Pharmacist-led interventions for vascular surgery patients: a prospective study on reducing drug-related problems

Abstract

Background

Vascular surgery patients are at a high risk of polypharmacy and drug-related problems. Only a limited number of studies have explored the impact of hospital pharmacists being members of a multidisciplinary team in the care of vascular surgery patients. The clinical study (Trial Registration Number NCT04930302, 16^th June 2021) aimed to assess the impact of pharmacist-led interventions on the prevalence of drug-related problems among patients hospitalised at the vascular surgery department.

Methods

The study, conducted at a specialised hospital in Slovakia during a 1-year period, included adult patients with carotid artery disease or lower extremity artery disease, taking ≥3 medications. Medication reconciliation and medication reviews were performed by hospital pharmacists at both admission and discharge. Pharmacist-proposed interventions were documented and communicated to the physician, patients were educated about their medications upon discharge.

Results

Among our study participants (n = 105), the average number of drug-related problems at admission was 2.3 ± 2.1, significantly decreasing to 1.6 ± 1.8 at discharge (p < 0.001). The predominant drug classes associated with drug-related problems were those related to the cardiovascular system (41.9%). At admission, the most frequent drug-related problem was untreated indication (40.3%), mostly caused by the failure to prescribe statin in patients with lower extremity artery disease. The highest acceptance rate of pharmacist-led interventions was at hospital admission (66.1%). More than 50% of patients were classified as those with good understanding of their pharmacotherapy.

Conclusions

This study demonstrates that pharmacist-led interventions significantly reduce drug-related problems in vascular surgery patients during hospitalisation, contributing to patient safety and clinical outcomes.

Trial registration

ClinicalTrials.gov Trial Registration Number: NCT04930302, 16^th June 2021.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-024-12015-7.

Background

Peripheral artery disease (PAD) is a condition with high worldwide prevalence, especially in the elderly population. Risk factors, such as hypertension, dyslipidaemia, diabetes and smoking, are associated with an increased risk of PAD. Moreover, patients with PAD are at increased risk of fatal and non-fatal cardiovascular events. Given these complexities, it is beneficial to involve pharmacists in managing drug-related problems (DRPs) in patients with PAD [1, 2] Pharmacists can optimise medication regimens, educate patients, monitor for side effects and drug interactions, and collaborate with other healthcare providers to ensure comprehensive care and better outcomes [3, 4].

Pharmacological treatment

The goal of pharmacological therapy patients with PAD is to slow or stop the progression of disease and to prevent cardiovascular events. A multidisciplinary treatment approach is recommended for these patients. Beside lifestyle management, the best medical therapy includes antihypertensive, lipid-lowering, antithrombotic medications and strict glycaemic control for patients with diabetes [1, 2, 5].

Vascular surgery patients often suffer from polypharmacy due to the need for stringent cardiovascular risk control [6, 7]. Age-related physiological changes, such as reduced renal and hepatic function, heighten the risk of drug interactions, adverse drug events, and altered pharmacokinetics. These factors complicate medication regimens, increase the likelihood of nonadherence, and make it challenging to monitor for side effects. Polypharmacy-related DRPs can result in missed doses, therapy failure, and poor clinical outcomes [8, 9]. Moreover, cognitive decline and functional limitations common in this population further exacerbate the complexity of managing their treatment plans. Proper management and regular review of the medication regimen by healthcare professionals are essential to mitigate these risks and ensure effective clinical results. Pharmacists can play a pivotal role in optimising pharmacotherapy by conducting regular medication reviews, identifying and resolving DRPs, and improving adherence.

Role of pharmacists in multidisciplinary teams

Only a limited number of studies have explored the impact of pharmacists being members of a multidisciplinary team in the care of vascular surgery patients [3, 6, 7, 10]. Pharmacists were included in a multidisciplinary team providing a novel model of care called Geriatric Comanagement of older Vascular surgery patients [10]. Drug-drug interactions and DRPs in patients with vascular diseases were investigated by clinical pharmacists and surgeons to identify problematic drug groups and develop prevention strategies [7]. DRPs in hospitalised vascular surgery patients were identified and studied by pharmacists [3]. The efficiency of pharmaceutical activity across the continuum of care was demonstrated by the high acceptance of pharmacists´ interventions in the vascular surgery department [10]. The role of a hospital pharmacist has traditionally focused mainly on the procurement, purchase, receipt, control, storage and expenditure of medicines. Currently, there is an increasing focus on the proactive involvement of hospital pharmacists in optimising the pharmacotherapy of both inpatients and outpatients. This emphasis also centres on enhancing the efficacy and safety of treatments through the reduction of DRPs [11, 12].

