Prevalence and associated factors for drug-related problems and polypharmacy identified by a pharmacist-led medication review in patients enrolled in a home healthcare program

Abstract

Background

Drug-related problems (DRPs) and polypharmacy are common among older adults with multimorbidity and limited functional capacity. These issues are associated with adverse outcomes and require targeted interventions. However, evidence regarding their prevalence and associated factors in home-based primary care settings—such as the ATDOM program in Catalonia—is still limited.

Methods

Aim: To determine the prevalence and types of DRPs and polypharmacy, and to identify factors associated with their presence in a home-based elderly population.

Design: Observational cross-sectional study analysing data collected during a medication review project for ATDOM patients.

Setting: Ten primary care centres in the Camp de Tarragona Primary Care Area, Catalonia, Spain.

Participants: ATDOM patients aged ≥ 65 years undergoing pharmacological treatment.

Measures: Sociodemographic and clinical characteristics (comorbidity, cognitive and functional status), medication data, and DRPs identified through pharmacist-led clinical medication review.

Analysis: Descriptive statistics and bivariate and multivariable logistic regression analyses were conducted to examine associations between patient characteristics and the presence of DRPs and polypharmacy. DRPs were an independent variable for polypharmacy, whereas the number of drugs was an independent variable for DRPs.

Ethics: Approved by the Research Ethics Committee of the Jordi Gol Primary Care Research Institute (19/141-P), Barcelona, Spain.

Results

We studied 923 patients with a mean age of 88.7 years (SD 6.7), and 72.6% of patients were women. Most had multiple comorbidities and high medication burden (56.3% polymedicated patients). The average number of prescribed drugs was 8.5 (SD 3.7). DRPs were identified in 78% of patients, with a mean of 1.7 (SD 1.4) DRPs per patient. The most common DRP subtypes were unnecessary indications and inappropriate drugs. In multivariable models, older age, greater number of prescribed drugs, diabetes and absence of cancer were associated with DRPs. Polypharmacy was associated with younger age, greater comorbidity, complex chronic condition, having DRPs, hypertension and respiratory system diseases.

Conclusions

In the ATDOM program, patients with polypharmacy and DRPs associated with various factors are quite prevalent. These patients could be prioritized for pharmacist-led clinical medication review within multidisciplinary teams. An ongoing clinical trial will assess the effectiveness of medication review in reducing polypharmacy and DRPs.

Trial registration

ClinicalTrials.gov identifier NCT05820945 (March 21, 2023).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12877-026-07045-1.

Background

In recent decades, the average human lifespan has notably increased. For example, in Catalonia, life expectancy has reached 83.4 years [1]. This demographic shift has led to a higher prevalence of chronic diseases and multimorbidity, resulting in increased medication use and a subsequent rise in potential therapeutic risks. A drug-related problem (DRP) is an event or circumstance involving drug therapy that actually or potentially interferes with desired health outcomes [2].

Several studies, both in Spain and internationally, have reported widely varying DRP prevalence rates, ranging from 3.9% to 59% [3, 4]. This disparity could be due to differences in the definition of DRP [5–7], as well as the heterogeneity of the populations studied. DRPs have been documented in both inpatient and outpatient settings [8, 9], including home care patients.

In the home care context, multiple factors could increase the risk of DRPs: multiple comorbidities, polypharmacy, limited follow-up with healthcare professionals, the involvement of multiple health care providers, insufficient coordination among interdisciplinary teams, and inadequate patient information and education regarding their treatment [10]. A recent systematic review highlighted the significant prevalence of DRPs in community-dwelling individuals, especially those receiving home care services [11].

Although most DRPs in outpatients do not cause serious harm, they can affect treatment efficacy [12]. However, some types—such as adverse drug reactions and drug‒drug interactions—can have considerable clinical relevance, including hospitalization [13].

A clinical medication review is a critical and structured, patient-centred process aimed at optimizing pharmacotherapy and minimizing medication-related risks, especially in polymedicated populations. This review identifies DRPs and proposes interventions to improve therapeutic outcomes and support healthcare system sustainability. Ideally, this process should be conducted collaboratively by an interdisciplinary team that includes physicians, nurses, social workers, and clinical pharmacists. Pharmaceutical services play a key role by enhancing communication between healthcare professionals and supporting clinical decisions [14]. Pharmacist-led medication reviews have demonstrated efficacy in detecting, reducing, and resolving DRPs among elderly patients in both primary care and home care settings [15–17].

In Catalonia, the Domiciliary Healthcare Program (ATDOM) is responsible for managing the care of patients with chronic or acute health conditions who face challenges accessing health centres. Its main goal is to maintain patient autonomy and quality of life through coordinated, continuous care, thanks to the coordination of all health areas [18]. However, the medication regimens of ATDOM patients are not systematically reviewed at enrolment or during follow-up. Considering the high vulnerability, comorbidity, and polypharmacy in this population, implementing structured medication reviews could improve both therapeutic efficacy and safety to detect DRPs.

Despite the relevance of DRPs in home care patients, there are limited and inconsistent data on their prevalence and characteristics in this context. Similarly, the role of demographic, clinical, social and pharmacological factors in the occurrence of DRPs and polypharmacy remains unclear and controversial [19–24].

Therefore, this study aimed to determine the prevalence of DRPs and polypharmacy, characterize the types of DRPs observed in ATDOM patients, and identify associated sociodemographic, clinical, and pharmacological factors.

Methods/design

Study design

We conducted an observational, cross-sectional study analysing data from a pharmacist-led clinical medication review project for patients enrolled in the ATDOM. The study was registered on ClinicalTrials.gov (identifier NCT05820945), and details of the study design have been previously reported [25].

Study setting and participants

The study was carried out in 10 primary care centres within the Camp de Tarragona Health Region (Catalan Health Institute) in the province of Tarragona, Catalonia (Spain).

Patients seen by physicians who agreed to a pharmacist-led medication review were eligible if they received home care through the ATDOM program, were aged 65 years or older, and were under active prescription involving at least one pharmacological treatment. Patients admitted to long-term care facilities, transferred to a center in another health region, or who died before the pharmacotherapeutic review could be performed were excluded from the analysis.

Study period

Patients who underwent clinical medication review from January 2021 to March 2024 were included.

