phases, each of which requires evolving pharmacological strategies. In palliative settings, symptom control often necessitates additional agents such as opioids, laxatives, and psychotropic agents. This dynamic layering of medications over time distinguishes oncology populations from the general population, where treatment regimens are usually more stable with negligible temporal evolution (Tables 1, 2, 3, and 4).

Table 1.

Characteristics of the included studies

Author, year of publication Type of study Country Setting Sample size Number of pp (%) Males (%) Mean age in years (± SD) Types of cancer Definition of pp
Ramsdale 2022 [26] RCT USA Community 718 440 (61.3%) 405 (56.4%) 77.6 Solid & hematological ≥ 5 medications
Lavan 2021 [15] Observational Ireland Inpatient 186 113 (60.8%) 100 (53.8%) 72.2 (5.8) Solid & hematological ≥ 6 medications
Leger 2017 [11] Observational France Outpatient 122 92 (75.4%) 55 (45.1) 81.54 (4.48) Hematological ≥ 5 medications
Hong 2020 [19] Observational Korea Inpatient 301 136 (45.1%) 208 (69.1%) 75 (70–93)a Solid ≥ 5 medications
VAN LOVEREN 2021 [16] Observational The Netherlands Inpatient 150 91 (61%) 88 (59%) 75 (65–90)a Solid & hematological ≥ 5 medications
Ramsdale 2018 [14] RCT USA Outpatient 40 37 (93%) 22 (55%) 77 (70–89)b Solid & hematological ≥ 5 medications
Lavan 2019 [27] Observational Ireland Inpatient 350 165 (47%) 167 (47.7%) 63.3 (12.1) Solid ≥ 6 medications
ANDREA C.Betts 2023 [18] Cross-sectional USA Not reported 601 189 (31.5%) 121 (20%) (18–39)c Solid & hematological ≥ 5 medications
HÉLÈNEPluchart 2023 [21] Retrospective cohort France Inpatient 633 154 (24.3%) 540 (71%) 66 (58–73%)a Lung cancer ≥ 5 medications
DONG-WOOChoi 2023 [25] Retrospective cohort Korea Inpatient 55,228 11,773 (21%) 30,431 (55%) ≥ 65 Colorectal cancer ≥ 5 medications
AA LABAN 2021 [20] Observational Turkey Inpatient 152 64 (42.1%) 54 (35.52%) ≥ 18 Solid & hematological ≥ 5 medications
Ortland 2022 [12] Observational Germany Inpatient 136 71 (52.2%) 68 (50%) 77 (4.53) Solid & hematological ≥ 5 medications
ESCUDERO-VILAPLANA 2020 [29] Cross-sectional Spain Outpatient 881 630 (71.5%) 495 (56.2%) 67.8 (22.5–94.4)a Solid & hematological ≥ 5 medications
M.Castro-Manzanares 2019 [13] Observational Spain Outpatient 273 203 (74.3%) 159 (58.2%) 61 (18)d Solid & hematological ≥ 4 medications
IVAN KRECAK 2023 [10] Observational Croatia Community 124 76 (61.3%) 48 (38.7%) 70 (21–92)a Hematological ≥ 5 medications
TAIKIHakozaki 2020 [23] Retrospective cohort Japan Outpatient 157 94 (59.9%) 100 (63.7%) 73 (65–88)a Lung cancer ≥ 5 medications
TAIKIHakozaki 2020 [23] Retrospective cohort Japan Inpatient 232 89 (38.4%) 60 (25.9%) 73 (65–88)a Lung cancer ≥ 5 medications
LAURIEN HAM 2022 [31] Observational Netherlands Outpatient 7864 4325 (55%) 5269 (67%) 74 (9)d Lung cancer ≥ 5 medications
JENNIFER L. Lund 2018 [22] Cohort USA Outpatient 33,838 Unspecified 5850 (17.3%) 66–69: 6973 (20.6%) Breast, colon, & lung cancer Undefined
70–74: 8381 (24.8%)
75–79: 7505 (22.2%)
80–84: 6004 (17.7%)
85 + : 4975 (14.7%)
NOBORUMorikawa 2024 [24] Observational Japan Outpatient 122 60 (49.2%) 92 (75.4%) 72 (65–89)a Lung cancer ≥ 5 medications
Open in a new tab

PP polypharmacy, SD standard deviation

aMedian (range)

bMean (range)

cRange

dMedian (interquartile range)

eYears (interquartile range)

f Patients with cancer diagnosis

Table 2.