According to the Pharmaceutical Care Network Europe Association (PCNE), DRP is defined as a problem, event or circumstance related to pharmacotherapy that affects or has the potential to affect a desired therapeutic effect. PCNE has developed a classification system to accurately identify DRPs [12].

Some common DRPs include:

• Adverse drug reactions (ADRs): unintended reactions that occur after medication administration [13–15]

• Medication errors (MEs): Any preventable error that may lead to improper use of a drug under the control of a healthcare professional or patient [14–16]

DRPs have been found to occur mainly during the transit between environments, such as admission to a hospital, between different hospitals or when transferring a patient to another department within a hospital [17]. Medication reconciliation (MedRec, the process of creating an accurate medication list at transitions of care), followed by a medication review (MedRev, a structured evaluation of a patient's medications to optimize outcomes) with a pharmaceutical intervention and finally education of the patient on proper use of drugs, are widely recognised strategies on how pharmacists can help reduce DRPs [17–21].

MedRec involves several steps

1. Obtain Best Possible Medication History (BPMH): Use all available information, such as previous medical records, transfer reports, outpatient reports, telephone calls, information from a community pharmacy, and discussions with the patient or family members [17, 20, 22–24].

2. Compare Medication Lists: Compare the BPMH with the patient’s current prescriptions, including the name, strength, dosage, and route of administration.

3. Identify and Discuss Discrepancies: Discrepancies, whether intentional or unintentional, are identified and discussed with healthcare professionals (physicians, nurses, etc.) and the patient.

4. Finalise Medication List: Create a new list of medications that the patient should actually take.

Studies show that pharmacist-led MedRec in transit across health care have reduced the number of patients with drug discrepancies by up to 66% [18]. A MedRec is required in many countries to ensure the quality of healthcare delivery [21, 25].

Another tool for reducing DRPs and optimising pharmacotherapy is MedRev. During MedRev:

• Unnecessary medications are identified.

• Dosages are adjusted.

• New medications may be added.

• Drug interactions are checked.

• The best route of administration is selected [26].

Pharmacist contributions

There is increasing evidence that the pharmacist involved in the process of optimising a patient’s pharmacotherapy and collaborating with the physician may not only identify and solve the problems [3, 6, 10, 12, 27, 28] but may also contribute to a reduction in repeat hospital visits [29] and to a reduction in patient mortality [18, 29, 30].

In Slovakia, hospital pharmacists typically do not engage in activities like MedRec, MedRev or patient education. Therefore, we aimed to assess the potential impact of pharmacist-led interventions among vascular surgery patients in our conditions.

Methods [26]

Setting and participants

The study “Optimising the pharmacotherapy of vascular surgery patients at hospital admission, at discharge and at post-discharge check-up (PHAROS)” was designed as a pilot single-centre prospective, uncontrolled study. It was conducted at the National Institute of Cardiovascular Diseases in Bratislava, Slovakia, a 280-bed specialised hospital, during a 1-year period from September 2021 to August 2022.

Inclusion criteria

• Adult patients (age ≥ 18 years at the date of admission for hospitalisation) electively admitted to the Vascular Surgery Department, with carotid artery disease or lower extremity artery disease as the cause of hospitalisation and taking at least 3 medications. Patients were included in the study consecutively after meeting the inclusion criteria; no randomization was performed.

Exclusion criteria

• Acute patients: Patients requiring urgent surgery were excluded from the study because their need for immediate surgical intervention did not allow sufficient time for them to be included in the study procedures.

• Patients unwilling to sign the informed consent form.

• Patients with any mental disorder affecting memory and recall ability.

• Patients participating in another clinical study.

Data collection

Patients’ demographic characteristics, social status, health conditions and list of medications were drawn from the hospital information system. The morbidity of patients was expressed by the Charlson Comorbidity Index, which is a method to predict mortality by classifying or weighting comorbidities [31]. Other sources of information, such as medical and nursing reports, patient interviews and interviews with primary and secondary care specialists and caregivers, were used. We used a structured process to ensure consistency and accuracy by integrating information from multiple sources, verifying it through cross-referencing, resolving discrepancies collaboratively, and maintaining up-to-date records.

Patients’ personal data were anonymised and inserted into the case report form on the online database MIA DMS.