Measurements and data sources

The outcome variables were obtained through a review of the medication prescription and laboratory modules of the electronic medical records and were collected via a structured form that captures data from the electronic health records. The pharmacist identified whether there were any DRPs in the patient's medication schedule through a structured clinical review of the medication. All study-related information was anonymized and securely stored at the research site.

The workflow consisted of a clinical pharmacist at the primary care centers conducting a clinical review of medical records and established drug treatment plans for patients enrolled in the ATDOM programme. Pharmaceutical review procedures are systematised in the reference document "Rational use of medicines. Basic medication management in chronic patients: reconciliation, review, deprescribing and adherence" [26], published by the Department of Health, which is part of the primary care services portfolio of the Regional Authority of Camp de Tarragona (Catalan Health Institute). The pharmacist reviewed the instructions for each medicine in the treatment plan in terms of indication, appropriateness, efficacy/efficiency and safety. This review was based on the Beers [27], STOPP-START [28], Priscus [29] and EU-7-PIM [30] criteria. Following the review process, a meeting was held between the pharmacist and the responsible physician to discuss the DRPs detected and formulate a series of recommendations for improvement and proposals for change. The physician then conducted an overall assessment of the treatment improvement and adaptation plan and, in the exercise of their responsibility, decided to what extent and/or which points of the plan to follow. In all cases, the proposals, their justification, the expected advantages and possible disadvantages were explained and agreed upon with the patient and, if necessary, with the caregiver or responsible family member.

Study variables

Primary outcomes

– Drug-related problems: number of patients with one or more DRPs and number of DRPs per patient, as identified by the pharmacist during the clinical medication review.

In this study, a DRP was defined as any event or condition associated with drug therapy that may cause, or has caused, an undesirable effect on the patient's health. The DRPs were categorized into four domains:

1. Indication-related problems: unnecessary medication; inappropriate indication; lack of prescription.

2. Adequacy-related problems: wrong drug for the patient’s condition or profile (e.g., age, pathology); incorrect dosage (higher or lower than the recommended); inappropriate frequency of administration; pharmaceutical formulation (incorrect or not recommended); treatment duration (that is, either longer or shorter than the recommended); or inappropriate timing of administration.

3. Effectiveness/efficiency-related problems: nonadherence to treatment; failure to achieve the therapeutic goal; therapeutic objective achieved, but with a need for de-escalation of the pharmacological treatment; or use of a suboptimal therapeutic alternative (i.e., not the most effective or efficient option available).

4. Safety-related problems: history of allergy or other adverse reactions to the same or similar medication; presence of contraindications; therapeutic duplication; clinically relevant drug‒drug or drug‒food interactions; or absence of required laboratory monitoring.

– Polypharmacy: the number of drugs per patient and the number of patients meeting the criteria for polypharmacy, defined as the concurrent consumption of eight or more different medications [31]. It is based on the number of active substances consumed by the patient, since some medicines may contain two or three active ingredients.

Secondary variables and covariates

– Sociodemographic data: age, sex, primary care center, and socioeconomic status (SES) according to place of residence. SES was measured via both a continuous variable and a categorical variable, following classifications established by the Catalan Health Service that assign indicators on the basis of the population served by each primary care centre [32]:

The continuous measure used was the Composite Socioeconomic Index from the Health Quality and Evaluation Agency of Catalonia, which assigns a value between −0.2 and + 6 to each centre, with higher values indicating lower levels of socioeconomic deprivation.

The categorical measure was the Socioeconomic Level classification from the Catalan Health Service Command Notebook, which classifies primary care centers into five categories on the basis of the socioeconomic characteristics of their catchment area: A (very high), B (high), C (moderate), D (low), and E (very low). This classification is based on the proportion of residents with annual incomes above €100,000 and those with incomes below €18,000.

– Clinical:

• Physical comorbidities: chronic pathologies such as hypertension, diabetes mellitus, respiratory diseases, coronary heart disease and other types of heart disease, cerebrovascular disease, musculoskeletal system and connective tissue diseases neoplasms, liver diseases, and chronic kidney disease.

• Mental health: dementia, depression, anxiety, and psychotic disorders.

• Charlson comorbidity index: a global comorbidity index that includes 19 conditions related to various chronic diseases. Combined with patient age, it is used to estimate the risk of mortality and comorbidity at one year and ten years [33].

• Adjusted morbidity groups (AMGs): a tool used to stratify the population on the basis of levels of multimorbidity and clinical complexity. It provides strong explanatory capacity for predicting healthcare resource usage, particularly in relation to drug consumption [34].

• Barthel index: this tool assesses the degree of a patient’s dependence on performing basic activities of daily living. The score ranges from 0 (maximum dependence) to 100 (total autonomy) and is based on an ordinal scale [35].

• Pfeiffer short portable mental status questionnaire (SPMSQ): a screening tool used to detect the presence and severity of cognitive impairment. It consists of ten brief questions that examine short- and long-term memory, orientation, daily knowledge, and calculation abilities. The number of errors provides an indication of the level of cognitive decline [36, 37].

• Complex chronic patient (CCP): patients with multimorbidity, extreme fragility, or a unique clinical condition that makes disease management difficult. These patients often have frequent visits to emergency hospital services, recurrent hospital admissions, and reduced personal autonomy (temporary or permanent) and are commonly polymedicated [38, 39].

• Advanced chronic patient (ACP): defined by the Catalan Health Service as a patient with a limited life expectancy and high healthcare need, requiring palliative care and advanced care planning. ACPs typically present with severe disease progression and reduced functional capacity, necessitating individualized and anticipatory management [39].

– Pharmacological treatment:

• Pharmacotherapy plan: described using the ATC code (Anatomical, Therapeutic, Chemical Classification System), which categorizes medications into therapeutic groups on the basis of the organ or system they act upon, their pharmacological effect, therapeutic indication, and chemical structure [40].

• DRPs detected: described and categorized according to typology: indication, adequacy, effectiveness/efficiency, and safety [13].

• Treatment adherence: assessed by comparing the number of medication units dispensed by the pharmacy office with the prescribed regimen over a six-month period. Adherence was considered adequate when the proportion of days covered was greater than 80%.

Data analysis

Descriptive statistics were used to characterize the study population. Continuous variables are reported as the means and standard deviations (SDs), whereas categorical variables are presented as absolute and relative frequencies.