Prevalence of PIMs and DDIs

Author, Year of publication Sample size Number with PIMS Percentage Number with DDIS Percentage Exact definition of PIMS in article
Ramsdale 2022 [26] 718 206 28 490 0.7 Beers & STOPP criteria
Lavan 2021 [15] 186 136 0.8 94 0.8 STOPP & OncPal criteria
Leger 2017 [11] 122 42 0.8 87 0.8 Criteria not defined
Hong 2020 [19] 301 137 0.7 92 0.7 Beers criteria
VAN LOVEREN 2021 [16] 150 98 0.9 NA NA STOPP criteria
Ortland 2022 [12] 128 72 0.9 42 0.9 Criteria not defined
IVAN KRECAK 2023 [10] 124 46 1 77 1 Criteria not defined
M.Castro-Manzanares 2019 [13] 273 NA NA 73 0.8 NA
TAIKIHakozaki 2020 [23] 157 60 38.2 NA NA STOPP criteria
TAIKIHakozaki 2020 [23] 232 74 31.9 NA NA STOPP criteria
LAURIEN HAM 2022 [31] 7,864 3,539 45 NA NA OncPal criteria
JENNIFER L.Lund 2017 33,838 13,028 38.5 1–31%c NA Beers criteria
Open in a new tab

PIMs potentially inappropriate medications, DDIs drug–drug interactions

aBeers criteria

bSTOPP criteria

cRange

dPatients with cancer diagnosis

Table 3.

Pooled estimates of PIMs using different models

Models Prevalence LCI HCI Cochrane Q
Fixed Effects model 0.38 0.36 0.40 223.39
Random Effects model 0.39 0.29– 0.50 223.39
Quality Effects model 0.39 0.25 0.54 98.43
Inverse Heterogeniety model 0.39 0.22 0.57 97.42
Open in a new tab

Table 4.

Pooled estimates of DDIs using different models

Models Prevalence LCI HCI Cochrane Q
Fixed Effects Model 0.55 0.53 0.57 324.18
Random Effects Model 0.51 0.40 0.62 324.18
Quality Effects Model 0.55 0.42 0.67 324.18
Inverse Heterogeniety Model 0.55 0.41 0.68 324.18
Open in a new tab

Thirdly, oncology care is inherently multi-disciplinary, involving oncologists, surgeons, radiation oncologists, primary care physicians, and palliative specialists. This fragmentation of care may contribute to therapeutic duplication, overlapping prescriptions, and less coordinated deprescribing efforts compared to non-cancer populations.

Finally, psychological distress, anxiety, and insomnia are frequent in cancer patients, often requiring the prescription of psychotropics or sedatives. These agents are less commonly initiated in the general population at the same rate but do contribute substantially to total drug counts in oncology patient’s cohorts.

Taken together, the higher prevalence of PP among cancer patients likely reflects a cumulative effect of multimorbidity, treatment complexity, supportive care requirements, fragmented prescribing practices, and cancer-specific symptomatic needs. This underscores the urgent importance of systematic medication reviews, deprescribing frameworks, and integration of clinical pharmacists within oncology teams to mitigate risks associated with PP in this vulnerable group.

Our findings underscore the gravity of the increasing burden of polypharmacy among patients with cancer and highlight the pressing need to implement interventions that have been proposed to reduce the burden of polypharmacy in the general population. This includes comprehensive medication reviews, deprescribing algorithms, PIM screening tools (e.g., Beers criteria), and clinical pharmacist-led interventions [38–40]. Accordingly, future research should prioritize the rigorous evaluation of the effectiveness of medication-related interventions in oncology populations. In parallel, there is a need for the development of multifaceted, theory-informed interventions specifically tailored to patients with cancer. Such approaches are more likely to produce meaningful and sustainable benefits, as prior intervention strategies based on predominantly pragmatic or ad hoc frameworks have frequently demonstrated limited effectiveness or, in some cases, unfavorable outcomes [41–43]. In cancer patients, the use of oncology-specific deprescribing tools such as Oncopal should be prioritized and validated across cancer populations around the world. This tool has been proven to rationalize the medication counts of cancer patients with resultant improvement in their quality of life [44].

A higher prevalence was observed in Europe (63%, 95% CI 38%–85%) compared with the Americas (32%, 95% CI 1%–79%) and Asia (22%, 95% CI 1%-61%). The highest national prevalence was reported in the Netherlands 96% (95% CI 92%–99%), whereas the lowest was recorded in Korea (21%, 95% CI 21%–22%) (24).