Three trained hospital pharmacists performed MedRec, followed by a MedRev at both hospital admission and hospital discharge. Pharmaceutical interventions were suggested to physicians based on the MedRev. All comments that included a proposal from the pharmacist were recorded in the patient’s medical record and communicated to the physician. Education of the patients was performed at hospital discharge.

Our study follows Standards for QUality Improvement Reporting Excellence (SQUIRE) guidelines [32].

The complete scheme of the procedure is illustrated in Fig. 1.

Fig. 1.

Procedure scheme. BPMH, Best Possible Medication History; DRPs, drug-related problems

The study was performed without external funding and was approved by the Ethics Committee of the National Institute of Cardiovascular Diseases.

Medication reconciliation and patient understanding

The patient was routinely admitted to the planned hospitalisation by the physician in cooperation with the nurse. If the patient was over 18 years of age, takes more than three medicines, speaks and understands the Slovak language and has signed an informed consent to participate in biomedical research, the hospital pharmacist came to the patient in person and performed MedRec. This process was administered according to the High 5s Project Standard Operating Protocol for MedRec [25].

Steps in the MedRec

Completion of the BPMH

Our primary source of information were admission reports and information from the patients or their family members. The types of drugs recorded for BPMH included prescription drugs, over-the-counter drugs, nutritional supplements, herbal medicines, and regular consumption of certain foods (e.g. grapefruit). If patients were unable to attend the interview, we interviewed their caregivers or family members.

Verification and documentation of BPMH

The BPMH was verified using multiple sources such as the hospital information system, outpatient reports, and patient records.

Comparison of medication lists, identification and discussion of discrepancies

Subsequently, the BPMH list was compared with the list of drugs currently prescribed to the patient. The list included the name of the medicinal product, the strength, the dosage and the route of administration. These lists were compared, and all discrepancies were identified. The discrepancies were then discussed with physicians and the patient. The finalised BPMH was documented in the case report form.

A detailed timeline of all the steps is provided in Supplement 1.

To clarify the timing and scope of the interventions conducted during the study, Supplement 2 presents a summary of the data collection timepoints, outlining the key activities, responsible team members, and time requirements across various stages, from patient admission to discharge.

Challenges and mitigation strategies in a process of the obtaining BPMH

The most common challenges encountered during the obtaining BPMH were:

• incomplete patient information—patients often forgot to mention all medications they were taking, including over-the-counter drugs, supplements, and herbal products. They didn´t know the names or dosages of their medications.

• multiple healthcare providers—fragmented medication records due to lack of coordination and communication among healthcare providers.

• inaccurate or outdated medication records.

We addressed these challenges through a combination of structured processes (asking specific questions about each type of medication, including over-the-counter drugs and supplements) and patient education (educating patients on the importance of proper use of their medications and preparing a list of actual pharmacotherapy, including dosages and frequencies).

Evaluating patient understanding

One of the recorded parameters at admission was patient understanding of their pharmacotherapy. This was evaluated for each medication on a three-point scale:

• the patient knows/does not know the name of the drug (1/0)

• the patient knows/does not know the indication of the drug (1/0)

• the patient knows/does not know the dosage of the drug (1/0).

Medication review

The next step performed by the hospital pharmacist was the MedRev. The patient’s BPMH list was reviewed to identify DRPs, such as duplicates in therapy, dosing, drug interactions, adverse drug reactions, potentially inappropriate drugs, dosage forms or route of administration. Both MedRec and MedRev were performed at hospital admission and at hospital discharge. All DRPs were classified according to the PCNE classification system, version V9.00 from 2019 [13]. Unlike other classification systems (Cipolle and Strand [33], Hepler and Strand [34]), PCNE system was prioritised because of its comprehensive nature, international recognition, and structured methodology to effectively identify and manage DRPs. Potentially inappropriate medications were identified in elderly study participants (aged ≥ 65 years) using EU(7)-PIMs list [35].

Patient education

At hospital discharge, the hospital pharmacists met patients individually and discussed with them the management of the further pharmacotherapy using teach-back method. After explaining each medication, including its name, indication, dosage, and potential side effects, pharmacists asked patients to describe how and when they would take their medications. Pharmacist prepared a written personalised medication list for every patient, summarised all medications the patient needed to take, including dosage and schedules in a clear and easy-to-understand format. Before the study, pharmacists trained under a supervisor and then independently provided patient education once deemed competent.

Outcome measures

We assessed the frequency and type of DRPs at hospital admission and at hospital discharge. We also identified ATC (Anatomical Therapeutic Chemical Classification) [36] groups of drugs and active substances with the highest incidence of DRPs.