The presence of DRPs and polypharmacy were treated as dependent variables. Sociodemographic and clinical characteristics were considered independent variables for both outcomes. Moreover, the presence of DRPs was evaluated as a potential independent variable for polypharmacy, and the number of drugs was evaluated as a potential independent variable for the presence of DRPs.

The normality of the distributions of continuous variables was assessed via the Kolmogorov–Smirnov test. Since the variables did not follow a normal distribution, nonparametric tests were applied: Pearson’s chi-square test for associations between categorical variables and the Mann–Whitney U test for comparisons of continuous variables.

Finally, to identify factors independently associated with the presence of DRPs and polypharmacy, multivariable logistic regression was performed. Adjusted odds ratios (ORs) with corresponding 95% confidence intervals (CIs) were calculated. Only cases with complete data were included in the regression analyses. Each age group was compared with the lowest age group (the reference category) to obtain an OR for each category compared separately (dummy variable). The same principle was applied to the SES and number of medications categories. For the SES category, we recoded categories A and B as the high level, category C as the medium level, and categories D and E as the low level. These categories were grouped together because there were no patients in category A and very few in category E. Thus, the medium and high levels were compared with the low level, and an OR was obtained for each category compared separately. Regarding the number of medications category, the ORs for each category in relation to the lowest category were reported. For the CCI, the OR is interpreted for each one-unit increase. For the other variables, the presence of the variable was compared with its absence. Statistical significance was set at a p value < 0.05 (two-tailed). Database management and statistical analyses were performed via the SPSS for Windows, version 29.0 statistical package (IBM Corp., Armonk, N.Y., USA).

Results

Study population characteristics

A total of 923 patients, aged 88.7 years (SD 6.7), from 10 different primary care centers were recruited, and 72.3% were women. SES, which is based on area of residence, was measured via two methods. As a continuous variable (Health Quality and Evaluation Agency of Catalonia), the mean score was 2.8 (SD 0.7). As a categorical variable (Catalan Health Service Command Notebook), over half of the patients (53.1%) were classified into category C (moderate level), followed by 35.4% in category B (high level). A smaller proportion fell into category D (low level, 9%) and category E (very low level, 2.5%), whereas no patients were classified into category A (very high level) (Table 1).

Table 1.

Patient characteristics

Number of patients		n = 923	%
Sociodemographic variables
Age groups	65–74	55	6.0
	75–84	236	25.6
	85–94	545	59.1
	≥ 95	87	9.4
Sex	Male	256	27.7
	Female	667	72.3
Socioeconomic status according to place of residencea	B	327	35.4
	C	490	53.1
	D	83	9.0
	E	23	2.5
Clinical variables
Physical comorbidity	Hypertension	745	80.7
	Diabetes mellitus	311	33.7
	Respiratory diseases	285	30.9
	Ischemic heart disease and other cardiopathies	387	41.9
	Cerebrovascular disease	101	10.9
	Musculoskeletal system and connective tissue diseases	702	76.1
	Cancer	282	30.6
	Liver diseases	26	2.8
	Chronic kidney disease	155	16.8
Psychiatric comorbidity	Dementia	315	34.1
	Depression	164	17.8
	Anxiety	222	24.1
	Psychotic disorder	20	2.2
Charlson comorbidity indexb		7.0	2.0
Complex chronic patientc		483	52.3
Advanced chronic patientd		31	3.4
Barthel index category	Total dependence (< 20)	90	10.2
	Severe dependence (20–35)	89	10.1
	Moderate dependence (40–55)	179	20.3
	Low dependence (> 60)	485	55.1
	Autonomous (100)	38	4.3
Pfeiffer test category	Severe cognitive impairment (> 8)	139	16.3
	Moderate cognitive impairment (5–7)	146	17.1
	Mild cognitive impairment (3–4)	144	16.9
	No cognitive impairment (0–2)	424	49.7

^aSocioeconomic status classification by the Catalan Health Service, categorising health centres into five categories: according to the area of residence: A (very high), B (high), C (moderate), D (low) and E (very low)

^bData presented as a mean and standard deviation

^cComplex chronic patient: defined as a patient with multimorbidity, functional or social frailty, and/or a complex clinical profile that complicates disease management and often requires frequent use of healthcare services, including emergency and hospital care

^dAdvanced chronic patient: defined as a patient with a limited life expectancy, high clinical complexity, and palliative care needs, requiring advanced care planning and a focus on quality of life

The most common physical comorbidities were hypertension (80.7%), musculoskeletal and connective tissue disorders (76.1%), and ischaemic heart disease and other cardiopathies (41.9%). Approximately one-third of the patients had diabetes mellitus, respiratory diseases, or cancer. Cerebrovascular, liver, and kidney diseases were less common. Among the psychiatric comorbidities, dementia was the most common, followed by anxiety. The mean Charlson Comorbidity Index score was 6.9 (SD 2), indicating a high comorbidity burden (Table 1). Most patients (50.8%) had an AMG score of 4, followed by 3 (41%), also reflecting high clinical complexity.

Regarding chronic condition status, approximately half of the patients (50.8%) were classified as CCP, whereas a small percentage (3.4%) were categorized as ACP. According to the Barthel Index, 4.3% of the patients were completely autonomous; approximately half of the patients (55.1%) had mild functional dependence, followed by moderate dependence (20.3%). According to the Pfeiffer test, more than half of the patients (50.3%) presented with cognitive impairment, with an approximately equal distribution across the mild, moderate, and severe categories (17%) (Table 1).

Prescription patterns

The average number of drugs prescribed per patient was 8.5 (SD 3.7), and 56.3% of patients were polymedicated (receiving ≥ 8 chronic drugs) (Table 2). The most commonly prescribed pharmacological groups were analgesics (mainly paracetamol), diuretics, psycholeptics (anxiolytics and hypnotics—mainly benzodiazepines—and antipsychotics), drugs for gastric acid-related disorders (mainly omeprazole), antithrombotics, and agents that act on the renin‒angiotensin system (Table S1). The main pharmacological groups involved in DRPs included psycholeptics (anxiolytics and hypnotics—mainly benzodiazepines—and antipsychotics), drugs for gastric acid-related disorders (mainly omeprazole), lipid-lowering agents, antidiabetics, antithrombotics, and psychoanaleptics (antidepressants and dementia drugs) (Fig. 1A, Table S2). The pharmacological groups most frequently involved in polypharmacy were analgesics (mainly paracetamol), diuretics, gastric acid-related drugs (mainly omeprazole), psycholeptics, bronchodilators, and antithrombotics (Fig. 1B, Table S3).