The overall pooled prevalence of polypharmacy among cancer patients, across all malignancy types, was 29% (95% CI 10–52%). When stratified by tumor type, marked differences were observed. Patients with hematologic malignancies exhibited the highest prevalence at 68% (95% CI 53–81%), followed by those with mixed solid and hematologic cancers (70%, 95% CI 59–80%), whereas patients with solid tumors alone had a markedly lower prevalence of 27% (95% CI 10–46%). The lowest cancer-specific prevalence was reported among individuals with colorectal cancer, at approximately 21% (95% CI 21–22%) [25]. This disparity likely reflects differences in treatment intensity and patient characteristics. Hematologic malignancies such as lymphoma and multiple myeloma frequently occur in older adults often over 75 years of age who typically have multiple comorbidities and require complex regimens including multi-agent chemotherapy, immunotherapy, and supportive care drugs. These factors collectively increase medication burden and the probability of polypharmacy [11]. However, age may have confounded these observations, as several hematologic malignancy studies (e.g., [24, 31]) involved older populations (median age ≥ 72), whereas studies in solid tumors have included slightly younger subsets within the older adult spectrum (age thresholds ≥ 65). These subtle differences in age distribution—especially when not matched by comorbidity burden could influence the observed variability in medication use. However, our exploration (through meta-regression analyses) revealed no significant effects of age, the comorbidity index, or cancer type on polypharmacy prevalence estimates.

Both point and period prevalence are affected by a whole range of socio-demographic factors. For example, Duerden et al. reported that older adults are more prone to developing both appropriate and inappropriate polypharmacy due to multiple comorbidities and associated complex health needs [45]. From our pooled synthesis, the high prevalence in the Americas aligns with this, as these studies likely include older populations [14, 18, 26, 35]. The agreement between the our prevalence estimates and that of Duerden et al. suggests that older age groups are consistently associated with higher polypharmacy rates [45]. Additionally, differences in health systems could account for differing prevalence estimates. Mair et al.’s [46]examination of this reported that patients from higher socio-economic backgrounds often have better access to healthcare services, which may increase polypharmacy rates [47]. Our review revealed that European-based studies show a wide range of polypharmacy prevalence estimates, reflecting socio-economic disparities within Europe as a continent (“Northern” vs. “Southern Europe”) [11, 15, 16, 27]. Furthermore, this variability (seen in Europe) could be due to different healthcare funding models and socio-economic inequalities, further supporting Mair et al.’s findings [47]. Regions with universal healthcare might show different trends than those with “out-of-pocket” models. A notably lower prevalence of polypharmacy was observed in the Korean studies [25, 31] compared with most European cohorts. Several factors within Korea’s healthcare system may contribute to this pattern. First, Korean oncology care is highly centralized and protocol-driven, with national insurance-linked clinical pathways that tightly regulate medication reimbursement and limit the routine use of non-essential supportive agents. Prior studies have shown that Korea has comparatively lower rates of potentially inappropriate medication use in older adults than many Western countries, a pattern attributed to stronger formulary restrictions, fewer over-the-counter medications, and more uniform prescribing norms across institutions. Moreover, prescribing in Korean cancer centers is often concentrated within specialist oncology teams, reducing fragmentation of care and the accumulation of medications from multiple providers; a known driver of polypharmacy in European settings. Additionally, the Korean studies included in our review were hospital-based cohorts with structured medication review processes led by oncology pharmacists, which may have actively suppressed polypharmacy prevalence compared with community-based European studies where medication reconciliation is less consistently implemented. Overall, these system-level, demographic, and organizational differences likely explain the markedly lower prevalence observed in Korea and underscore the importance of reading polypharmacy estimates within the context of local healthcare structures [46, 48].

The marked disparities (with high heterogeneity) in the prevalence estimates stratified by cancer type may suggest an underlying role of other covariates such as age, and the comorbidity index, among other variables. To explore this, we have executed a meta-regression analysis to ascertain whether differences in study-level characteristics could account for the observed heterogeneity in polypharmacy prevalence among cancer patients. The covariates assessed included median age, Charlson Comorbidity Index (CCI) score, and cancer type (solid vs. hematologic). Although the coefficient for hematologic cancer type was positive (β = 0.313), it was not statistically significant (p = 0.45, 95% CI: − 0.472 to 0.193). Similarly, neither the median age (β = − 0.007, p = 0.284) nor the CCI score (β = 0.035, p = 0.411) was significantly associated with polypharmacy prevalence. These results suggest that none of the examined covariates explained the substantial heterogeneity (I2 = 100%) across studies, and that other unmeasured or study-specific factors may have contributed to this variability. Further investigations using a larger pool of studies and additional clinical or methodological variables may help clarify sources of heterogeneity.