Primary outcome

• Change in the prevalence of DRPs at hospital admission vs. hospital discharge.

Secondary outcomes

• Acceptance rate of pharmaceutical interventions by physicians. The proportion of accepted interventions by physicians was calculated.

• Patients’ understanding of their pharmacotherapy, assessed on a three-point scale at hospital admission.

Sample size

To calculate the sample size, we planned to accept a Type I error rate with a p-value of = 0.05. The study aimed to achieve 80% power to detect a small effect size of 0.3 using the Wilcoxon signed-rank test. Based on these assumptions, the required sample size was calculated to be 94 patients. To account for potential dropouts, the sample size was increased to 120 patients. The sample size calculation was performed using G*Power software version 3.1 [37]. Similar studies published in this field had comparable sample sizes [10].

Statistical analysis

For statistical analysis, the statistical program Unistat® version 10 and MIA DMS were used. Continuous variables are expressed as the mean with the standard deviation. If relevant, the maximum and minimum values are displayed. Categorical variables are characterised as frequencies and percentages.

To compare the number of drugs, active substances and DRPs at hospital admission and hospital discharge, the paired Wilcoxon signed-rank test was used given the non-normal distribution of the differences between the two time points. Normality of the distribution was tested using the Shapiro-Wilk test. To determine the effect size of the intervention, Cohen's d was calculated [38]. The effect size was subsequently categorised as small, intermediate, or strong based on established intervals [39].

For the univariate analysis (Supplement 3), the Mann-Whitney U test was used to assess the distribution characteristics of continuous variables between two independent groups. Categorical variables between the two groups were compared using Pearson's Chi-square test. Fisher's exact test was used when cell counts were less than 5.

Patients’ understanding

Patients’ understanding of their pharmacotherapy was assessed on the three-point scale at hospital admission. The average value of this score per medication was calculated, and patients were classified as those with good understanding of pharmacotherapy (2-3 points per medication), modest (1-2 points per medication) and poor understanding of pharmacotherapy (0-1 point per medication).

The three point scale was based on previously published studies of Cline et al., Boonstra et al. and Marfo et al. who evaluated patients' knowledge based on their understanding of the drug name, dosage, duration of treatment, indication, relationship of the drug to food, duration of therapy or route of administration [40–42].

Results

Study sample characteristics

During the one-year period, 120 patients who satisfied the inclusion criteria were included. Fourteen patients were excluded from the dataset due to their premature discharge from hospitalisation caused by SARS-CoV-2 infection spreaded in the department, which led to the immediate discharge of all non-acute patients. Additionally, one patient died as a result of post-surgical complications. Among these participants, 80 patients (76.2%) were 65 years or older, and within this group, 19 (18.1%) were 75 years or older. The majority (73.3%) of participants were male. Out of the total, 29 patients (27.6%) exhibited a Body Mass Index (BMI) surpassing 30, pointing toward a notable prevalence of obesity.

In cases (8 participants) where the Education level is mentioned as “No answer”, participants’ response was not recorded (Table 1).

Table 1.

Patients´ baseline social, demographic and clinical characteristics (n = 105)

	Study population (n=105)
Age (mean, SD)	68.5 ± 7.2
Patients aged ≥ 65 years	80 (76.2%)
Patients aged ≥ 75 years	19 (18.1%)
Sex
Female	28 (26.7%)
Male	77 (73.3%)
BMI (mean, SD)	27.8 ± 4.6
Social status
Lives with family/caregiver	81 (77.1%)
Social service home, retirement home	1 (1.0%)
Lives alone	23 (21.9%)
Employment status
Employed	22 (21.0%)
Unemployed	2 (1.9%)
Pensioner	81 (77.1%)
Educational level
Elementary school	20 (19.0%)
High school	64 (61.0%)
University	13 (12.4%)
No answer	8 (7.6%)
Cause of hospitalisation
Carotid artery disease	51 (48.6%)
Lower extremity artery disease	54 (51.4%)
Charlson Comorbidity Index (mean, SD) (27)	4.5 ± 1.9
Patients with CCI < 3	14 (13.3%)
Patients with CCI 3 to 4	43 (41.0%)
Patients with CCI ≥ 5	48 (45.7%)

BMI Body Mass Index, SD Standard deviation

Univariate analysis showed that the presence of DRPs at hospital admission was not associated with patient characteristics (age, sex, BMI, social status, employment status, educational level), polypharmacy, or the Charlson Comorbidity Index (see Supplement 3).