Table 2.

Pharmacotherapeutic characteristics and related problems

				DRPs
		All		Yes		No
		n = 923	%	n = 725	%	n = 198	%	p-value
Number of drugs/patienta		8.5	3.7	9.0	3.7	6.8	3.3	< 0.001
Number of drugs category	< 5	111	12.0	70	9.7	41	20.7	< 0.001
	5—9	485	52.6	366	50.5	119	60.1
	10—14	261	28.3	227	31.3	34	17.2
	> 14	66	7.2	62	8.6	4	2.0
Polypharmacyb		520	56.3	450	62.1	70	35.4	< 0.001
Adherencec		820	88.8	629	86.8	191	96.5	< 0.001

Statistical tests: Pearson’s chi-squared test for categorical variables; Mann–Whitney U test for continuous variables

^aData presented as a mean and standard deviation

^bPolypharmacy defined as ≥ 8 concurrent medications

^cAdherence defined as > 80% of prescribed medication collected over the previous 6 months

Fig. 1.

The ten most frequently involved drug classes, accounting for 76% of drug-related problems (DRPs) (A) and 66% of polypharmacy cases (B)

Prevalence and types of DRPs

The mean number of drugs with some DRPs per patient was 1.7 (SD 1.4). A total of 21.3% of patients had no DRPs, 30.6% had one, and 24.3% had two (Table S4), for a total of 1,692 DRPs affecting 1,534 drugs. The most frequent DPR type was the use of unnecessary medication, followed by inappropriate medication for the patient’s age, medical condition, or comorbidities and a lack of adherence. Overall, the most frequent DRPs were related to indication, followed by efficacy/effectiveness, adequacy, and safety. Underprescribing issues—such as missing necessary prescriptions, underdosing, or shorter-than-recommended treatment durations—were less common than overprescribing (Table 3).

Table 3.

Types of drug-related problems (DRPs) according patient and DRPs level^a

		Patient level		DRP level
		n = 1348	%	n = 1692	%
Indication	Unnecessary medicine	424	46.0	609	39.7
	Medicine no indicated or appropriate for the condition being treated	31	3.4	31	2.0
	Lack of prescription for a necessary medicine	8	0.9	8	0.5
Adequacy	Inappropriate medicine for the patient owing to age, medical situation or underlying pathology	167	18.1	192	12.5
	Dose is higher than the correct or recommended dose	61	6.6	66	4.3
	Dose is lower than the correct or recommended dose	4	0.4	4	0.3
	Frequency of administration incorrect or not recommended	29	3.1	31	2.0
	Drug form incorrect or not recommended	2	0.2	2	0.1
	Treatment is longer than the correct or recommended period	89	9.6	93	6.1
	Treatment is shorter than the correct or recommended period	3	0.3	3	0.2
	Administration time incorrect or not recommended	3	0.3	3	0.2
Effectiveness/efficiency	Lack of adherence	143	15.5	215	14.0
	Therapeutic goal is not reached	12	1.3	12	0.8
	Objective achieved but the intensity of the pharmacological treatment must be reduced	107	11.6	122	8.0
	Not the most effective-efficient alternative	163	17.7	194	12.6
Safety	History of allergy or other adverse reaction to the same medication or similar	26	2.8	27	1.8
	Drug with contraindications	19	2.1	20	1.3
	Therapeutic duplication	23	2.5	25	1.6
	Drug-drug interaction	1	0.1	1	0.1
	Drug-food interaction	1	0.1	1	0.1
	Lack of analytical control	32	3.5	33	2.2

^aEach patient may present more than one DRP; categories are not mutually exclusive

Factors associated with DRPs and polypharmacy

Patients with diabetes mellitus had a greater presence of DRPs, whereas patients with a diagnosis of cancer had a lower prevalence (Table 4). The analysis of the association between cognitive impairment and the presence of DRPs was stratified by Pfeiffer score < 3 and ≥ 3, i.e., without cognitive impairment and with cognitive impairment, respectively. This analysis failed to reveal any statistically significant difference (p = 0.809). A greater number of prescribed drugs was associated with a greater number of DRPs, with statistically significant differences between drug count categories. There was a clear trend: patients in the 10–14 or > 14 drug groups had more DRPs, whereas those in the < 5 or 5–9 drug groups had fewer DRPs. DRPs were also more common in polymedicated patients (≥ 8 chronic prescription drugs). Moreover, an inverse relationship was observed between DRP presence and treatment adherence (Table 2).

Table 4.

Characteristics of patients with and without drug-related problems (DRPs)

		With DRPs		Without DRPs		p-value
		n = 725	%	n = 198	%
Sociodemographic variables
Age groups	65–74	38	5.2	17	8.6	0.375
	75–84	187	25.8	49	24.8
	85–94	431	59.5	114	57.6
	> 95	69	9.5	18	9.1
Sex	Male	200	27.6	56	28.3	0.846
	Female	525	72.4	142	71.7
Socioeconomic status according to place of residencea	B	265	36.6	62	31.3	0.461
	C	379	52.3	111	56.1
	D	62	8.6	21	10.6
	E	19	2.6	4	2.0
Clinical variables
Physical comorbidity	Hypertension	591	81.5	154	77.8	0.280
	Diabetes mellitus	269	37.1	42	21.2	0.000
	Respiratory diseases	228	31.5	57	28.8	0.528
	Ischemic heart disease and other cardiopathies	303	41.8	84	42.4	0.938
	Cerebrovascular disease	73	10.1	28	14.1	0.134
	Musculoskeletal system and connective tissue diseases	561	77.4	141	71.2	0.088
	Cancer	207	28.6	75	37.9	0.015
	Liver disease	22	3.0	4	2.0	0.602
	Chronic kidney disease	130	17.9	25	12.6	0.096
Psychiatric comorbidity	Dementia	246	33.9	69	34.9	0.875
	Depression	132	18.2	32	16.2	0.574
	Anxiety	181	25.0	41	20.7	0.251
	Psychotic disorder	12	1.7	8	4.0	0.077
Charlson comorbidity indexb		6.9	1.9	6.7	2.2	0.255
Complex chronic patientc		391	53.9	92	46.5	0.074
Advanced chronic patientd		23	3.2	8	4.0	0.705
Barthel index category	Total dependence (< 20)	69	9.5	21	10.6	0.583
	Severe dependence (20–35)	66	9.1	23	11.6
	Moderate dependence (40–55)	137	18.9	42	21.2
	Low dependence (> 60)	390	53.8	95	48.0
	Autonomous (100)	31	4.3	7	3.5
Pfeiffer test category	Severe cognitive impairment (> 8)	114	15.7	25	12.6	0.420
	Moderate cognitive impairment (5–7)	112	15.5	34	17.2
	Mild cognitive impairment (3–4)	107	14.7	37	18.7
	No cognitive impairment (0–2)	335	46.2	89	45.0