Major asymmetry was observed in the funnel plot and was corroborated by the Doi plot. Closer inspection suggests that this asymmetry is largely driven by a subset of smaller studies reporting disproportionately high prevalence estimates with wide confidence intervals, including studies reporting PIM prevalences of 73% [12] and 65% [16]. These studies likely reflect specific clinical contexts such as inpatient oncology services or geriatric cancer clinics where medication complexity and polypharmacy risk are intrinsically higher. The combination of small sample sizes and large effect estimates likely contributed to the skewed distribution. In contrast larger studies most notablyLund 2018 which accounted for approximately 70.6% of the weight in the PIM analysis occupied a more central position in the plot and contributed substantially to the precision of the pooled estimate. Nevertheless this stabilizing influence was insufficient to fully counterbalance the upward bias introduced by smaller high-prevalence studies. Collectively this pattern is consistent with the presence of small-study effects and possible publication bias whereby studies reporting more extreme estimates are preferentially published or indexed. Alternatively the observed asymmetry may partly reflect contextual heterogeneity arising from differences in population characteristics healthcare settings or the prescribing assessment tools applied (e.g. Beers versus STOPP/START criteria) rather than methodological bias alone. These findings are consistent with a cross-sectional study byIsmail et al. (2020) which reported a high prevalence of potential drug–drug interactions among chemotherapy-treated cancer patients in Pakistan, with over two-thirds classified as major interactions. Importantly, the risk of DDIs increased markedly with higher medication burden and greater use of anticancer agents, underscoring that simple screening or pragmatic interventions alone may be insufficient. Collectively, these findings underscore the need for structured, theory-informed, and multifaceted intervention strategies that better reflect the complexity of prescribing in oncology, rather than reliance on ad hoc or single-component approaches, which have demonstrated limited or inconsistent benefit [49]. In the context of recent literature the reported prevalence represents one of the highest estimates to date and lies at the upper bound of our pooled estimate. More broadly contemporary studies from European and Asian settings have reported DDI prevalence rates ranging from approximately 45% to 73% depending on cancer type and clinical context. For exampleGholipourshahraki et al. (2023) reported a prevalence of 95% among patients with hematologic malignancies admitted to tertiary oncology centers [50], while Oliviera et al. (2024) observed DDI rates of 76.45% among older adults (≥ 65 years) with cancer receiving antineoplastic agents [51]. Notably, recent studies consistently demonstrate that major pDDIs frequently involve combinations including fluoroquinolones, corticosteroids, and QT-prolonging antiemetics. Taken together, the persistently high prevalence of clinically relevant interactions across settings highlights the need for systematic medication review processes, pharmacist-led interventions, and the integration of robust DDI screening tools into electronic prescribing systems.

Strengths and limitations

The principal strength of this synthesis lies in its novelty at estimating the overall pooled prevalence of PP among patients with cancer, the prevalence of PIMs, and the prevalence of DDIs. Additionally, the pooled estimates by type of cancer and geographical region will potentially provide additional “scaffolds” for the strengthening of PP interventional measures. The use of three major bibliographic databases, searched using established and systematic methods, strengthened the methodological rigor of the review. Several important limitations should be acknowledged. First, this review was restricted to English-language publications, which may have led to the exclusion of relevant non-English studies and introduced both language and publication bias. Second, polypharmacy and related outcomes were not the primary focus of many included studies, resulting in variable depth, inconsistent reporting, and limited availability of granular data relevant to medication burden, prescribing appropriateness, and clinical context. Third, substantial between-study heterogeneity was observed across nearly all pooled analyses, reflecting marked differences in polypharmacy definitions, patient populations, cancer types, clinical settings, and study designs. Although the application of the quality-effects meta-analytical model partially mitigates the influence of methodological quality on pooled estimates, it does not eliminate heterogeneity arising from genuine clinical and methodological diversity. Additional sources of heterogeneity likely include geographic variation in prescribing practices, differences in healthcare systems, variation in age distributions, and inconsistent use of prescribing assessment tools. Finally, the predominance of observational designs limits causal inference, and the potential impact of residual confounding and small-study effects cannot be fully excluded.

Conclusion

From this meta-analysis of studies with diverse designs, we found a high pooled prevalence of polypharmacy among patient cohorts with cancer, with marked variability across studies. Given this level of heterogeneity, future prospective and systematic studies are needed to better characterize the determinants and consequences of polypharmacy in order to guide strategies that may improve patient outcomes in these cohorts.

Supplementary Information

13643_2026_3068_MOESM1_ESM.docx (34.5KB, docx)

Supplementary Material 1: A forest plot of the prevalence of polypharmacy among cancer patients across reviewed studies using fixed effects model. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals.

13643_2026_3068_MOESM2_ESM.docx (34.5KB, docx)

Supplementary Material 2: A forest plot of the prevalence of polypharmacy among cancer patients across reviewed studies using the inverse heterogeneity model. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals.

13643_2026_3068_MOESM3_ESM.docx (614.6KB, docx)

Supplementary Material 3: Risk of bias assessment of included studies. The figure visualizes the domain-specific and overall risk of bias for each study included in the meta-analysis of polypharmacy and outcomes among patients with cancer.

13643_2026_3068_MOESM4_ESM.docx (26.4KB, docx)

Supplementary Material 4: A forest plot of the prevalence of PIM use among cancer patients across the reviewed studies using the quality effects model by geographical region. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals.

13643_2026_3068_MOESM5_ESM.docx (24.9KB, docx)

Supplementary Material 5: A forest plot of the prevalence of DDIs among cancer patients across reviewed studies using the quality effects model by geographical region. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals.