Among the 105 patients in the sample (Table 2), the most prevalent comorbidity was hypertension (I10), affecting 88.6% of the patients. Carotid artery disease (I65) and atherosclerosis (I70) were each present in 62.9% of patients, while dyslipidaemia (E78) affected 60.0%. Almost half (48.52%) of the patients had active nicotinism or had it documented in the past, and type 2 diabetes mellitus (E11) was observed in 36.2%. This distribution underscores the high prevalence of cardiovascular and metabolic disorders within the patient population.

Table 2.

The most common patient comorbidities in the sample population (n = 105)

Category	Diagnosis	No of patients (%)
I10	Hypertension	93 (88,6%)
I65	Carotid artery disease	66 (62,9%)
I70	Atherosclerosis	66 (62,9%)
E78	Dyslipidaemia	63 (60,0%)
Z81	Nicotinism	52 (48,52%
E11	Type 2 diabetes mellitus	38 (36,2%)
I25	Chronic ischemic heart disease	29 (27,6%)
G45	Transient cerebral ischemic attacks	27 (25,7%)
M53	Vertebrogenic algic syndrome	24 (22,9%)
K76	Liver disease	17 (16,2%)
I48	Atrial fibrillation and flutter	14 (13,3%)
E66	Overweight and obesity	13 (12,4%)
N18	Chronic kidney disease	13 (12,4%)

Patients’ pharmacotherapy and DRPs

The mean number of drugs prescribed at hospital admission was 10.6 ± 3.7 (minimum 4, maximum 19). Hospitalisation led to significant reduction in the number of DRPs per patient (< 0,001), as well as drugs per patient (p < 0,01) and number of active substances per patient (p < 0,05), as detailed in Table 3. The calculated Cohen's d to estimate the effect size resulted in a value of 0.36. According to Cohen´s guidelines the effect size of the intervention is classified as intermediate [39]. Polypharmacy, defined as 10 or more drugs, occurred in 55 patients (52.4%) at admission and in 57 patients (54.3%) at hospital discharge. Among the 105 patients, a total of 64 (61.0%) patients reported regular (49.5%) or occasional (11.4%) use of self-administered medications, specifically over-the-counter drugs.

Table 3.

Medications and DRPs at hospital admission and discharge

	At admission	At discharge	Statistical significance
No. of drugs per patient (mean, SD; min - max)	10.6 ± 3.7 (4 - 19)	10.0 ± 3.9 (3 - 19)	0.004
No. of active substances per patient (mean, SD; min - max)	12.0 ± 4.3 (4 -23)	11.5 ± 4.6 (3 - 24)	0.033
No. of DRPs per patient (mean, SD; min - max)	2.3 ± 2.1 (0 - 9)	1.6 ± 1.8 (0 - 10)	<0.001

DRPs Drug-related problems, SD Standard deviation

Primary outcome: DRPs reduction

The provided histogram (Fig. 2) visualises the distribution of the number of detected DRPs among patients at the time of admission and discharge.

Fig. 2.

Distribution of DRPs among patients at admission and at discharge. The x-axis represents the number of DRPs encountered by patients. The values range from 0 to 10, indicating the count of DRPs per patient. The y-axis shows the frequency of patients corresponding to each DRP count on the x-axis. The orange bars show the distribution of the number of DRPs at the time of patient admission, while the blue bars show the number of DRPs at the time of patient discharge

Overall, 236 DRPs at admission and 163 DRPs at discharge were identified in our study. Among the cohort of study participants, 80 patients (76.2%) exhibited at least 1 DRP upon hospital admission, while 79 patients (75.2%) presented with 1 or more DRP upon discharge. Among the study cohort, a small number of cases stood out due to an unusually high number of DRPs. These cases provide valuable insights into the challenges of managing patients with complex medical conditions and polypharmacy.

For example:

• One patient was identified with 9 DRPs at admission, primarily related to treatment effectiveness and safety. This patient had an extensive medication regimen, including 19 medications, and was managing 16 comorbidities.

• Another case involved a patient with 8 DRPs at discharge, related to a complex medication plan of 16 medications and 16 comorbidities. Due to the complexity of the pharmacotherapy, the recommendations to resolve these DRPs were addressed not only to hospital physicians but also to the ambulatory care physician to ensure a comprehensive and coordinated approach to the patient’s care.