Statistical tests: Pearson’s chi-squared test for categorical variables; Mann–Whitney U test for continuous variables

^aSocioeconomic status classification by the Catalan Health Service, categorising health centres into five categories: according to the area of residence: A (very high), B (high), C (moderate), D (low) and E (very low)

^bData presented as a mean and standard deviation

^cComplex chronic patient: defined as a patient with multimorbidity, functional or social frailty, and/or a complex clinical profile that complicates disease management and often requires frequent use of healthcare services, including emergency and hospital care

^dAdvanced chronic patient: defined as a patient with a limited life expectancy, high clinical complexity, and palliative care needs, requiring advanced care planning and a focus on quality of life

Polypharmacy was associated with age, increasing especially in the 75–84 years age group and decreasing in patients aged > 95 years. It was more common among patients with hypertension, diabetes mellitus, respiratory diseases, ischaemic heart disease and other heart diseases, and musculoskeletal or connective tissue disorders. Polypharmacy was also more common in patients without dementia or with anxiety, with higher Charlson scores, and among those classified as CCP. According to the Pfeiffer test, statistically significant differences were observed, with a higher frequency of polypharmacy in patients without cognitive impairment and a lower frequency among those with severe impairment (Table 5).

Table 5.

Characteristics of patients with and without polypharmacy

		Polypharmacy
		Yesa		No		p-value
		n = 520	%	n = 403	%
Sociodemographic variables
Age groups	65–74	34	6.5	21	5.2	0.000
	75–84	159	30.6	77	19.1
	85–94	293	56.4	252	62.5
	> 95	34	6.5	53	13.2
Sex	Male	146	28.1	110	27.3	0.850
	Female	374	71.9	293	72.7
Socioeconomic status according to place of residenceb	B	186	35.8	141	35.0	0.206
	C	265	51.0	225	55.8
	D	55	10.6	28	7.0
	E	14	2.7	9	2.2
Clinical variables
Physical comorbidity	Hypertension	452	86.9	293	72.7	0.000
	Diabetes mellitus	216	41.5	95	23.6	0.000
	Respiratory diseases	194	37.3	91	22.6	0.000
	Ischemic heart disease and other cardiopathies	241	46.4	146	36.2	0.003
	Cerebrovascular disease	60	11.5	41	10.2	0.581
	Musculoskeletal system and connective tissue diseases	418	80.4	284	70.5	0.001
	Cancer	161	31.0	121	30.0	0.815
	Liver disease	16	3.1	10	2.5	0.733
	Chronic kidney disease	100	19.2	55	13.7	0.031
Psychiatric comorbidity	Dementia	157	30.2	158	39.2	0.005
	Depression	103	19.8	61	15.1	0.079
	Anxiety	147	28.3	75	18.6	0.001
	Psychotic disorder	11	2.1	9	2.2	1.000
Charlson comorbidity indexc		7.3	2.0	6.5	1.8	0.000
Complex chronic patientd		301	57.9	182	45.2	0.000
Advanced chronic patiente		15	2.9	16	4.0	0.469
Barthel index category	Total dependence (< 20)	50	9.6	40	9.9	0.509
	Severe dependence (20–35)	47	9.0	42	10.4
	Moderate dependence (40–55)	105	20.2	74	18.4
	Low dependence (> 60)	280	53.9	205	50.9
	Autonomous (100)	17	3.3	21	5.2
Pfeiffer test category	Severe cognitive impairment (> 8)	56	10.8	83	20.6	0.000
	Moderate cognitive impairment (5–7)	83	16.0	63	15.6
	Mild cognitive impairment (3–4)	81	15.6	63	15.6
	No cognitive impairment (0–2)	261	50.2	163	40.5

Statistical tests: Pearson’s chi-squared test for categorical variables; Mann–Whitney U test for continuous variables

^aPolypharmacy defined as ≥ 8 concurrent medications

^bSocioeconomic status classification by the Catalan Health Service, categorising health centres into five categories: according to the area of residence: A (very high), B (high), C (moderate), D (low) and E (very low)

^cData presented as a mean and standard deviation

^dComplex chronic patient: defined as a patient with multimorbidity, functional or social frailty, and/or a complex clinical profile that complicates disease management and often requires frequent use of healthcare services, including emergency and hospital care

^eAdvanced chronic patient: defined as a patient with a limited life expectancy, high clinical complexity, and palliative care needs, requiring advanced care planning and a focus on quality of life

In the multivariable logistic regression analysis, factors independently associated with the presence of DRPs were older age (OR 2.29, 95% CI 1.13–4.65, p = 0.022 for 85–94 vs 65–74; OR 3.06, 95% CI 1.24–7.51, p = 0.015 for > 95 vs 65–74), a greater number of drugs (OR 1.97, 95% CI 1.22–3.19, p = 0.006 for 5–9 vs < 5; OR 4.76, 95% CI 2.60–8.72, p = 0.000 for 10–14 vs < 5; OR 11.14, 95% CI 3.51–35.37, p = 0.000 for > 14 vs < 5), diagnosis of diabetes (OR 2.16, 95% CI 1.40–3.33, p = 0.000), and not having cancer (OR 0.67, 95% CI 0.47–0.97, p = 0.032). The factors associated with polypharmacy were younger age (OR 0.30, 95% CI 0.15–0.61, p = 0.001 for 85–94 vs 65–74; OR 0.18, 95% CI 0.08–0.41, p = 0.000 for > 95 vs 65–74), a higher Charlson comorbidity score (OR 1.18, 95% CI 1.07–1.30, p = 0.001), CCP (OR 1.57, 95% CI 1.16–2.14, p = 0.004), the presence of DRPs (OR 3.24, 95% CI 2.25–4.67, p = 0.000), diagnosis of hypertension (OR 2.02, 95% CI 1.38–2.95, p = 0.000), and respiratory system diseases (OR 1.87, 95% CI 1.34–2.60, p = 0.000) (Table 6).