13643_2026_3068_MOESM6_ESM.docx (23.2KB, docx)

Supplementary Material 6: Funnel plot of the prevalence of polypharmacy among patients with cancer.

13643_2026_3068_MOESM7_ESM.docx (46.7KB, docx)

Supplementary Material 7: Doi plot of the prevalence of polypharmacy among patients with cancer.

Acknowledgements

Funding for the article processing charge (APC) of this manuscript was provided by the Medical Research Centre (MRC) Hamad Medical Corporation, Doha State of Qatar.

Abbreviations

PP

Polypharmacy

PIMs

Potentially inappropriate medications

DDIs

Drug-drug interactions

OTC

Over-the-counter

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

PROSPERO

Prospective Register of Systematic Reviews

ADRs

Adverse drug reactions

CCI

Charlson comorbidity index

MASTER

Methodological Standards for Epidemiological Research

SD

Standard deviation

IQR

Interquartile range

REM

Random effects model

FEM

Fixed effects model

QEM

Quality effects model

Qi

Quality index

Authors’ contributions

Conceptualization: MD. Literature search: MD and RA. Independent review and risk of bias assessment: JB, MT, RA, and MD. Data curation: JB and MT. Data analysis and synthesis: MD. Initial draft of the manuscript: JB and MT. Final manuscript: JB, MT, RA, and MD.

Funding

Funding for the article processing charge (APC) of this manuscript was provided by the Medical Research Centre (MRC) Hamad Medical Corporation, Doha State of Qatar.

Data availability

The original contributions presented in the study are included in the article/supplementary materials, and further inquiries can be directed to the corresponding author. A preprint version of this systematic review and meta-analysis was previously posted on Research Square pre-print database [52] and is available at [10.21203/rs.3.rs-6896596/v1]. This preprint has not undergone peer review. The present manuscript represents the revised, peer-reviewed version of it.

Declarations

Ethics approval and consent to participate

Not applicable.

Competing interests

The authors declare no conflict of interest.

Footnotes

"The original online version of this article was revised: the authors would like to update the Funding and Acknowledgement statements of the article".

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Mohamed M. Tawengi, Jawaher Baraka and Rafal Al Shibly contributed equally to this work.

Change history

7/11/2026

The original online version of this article was revised: authors would like to update the Funding and Acknowledgement statements of the article