Table 6.

Multivariable logistic regression for drug related problems (DRPs) and polypharmacy

Patient characteristic			Multivariable logistic regression
			Presence of DRPs		Polypharmacy
			OR (95% CI)	p-value	OR (95% CI)	p-value
Age groups		65–74	ref	ref	ref	ref
		75–84	1.71 (0.83–3.54)	0.149	0.61 (0.30–1.26)	0.184
		85–94	2.29 (1.13–4.65)	0.022	0.30 (0.15–0.61)	0.001
		> 95	3.06 (1.24–7.51)	0.015	0.18 (0.08–0.41)	0.000
Sex		Male	ref	ref	ref	ref
		Female	0.80 (0.54–1.19)	0.273	1.36 (0.96–1.94)	0.083
Socioeconomic status according to place of residencea		Lower	ref	ref	ref	ref
		Middle	1.30 (0.76–2.21)	0.340	0.71 (0.43–1.16)	0.175
		Upper	1.50 (0.86–2.63)	0.155	0.72 (0.43–1.20)	0.208
Charlson comorbidity index (per unit increase)			0.94 (0.84–1.04)	0.238	1.18 (1.07–1.30)	0.001
Complex chronic patientb		No	ref	ref	ref	ref
		Yes	1.15 (0.81–1.63)	0.443	1.57 (1.16–2.14)	0.004
Advanced chronic patientc		No	ref	ref	ref	ref
		Yes	0.91 (0.38–2.25)	0.850	0.73 (0.30–1.75)	0.477
Physical comorbidity	Hypertension	No	ref	ref	ref	ref
		Yes	0.95 (0.62–1.46)	0.828	2.02 (1.38–2.95)	0.000
	Diabetes mellitus	No	ref	ref	ref	ref
		Yes	2.16 (1.40–3.33)	0.000	1.30 (0.91–1.85)	0.146
	Respiratory system diseases	No	ref	ref	ref	ref
		Yes	0.93 (0.63–1.35)	0.686	1.87 (1.34–2.60)	0.000
	Ischemic heart disease and other cardiopathies	No	ref	ref	ref	ref
		Yes	0.84 (0.59–1.20)	0.328	1.27 (0.93–1.73)	0.140
	Cerebrovascular disease	No	ref	ref	ref	ref
		Yes	0.64 (0.39–1.07)	0.091	1.10 (0.67–1.78)	0.708
	Musculoskeletal system and connective tissue diseases	No	ref	ref	ref	ref
		Yes	1.28 (0.87–1.88)	0.211	1.41 (0.99–2.00)	0.056
	Cancer	No	Ref	ref	ref	ref
		Yes	0.67 (0.47–0.97)	0.032	0.89 (0.64–1.24)	0.501
	Liver disease	No	ref	ref	ref	ref
		Yes	1.38 (0.44–4.35)	0.575	0.80 (0.33–2.15)	0.719
	Chronic kidney disease	No	Ref	ref	ref	ref
		Yes	1.37 (0.82–2.27)	0.225	0.92 (0.60–1.40)	0.701
Number of drugs		< 5	ref	ref
		5–9	1.97 (1.22–3.19)	0.006
		10–14	4.76 (2.60–8.72)	0.000
		> 14	11.14 (3.51–35.37)	0.000
DRP		No			ref	ref
		Yes			3.24 (2.25–4.67)	0.000

All variables were included in multivariate logistic regression models to control for potential confounders

OR Odds Ratio, CI Confidence interval

^aSocioeconomic status classification by the Catalan Health Service, categorising health centres into five categories: according to the area of residence: A (very high), B (high), C (moderate), D (low) and E (very low)

^bComplex chronic patient: defined as a patient with multimorbidity, functional or social frailty, and/or a complex clinical profile that complicates disease management and often requires frequent use of healthcare services, including emergency and hospital care

^cAdvanced chronic patient: defined as a patient with a limited life expectancy, high clinical complexity, and palliative care needs, requiring advanced care planning and a focus on quality of life

Discussion

The present cross-sectional observational study aimed to determine the prevalence of DRPs and polypharmacy, describe the types of DRPs, and identify factors associated with their occurrence. To our knowledge, this is one of the few studies to evaluate the prevalence and associated factors of DRPs in real-world primary care settings among home-based patients receiving pharmacist-led clinical medication reviews [41, 42]. It is also the only one study focusing specifically on patients enrolled in the ATDOM program.

Study population characteristics

Given that the study population comprised patients aged 65 years or older, the mean age was 88.7 years, which is higher than that reported in other studies investigating DRPs in home care settings, regardless of age [10]. The majority of patients included were women, a finding that is consistent with previous research on similar populations, although our study revealed a greater female predominance [43, 44]. The average Charlson Comorbidity Index score was severe, and more than 90% of patients scored 3 or 4 on the AMG scale. This suggests a greater comorbidity burden than that reported in other studies in similar settings [45–47]. This may be because our study focused specifically on elderly people with multimorbidity, disability, and dependency, who are unable to attend primary care centers [48].

The most frequent physical comorbidities were hypertension, musculoskeletal and connective tissue disorders, and ischaemic and other heart diseases, findings that align with those of previous studies with similar populations. Dementia and anxiety were the most common psychiatric comorbidities, although these were not consistently reported in comparable studies [43, 49, 50].

According to the Barthel index, most patients had low to moderate functional dependence. According to the Pfeiffer test, more than half of the patients had some degree of cognitive impairment, mainly mild or moderate disability. This may indicate that patients with severe functional dependence or cognitive impairment are more likely to be institutionalized, as shown in other studies [47].

Prescription patterns

The average number of medications prescribed per patient was 8.5, similar to that found in comparable studies [43, 50] but lower than that reported in patients recently discharged from the hospital (11.9 drugs per patient) [49]. Approximately one-third of the study population (35.5%) was exposed to polypharmacy, defined as ≥ 10 chronic medications, and more than half of the patients (56.3%) were taking ≥ 8 drugs. Taking into account the most common threshold of ≥ 5 medications, more than four-fifths of patients (88%) were classified as polymedicated, which is consistent with other studies conducted with older people in Central Catalonia [47] and even geographically different settings, such as Australia [50]. A study conducted in the USA using the criterion of excessive polypharmacy reported a lower proportion (25.9%) and a lower average number of medications (6.6 per patient) [51].

The most frequently prescribed pharmacological groups were analgesics, diuretics, psycholeptics, drugs for gastric acid-related disorders, antithrombotics, and agents that act on the renin‒angiotensin system, similar to other reports [50].

The pharmacological groups mainly implicated in DRPs include psycholeptics, drugs for gastric acidity, lipid-lowering agents, antidiabetics, antithrombotics, and psychoanaleptics. Psycholeptics include anxiolytics, hypnotics and antipsychotics. Anxiolytics and hypnotics—particularly benzodiazepines—have immediate effects but should be prescribed only in the short term in patients with severe anxiety or insomnia. It is essential to prioritize nonpharmacological treatments, especially in frail or polymedicated individuals, owing to the risks of tolerance, dependence, and significant adverse effects, especially in people over 65 years of age, associated with prolonged use. Therefore, treatment efficacy and safety must be regularly assessed to avoid harmful outcomes [52].

In terms of lipid-lowering therapy, evidence of its benefit is limited for frail and elderly populations, especially for patients over 80 years of age. While secondary prevention may justify treatment, data remain scarce concerning primary prevention. In older people, statins are associated with increased risks, including myalgias, hepatic dysfunction, and new-onset diabetes mellitus [53]. For elderly patients with type 2 diabetes, glycemic targets should be adjusted according to frailty. Excessive strict glycemic control must be avoided to prevent hypoglycemia and other adverse effects. Treatment, both dietary and pharmacological, should be personalized, considering the clinical condition, renal function, cardiovascular risk, and patient preferences. Deintensification or deprescribing is recommended for frail individuals, those with limited life expectancies, or those for whom the degree of risk outweighs the degree of benefit [54].

While gastroprotection may be warranted in patients at risk for ulcers, polypharmacy alone does not justify its use. Proton pump inhibitors are not clearly indicated for patients taking corticosteroids, anticoagulants, or antiplatelet agents without additional risk factors [55].

The pharmacological groups most frequently associated with polypharmacy were analgesics, diuretics, drugs for gastric acid-related disorders, psycholeptics, bronchodilators, and antithrombotics. Paracetamol was the most widely used analgesic and was often prescribed by inertia at a dose of 1 g. However, the recommended dose in patients over 65 years of age or with hepatic or renal insufficiency is 500–650 mg every 6–8 h, with a maximum of 2 g/day. In addition, coadministration of anticonvulsants increases the risk of hepatotoxicity, and vitamin K antagonists (acenocoumarol and warfarin) may enhance the anticoagulant effect [56]. Loop diuretics should be continued only if clinically justified, for example, in patients with congestive heart failure [57]. For bronchodilators, particularly inhaled corticosteroids, indications must be regularly reassessed [58].

Prevalence and types of DRPs

DRPs were identified in 78% of the patients, with a mean of 1.7 drugs with some DRPs per patient, which was slightly lower than that reported in other studies. In a study of recently hospital-discharged patients, multiple visits were conducted, potentially allowing the detection of a greater number of DRPs [49]. On the other hand, differences in the professional conducting the review (community vs. primary care pharmacist) may also influence DRP identification [10, 47, 59]. Our results revealed that nearly one-third (30.6%) of patients had a single DRP, similar to the 31% reported in a USA study of elderly patients receiving home health care [43]. Fewer than 2% of patients presented with 6–10 DPRs.

The most common DRPs types were related to indications, followed by effectiveness/efficiency, adequacy, and safety. The relatively lower prevalence of safety-related DRPs may reflect the presence of electronic prescribing alerts, mainly for critical safety issues (e.g., drug interactions, duplicates, contraindications, allergies, and required monitoring). Additionally, drug‒food interactions are more difficult to detect without direct patient interviews.

Among DRP subtypes, unnecessary medication was the most prevalent, with a higher percentage than in studies conducted in nursing homes, settings where a stronger culture of deprescribing may exist [60]. Our findings indicate more overtreatment than in a Dutch study of home-dwelling elderly patients with polypharmacy, which reported 25.5% overtreatment and 5% excessive dosage [61]. Our proportion of unnecessary medication exceeds the combined overtreatment and overdosing found in that study. This discrepancy may be attributable to our patients being older and frailer. Cases where therapeutic targets are met but deintensification is warranted or where dosing exceeds recommendations must also be considered. Underprescribing was less common than overprescribing, which is consistent with the high rates of polypharmacy observed.

Factors associated with DRPs and polypharmacy

Our results from the multivariable logistic regression revealed that older age and a greater number of prescribed drugs were significantly associated with DRPs. These findings are consistent with some studies involving populations with extensive prescribing, both in home care and primary care settings [43, 59]. We also find associations with age in the literature; some studies have reported a greater number of DRPs in younger elderly patients (between 65 and 69 years of age) [22, 62].

While polypharmacy alone was clearly associated with DRPs, age was masked when it was considered in isolation. Diabetes mellitus was associated with a greater presence of DRP and cancer with a lower presence in the bivariate analysis and the multivariable logistic regression. The former might reflect overtreatment in diabetic patients, and the latter might reflect more conservative management in oncology patients. On the other hand, greater adherence was associated with the absence of DRP in the bivariate analysis, possibly because a lower incidence of adverse medication outcomes may enhance patient satisfaction and adherence to treatment.

With respect to polypharmacy, multivariable logistic regression revealed associations with younger age, higher Charlson comorbidity index, CCP condition the presence of DRPs, and with hypertension and respiratory system diseases. When analysed separately, polypharmacy was also more common in patients with diabetes mellitus, ischaemic and other heart diseases, musculoskeletal and connective tissue diseases, and those without dementia or anxiety. These physical comorbidities were also included in the multivariable logistic regression but were not statistically significant. Polypharmacy increased among patients aged 75–84 years and then decreased, especially among those aged ≥ 95 years. This may be due to a greater emphasis on deprescription for these older patients, focusing on symptom control rather than preventive treatments to increase survival. The same explanation could be given for patients with dementia. Whereas for other comorbidities where a higher proportion of polypharmacy is seen, according to the guidelines, without taking the patient as a whole into account, we would treat them with various medications. Higher comorbidity index scores and classification as a CCP were also associated with polypharmacy. This trend may reflect a tendency to adhere to disease-specific guidelines without considering the patient holistically, as well as an expansion of diagnostic thresholds that contributes to the medicalization of life. According to the Pfeiffer test, in the bivariate analysis polypharmacy was more common among cognitively intact patients and less common among those with severe cognitive impairment. This finding is consistent with the lowest polypharmacy found in patients with dementia and nursing home studies showing an inverse relationship between cognitive impairment and excessive polypharmacy [63].