References

  • 1.Pazan F, Wehling M. Polypharmacy in older adults: a narrative review of definitions, epidemiology and consequences. Eur Geriatr Med. 2021;12(3):443–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Khezrian M, McNeil CJ, Murray AD, Myint PK. An overview of prevalence, determinants and health outcomes of polypharmacy. Ther Adv Drug Saf. 2020;11:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, Shamseer L, Tetzlaff JM, Akl EA, Brennan SE, Chou R, Glanville J, Grimshaw JM, Hróbjartsson A, Lalu MM, Li T, Loder EW, Mayo-Wilson E, McDonald S, McGuinness LA, Stewart LA, Thomas J, Tricco AC, Welch VA, Whiting P, Moher D. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. [DOI] [PMC free article] [PubMed]
  • 4.Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, et al, editor(s). Cochrane Handbook for Systematic Reviews of Interventions version 6.5 (updated August 2024). Cochrane, 2024. Available from: www.cochrane.org/handbook.
  • 5.Stone JC, Glass K, Clark J, Ritskes-Hoitinga M, Munn Z, Tugwell P, et al. The MethodologicAl STandards for Epidemiological Research (MASTER) scale demonstrated a unified framework for bias assessment. J Clin Epidemiol. 2021;134:52–64. [DOI] [PubMed] [Google Scholar]
  • 6.Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. A basic introduction to fixed-effect and random-effects models for meta-analysis. Res Synth Methods. 2010;1(2):97–111. [DOI] [PubMed] [Google Scholar]
  • 7.Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557–60. Available from: https://www.bmj.com/content/327/7414/557. [DOI] [PMC free article] [PubMed]
  • 8.Quintana DS. From pre-registration to publication: a non-technical primer for conducting a meta-analysis to synthesize correlational data. Front Psychol. 2015;6:1549. Available from: https://pubmed.ncbi.nlm.nih.gov/26500598. [DOI] [PMC free article] [PubMed]
  • 9.Furuya-Kanamori L, Barendregt JJ, Doi SAR. A new improved graphical and quantitative method for detecting bias in meta-analysis. Int J Evid Based Healthc. 2018;16(4):195–203. [DOI] [PubMed] [Google Scholar]
  • 10.Krečak I, Pivac L, Lucijanić M, Skelin M. Polypharmacy, potentially inappropriate medications, and drug-to-drug interactions in patients with chronic myeloproliferative neoplasms. Biomedicines. 2023;11(5):1301 [DOI] [PMC free article] [PubMed]
  • 11.Leger DY, Moreau S, Signol N, Fargeas JB, Picat MA, Penot A, et al. Polypharmacy, potentially inappropriate medications and drug-drug interactions in geriatric patients with hematologic malignancy: observational single-center study of 122 patients. J Geriatr Oncol. 2018;9(1):60–7. 10.1016/j.jgo.2017.07.015. [DOI] [PubMed] [Google Scholar]
  • 12.Ortland I, Mendel Ott M, Kowar M, Sippel C, Ko YD, Jacobs AH, et al. Medication risks in older patients (70 +) with cancer and their association with therapy-related toxicity. BMC Geriatr. 2022;22(1):716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Castro-Manzanares M, do Pazo-Oubiña F, Rodríguez-Rincón RM. Prevalence of potential drug–drug interactions in outpatients on treatment with parenteral antineoplastic drugs. Int J Clin Pharm. 2019;41(6):1429–33. [DOI] [PubMed] [Google Scholar]
  • 14.Ramsdale E, Lemelman T, Loh KP, Flannery M, Kehoe L, Mullaney T, et al. Geriatric assessment-driven polypharmacy discussions between oncologists, older patients, and their caregivers. J Geriatr Oncol. 2018;9(5):534–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Lavan AH, O’Mahony D, O’Mahony D, Gallagher P. Potentially inappropriate medication (PIM) use and severe drug interactions (SDIs) in older adults with cancer. J Geriatr Oncol. 2021;12(6):872–80. [DOI] [PubMed] [Google Scholar]
  • 16.van Loveren FMAM, van Berlo - van de Laar IRF, Imholz ALT, van ’t Riet E, Taxis K, Jansman FGA. Prevalence and follow-up of potentially inappropriate medication and potentially omitted medication in older patients with cancer – The PIM POM study. J Geriatr Oncol. 2021;12(1):80–4. [DOI] [PubMed]
  • 17.Mohamed MR, Ramsdale E, Loh KP, Arastu A, Xu H, Obrecht S, et al. Associations of polypharmacy and inappropriate medications with adverse outcomes in older adults with cancer: a systematic review and meta-analysis. Oncologist. 2020;25(1):e94–108. 10.1634/theoncologist.2019-0406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Betts AC, Murphy CC, Shay LA, Balasubramanian BA, Markham C, Allicock M. Polypharmacy and prescription medication use in a population-based sample of adolescent and young adult cancer survivors. J Cancer Surviv. 2023;17(4):1149–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Hong S, Lee JH, Chun EK, Kim KI, Kim JW, Kim SH, et al. Polypharmacy, inappropriate medication use, and drug interactions in older korean patients with cancer receiving first-line palliative chemotherapy. Oncologist. 2020;25(3):e502–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.AA Laban, N Birand, U Chukwunyere1, A Abdi BB. Evaluation of drug‑drug interactions in cancer patients treated at a University Hospital in North Cyprus using two interaction databases. Niger J Clin Pract. 2021;24(7):1067–71. [DOI] [PubMed]
  • 21.Pluchart H, Bailly S, Chanoine S, Moro-Sibilot D, Bedouch P, Toffart AC. Impact of polypharmacy and comorbidity on survival and systemic parenteral treatment administration in a cohort of hospitalized lung-cancer patients. BMC Cancer. 2023;23(1):1–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Lund JL, Sanoff HK, Hinton SP, Muss HB, Pate V, Stürmer T. Potential medication-related problems in older breast, colon, and lung cancer patients in the United States. Cancer Epidemiol Biomarkers Prev. 2018;27(1):41–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Hakozaki T, Hosomi Y, Shimizu A, Kitadai R, Mirokuji K, Okuma Y. Polypharmacy as a prognostic factor in older patients with advanced non-small-cell lung cancer treated with anti-PD-1/PD-L1 antibody-based immunotherapy. J Cancer Res Clin Oncol. 2020;146(10):2659–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Morikawa N, Naito T, Morita M, Sekikawa M, Doshita K, Yabe M, et al. Effect of polypharmacy on the outcomes of older patients with advanced non-small-cell lung cancer treated with PD-1/PD-L1 inhibitors: a retrospective cohort study. J Geriatr Oncol. 2024;15(7):101832. [DOI] [PubMed] [Google Scholar]
  • 25.Choi DW, Kang H, Zhang HS, Jhang H, Jeong W, Park S. Association of polypharmacy with all-cause mortality and adverse events among elderly colorectal cancer survivors. Cancer. 2023;129(17):2705–16. [DOI] [PubMed] [Google Scholar]
  • 26.Ramsdale E, Mohamed M, Yu V, Otto E, Juba K, Awad H, et al. Polypharmacy, potentially inappropriate medications, and drug-drug interactions in vulnerable older adults with advanced cancer initiating cancertreatment. Oncologist. 2022;27(7):E580–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Lavan AH, O’Mahony D, Buckley M, O’Mahony D, Gallagher P. Adverse drug reactions in an oncological population: prevalence, predictability, and preventability. Oncologist. 2019;24(9):e968–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Hakozaki T, Matsuo T, Shimizu A, Ishihara Y, Hosomi Y. Polypharmacy among older advanced lung cancer patients taking EGFR tyrosine kinase inhibitors. J Geriatr Oncol. 2021;12(1):64–71. [DOI] [PubMed] [Google Scholar]
  • 29.Escudero-Vilaplana V, Collado-Borrell R, Hoyo-Muñoz A, Gimenez-Manzorro A, Calles A, Osorio S, et al. Potential drug interactions between targeted oral antineoplastic agents and concomitant medication in clinical practice. Expert Opin Drug Saf. 2020;19:1041–1048. [DOI] [PubMed]
  • 30.Kurisu K, Miyabe D, Furukawa Y, Shibayama O, Yoshiuchi K. Association between polypharmacy and the persistence of delirium: a retrospective cohort study Abstract BioPsychoSocial Medicine. 2020;14(1). 10.1186/s13030-020-00199-3. [DOI] [PMC free article] [PubMed]
  • 31.Ham L, Geijteman ECT, Aarts MJ, Kuiper JG, Kunst PWA, Raijmakers NJH, et al. Use of potentially inappropriate medication in older patients with lung cancer at the end of life. J Geriatr Oncol. 2022;13(1):53–9. 10.1016/j.jgo.2021.07.009. [DOI] [PubMed]
  • 32.Higgins JPT, Green S. Cochrane handbook for systematic reviews of interventions (Version 5.1.0). The Cochrane Collaboration; 2011.
  • 33.Veehof L, Stewart R, Haaijer-Ruskamp F, Jong BM. The development of polypharmacy. A longitudinal study. Fam Pract. 2000;17(3):261–7. [DOI] [PubMed] [Google Scholar]
  • 34.Danjuma MIMU, Naseralallah L, Ansari S, Al Shebly R, Elhams M, Alshamari M, et al. Prevalence and global trends of polypharmacy in patients with chronic liver disease: A systematic review and meta-analysis. Medicine (Baltimore). 2023;102(19):32608 [DOI] [PMC free article] [PubMed]
  • 35.Mohammed I. Danjuma , Oyelola A. Adegboye, Ahmed Aboughalia NS, Ruba Almishal, Haseeb Abdul, Mohamad Faisal Hamad Mohamed MNE, Abdulatif AlKhal, Abdelnaser Elzouki, Arwa Al-Saud MC, Bidmos and MA. Prevalence and global trends of polypharmacy among people living with HIV: a systematic review and meta-analysis. Ther Adv Drug Saf. 2022 Aug 23;13:20420986221080795 [DOI] [PMC free article] [PubMed]
  • 36.Beezer J, Al Hatrushi M, Husband A, Kurdi A, Forsyth P. Polypharmacy definition and prevalence in heart failure: a systematic review. Heart Fail Rev. 2022; 27(2): 465-492. [DOI] [PMC free article] [PubMed]
  • 37.Chen LJ, Trares K, Laetsch DC, Nguyen TNM, Brenner H, Schöttker B. Systematic review and meta-analysis on theassociations of polypharmacy and potentially inappropriate medication with adverse outcomes in older cancer patients. J Gerontol A Biol Sci Med Sci. 2021;76. [DOI] [PubMed]
  • 38.Cooper JA, Cadogan CA, Patterson SM, Kerse N, Bradley MC, Ryan C, Hughes CM. Interventions to improve the appropriate use of polypharmacy in older people: a Cochrane systematic review. BMJ Open. 2015;5(12):e009235. [DOI] [PMC free article] [PubMed]
  • 39.El-Awaisi A, Al-Shaibi S, Al-Ansari R, Naseralallah L, Awaisu A. A systematic review on the impact of pharmacist-provided services on patients’ health outcomes in Arab countries. J Clin Pharm Ther. 2022;47(7):879–96. [DOI] [PubMed] [Google Scholar]
  • 40.Isleem N, Shoshaa S, AbuGhalyoun A, Khatib M, Naseralallah LM, Ibn-Mas’ud Danjuma M, et al. Critical care tele-pharmacy services during COVID-19 pandemic: a qualitative exploration of healthcare practitioners’ perceptions. J Clin Pharm Ther. 2022;47(10):1591–9. [DOI] [PMC free article] [PubMed]
  • 41.Steinmo SH, Michie S, Fuller C, Stanley S, Stapleton C, Stone SP. Bridging the gap between pragmatic intervention design and theory: using behavioural science tools to modify an existing quality improvement programme to implement “Sepsis Six.” Implement Sci. 2016;11(1):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Michie S, Johnston M, Abraham C, Lawton R, Parker D, Walker A. Making psychological theory useful for implementing evidence based practice: a consensus approach. Qual Saf Health Care. 2005;14(1):26–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Hughes CM, Cadogan CA, Ryan CA. Development of a pharmacy practice intervention: lessons from the literature. Int J Clin Pharm. 2016;38(3):601–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Lindsay J, Dooley M, Martin J, Fay M, Kearney A, Khatun M, et al. The development and evaluation of an oncological palliative care deprescribing guideline: the “OncPal deprescribing guideline.” Support care cancer Off J Multinatl Assoc Support Care Cancer. 2015;23(1):71–8. [DOI] [PubMed] [Google Scholar]
  • 45.Duerden Martin, Avery Tony PR. Polypharmacy and medicines optimisation Making it safe and sound. The King’s Fund; 2013.
  • 46.Park EH, Jung KW, Park NJ, Kang MJ, Yun EH, Kim HJ, et al. Cancer statistics in Korea: incidence, mortality, survival, and prevalence in 2022. Cancer Res Treat. 2025;57(2):312–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Mair A, Fernandes-Llimos F, Alonso A, Harrison C, Hurding S, Kempen T, Kinnear M, Michael N, Mcintosh J, Wilson M. Polypharmacy management by 2030: a patient safety challenge. Simpathy Consortium; 2017.
  • 48.Dong JY, Ju JH, Yang YM. Analysis of the prescription trends of potentially inappropriate medications in Korean older outpatients by sex: a retrospective study using data from the health insurance review and assessment service. Medicine (Baltimore). 2023;102(34):e34818. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Ismail M, Noor S, Harram U, Haq I, Haider I, Khadim F, et al. Potential drug-drug interactions in outpatient department of a tertiary care hospital in Pakistan: a cross-sectional study. BMC Health Serv Res. 2018 t;18(1):762. [DOI] [PMC free article] [PubMed]
  • 50.Gholipourshahraki T, Aria A, Sharifi M, Moghadas A, Moghaddas A. Potential drug interactions in hospitalized hematologic cancer patients: new update with new chemotherapy regimens. J Res Pharm Pract. 2023;12(4):115–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Oliveira RF, Oliveira AI, Cruz AS, Ribeiro O, Afreixo V, Pimentel F. Polypharmacy and drug interactions in older patients with cancer receiving chemotherapy: associated factors. BMC Geriatr. 2024;24(1):557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Mohamed M Tawengi, Jawaher Baraka, Rafal Al Shebly et al. Prevalence and global factors influencing polypharmacy and its clinical phenotypes in cancer patients: a systematic review and meta-analysis, 08 August 2025, PREPRINT (Version 1) available at Research Square. 10.21203/rs.3.rs-6896596/v1.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