The present study considered a number of factors, some of which were associated with DRPs or polypharmacy. However, the existence of other associated factors cannot be ruled out. In fact, two recent studies reported that DRPs and polypharmacy were also associated with living alone or with family members/caregivers who have a high burden or provide insufficient supervision, poorer economic status, lack of coordination among professionals, low health literacy of the patient/family/caregiver, limited time/expertise of home care professionals, or storage or accessibility problems in the patient's home [41, 42].

Clinical implications

Polypharmacy and DRPs represent significant public health concerns requiring proactive pharmaceutical care. The integration of clinical pharmacists into multidisciplinary teams facilitates medication review and DPR detection. Deprescribing and critical assessment of guideline adherence during these reviews can reduce DRPs and polypharmacy by tailoring treatments to individual care goals.

Limitations and strengths

The present study has several limitations and strengths. The ATDOM is specific to Catalonia, which may limit the generalisability of findings to health services in other geographical areas. Assessing adherence in a population with 34% dementia and 50% cognitive impairment is particularly important. It was measured solely through pharmacy dispensing data, which represent a possible upper limit of adherence. This indirect method should be complemented by patient-reported measures through an adherence test. Nonreimbursed medications not captured in electronic prescriptions were excluded, although they likely represented a minority. The clinical pharmacist did not conduct home visits, potentially limiting the identification of DRPs related to administration errors, nonadherence or drug-food interactions. Although the study was conducted at multiple centers, all the reviews were provided by a single clinical pharmacist. While this avoids interrater variability, it also restricts the diversity of interpretations. Causal inferences regarding risk factors for DRPs and polypharmacy should be interpreted with caution. This is a cross-sectional study, so the found associations do not necessarily imply causality.

Determining the factors associated with a higher number of pharmacotherapy-related and polypharmacy problems is indeed crucial, as it would allow us to focus our resources on the most at-risk populations. The strength of the study certainly lies in the number of factors considered. It is also important to focus on older people receiving home care, about whom we find less information in the literature. Another notable strength of this study is its foundation in real-world clinical practice, providing insights relevant to the routine care of ATDOM patients. The study was conducted over a relatively long period of three years and two months. The large sample size of 923 patients from ten different centers enhances the robustness and external validity of the findings.

Conclusion

This study revealed a high prevalence of drug-related problems (DRPs) and polypharmacy among older, home-based patients enrolled in the ATDOM programme. Most DRPs were related to unnecessary medication, suggesting a pattern of overtreatment in a population characterized by advanced age, multimorbidity, and frailty (complex and advanced chronic patients, with altered Barthel index and Pfeiffer test categories). Conversely, underprescription issues were less common, which aligns with the high rates of polypharmacy observed.

Factors significantly associated with the presence of DRPs included older age, a greater number of prescribed medications, and those patients diagnosed with Diabetes mellitus and those without a cancer diagnosis. Polypharmacy was more common in younger age, those with a higher comorbidity burden (Charlson Index), those classified as having complex chronic conditions, those presenting with DRPs, and those diagnosed with hypertension and respiratory systems diseases.

These findings, which identify factors associated with a higher number of pharmacotherapy-related problems and polypharmacy, allow us to focus our resources on the most at-risk populations by prioritizing these patients for medication review in primary care. Pharmacist-led clinical reviews—particularly when integrated into multidisciplinary teams with physicians, nurses ans social workers—play a key role in identifying DRPs and guiding deprescribing, facilitating safer and more individualized pharmacotherapy tailored to patient needs, preferences, and care goals. This approach supports better health outcomes and contributes to the sustainability of health systems that care for increasingly complex elderly populations.

A randomized clinical trial based on the data from this observational study is currently underway [25]. This study will provide evidence of the effectiveness of structured medication reviews compared with usual management practice. Future research should also evaluate the impact of these interventions on clinical outcomes, quality of life, healthcare utilization, and cost-effectiveness in this high-risk population.

Abbreviations

ACP

Advanced chronic patient

AMG

Adjusted morbidity groups

ATC

Anatomical therapeutic chemical classification system

CCP

Complex chronic patient

DRP

Drug-related problem

IDIAPJGol

Jordi Gol primary care research institute

SES

Socioeconomic status

Authors’ contributions

CSG is the principal investigator who developed the original idea for the study and carried out the clinical medication reviews. CSG, EA, MG, CCP, FBR, FML and LC contributed to the study design and writing of the funding applications. CSG, MG, MR and EA developed the statistical analysis plan. CSG drafted the manuscript. All the authors critically reviewed and revised the manuscript content before approving the final version.

Funding

This study is partially funded by a predoctoral grant from the Catalan Institute of Health (Institut Català de la Salut) and IDIAPJGol under the “21 ICS Grant for Research Training and Doctoral Studies in Primary Care” (reference 7Z21/011). The funding bodies had no role in the study design; data collection, analysis, or interpretation; or writing or decision to publish the manuscript.

Data availability

The datasets used and/or analysed during this study are available from the corresponding author upon reasonable request. The findings will be shared with participating physicians, patients, and the broader medical and scientific community.

Declarations

Ethics approval and consent to participate

This study was designed and conducted in accordance with the Guide to Good Practice in Health Sciences Research of the Catalan Institute of Health (Institut Català de la Salut), the amended principles of the Declaration of Helsinki, and all applicable regulatory requirements. The responsible physician informed patients, their caregivers and/or legal representatives about the study's objectives and general procedures. Verbal informed consent was obtained from all participants. The study protocol, including the verbal informed consent procedure, was approved by the Clinical Research Ethics Committee of Fundació Institut Universitari per a la Recerca a l’Atenció Primària de Salut Jordi Gol i Gurina (IDIAPJGol, Barcelona; February 2, 2020; Approval no: 19/141-P).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.