13643_2026_3068_MOESM1_ESM.docx (34.5KB, docx)

Supplementary Material 1: A forest plot of the prevalence of polypharmacy among cancer patients across reviewed studies using fixed effects model. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals.

13643_2026_3068_MOESM2_ESM.docx (34.5KB, docx)

Supplementary Material 2: A forest plot of the prevalence of polypharmacy among cancer patients across reviewed studies using the inverse heterogeneity model. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals.

13643_2026_3068_MOESM3_ESM.docx (614.6KB, docx)

Supplementary Material 3: Risk of bias assessment of included studies. The figure visualizes the domain-specific and overall risk of bias for each study included in the meta-analysis of polypharmacy and outcomes among patients with cancer.

13643_2026_3068_MOESM4_ESM.docx (26.4KB, docx)

Supplementary Material 4: A forest plot of the prevalence of PIM use among cancer patients across the reviewed studies using the quality effects model by geographical region. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals.

13643_2026_3068_MOESM5_ESM.docx (24.9KB, docx)

Supplementary Material 5: A forest plot of the prevalence of DDIs among cancer patients across reviewed studies using the quality effects model by geographical region. Each point represents the prevalence rate for a study, with the horizontal lines representing the 95% confidence intervals.

13643_2026_3068_MOESM6_ESM.docx (23.2KB, docx)

Supplementary Material 6: Funnel plot of the prevalence of polypharmacy among patients with cancer.

13643_2026_3068_MOESM7_ESM.docx (46.7KB, docx)

Supplementary Material 7: Doi plot of the prevalence of polypharmacy among patients with cancer.

Data Availability Statement

The original contributions presented in the study are included in the article/supplementary materials, and further inquiries can be directed to the corresponding author. A preprint version of this systematic review and meta-analysis was previously posted on Research Square pre-print database [52] and is available at [10.21203/rs.3.rs-6896596/v1]. This preprint has not undergone peer review. The present manuscript represents the revised, peer-reviewed version of it.

Articles from Systematic Reviews are provided here courtesy of BMC

RESOURCES