While such cases are not representative of the entire cohort, they underscore the critical role of pharmacists in optimising medication therapy for patients with significant clinical complexities. These outlier cases were included in the analysis to ensure a comprehensive understanding of the DRP landscape within the study population.

Characteristics of the DRPs are described in Table 4.

Table 4.

DRPs characteristics according to PCNE classification

DRPs No. (%)
	At admission (n=236)	At discharge (n=163)
No effect of drug treatment	3 (1.3%)	1 (0.6%)
Effect of drug treatment not optimal	75 (31.8%)	55 (33.8%)
Untreated symptoms or indication	95 (40.3%)	44 (27.0%)
Adverse drug event (possibly) occurring	39 (16.5%)	23 (14.1%)
Problem with cost-effectiveness of the treatment	3 (1.2%)	3 (1.8%)
Unnecessary drug-treatment	20 (8.5%)	37 (22.7%)
Unclear problem/complaint	1 (0.4%)	0 (0.00%)
Acceptance of Intervention proposals No. (%)
	At admission (n=236)	At discharge (n=163)
Intervention accepted	156 (66.1%)	67 (41.1%)
Intervention not accepted	47 (19.9%)	22 (13.5%)
Other	33 (14.0 %)	74 (45.4%)
Status of the DRPs No. (%)
	At admission (n=236)	At discharge (n=163)
Solved	112 (47.5%)	42 (25.8%)
Partially solved	1 (0.4%)	0 (0.0%)
Not solved	77 (32.6%)	49 (30.0%)
Not known	46 (19.5%)	72 (44.2%)
Causes No. (%)
	At admission (n=325)	At discharge (n=201)
Drug selection	52 (16.0%)	82 (40.8%)
Drug form	4 (1.2%)	4 (2.0%)
Dose selection	56 (17.2%)	30 (15.0%)
Treatment duration	0 (0.0%)	2 (1.0%)
Dispensing	1 (0.3%)	1 (0.5%)
Drug use process	0 (0.0%)	0 (0.0%)
Patient related	49 (15.1%)	23 (11.4%)
Patient transfer related	36 (11.1%)	24 (11.9%)
Other	127 (39.1%)	35 (17.4%)
The Planned Interventions No. (%)
	At admission (n=311)	At discharge (n=274)
No intervention	2 (0.6%)	1 (0.4%)
At prescriber level	234 (75.2%)	159 (58.0%)
At patient level	30 (9.7%)	32 (11.7%)
At drug level	41 (13.2%)	82 (29.9%)
Other intervention or activity	4 (1.3%)	0 (0.0%)

DRPs Drug-related problems

Medications involved

The top five drug classes contributing to all 399 DRPs, as categorised by the ATC, were predominantly associated with drugs affecting the cardiovascular system (167; 41.9%), followed by alimentary tract and metabolism (61; 15.3%), the nervous system (32; 8.0%), blood and blood forming organs (30; 7.5%) and the respiratory system (14; 3.5%). At admission, the most frequent active substance with DRPs was atorvastatin (due to untreated symptoms or indication), while at discharge it was pantoprazole (attributed to unnecessary drug treatment) (Table 5).

Table 5.

Top 5 active substances associated with DRPs with examples

ATC	DRPs (n=399) No. (%)	Most frequent DRPs with examples
Atorvastatin	42 (10.5%)	Untreated symptoms or indication Man, in his early 60s with lower extremity artery disease diagnosed 5 years ago, after multiple interventions. Patient was admitted in a state of critical limb ischemia with very high cardiovascular disease risk (arterial hypertension, diabetes, smoking) with no statin since the diagnosis of disease. Pharmacist advised initiation of statin therapy.
Pantoprazole	36 (9.0%)	Unnecessary drug-treatment Man, in his late 60s admitted for carotid artery stenosis without any history of gastrointestinal bleeding or multiple risk factors for gastrointestinal bleeding. Pantoprazole started at admission as a stress ulcer prophylaxis. Continuation of pantoprazole treatment at discharge without any indication. Pharmacist advised discontinuation of gastroprotection.
Naftidrofuryl	26 (6.1%)	Effect of drug treatment not optimal Man, in his early 74s admitted for lower extremity artery disease with claudication taking naftidrofuryl at a dose of 200 mg per day. To optimise treatment efficacy pharmacist advised to uptitrate to maximal dose 600 mg per day.
Sulodexide	11 (2.6%)	Untreated symptoms or indication Man, in his late 40s admitted for lower extremity artery disease. Patient received a new prescription for sulodexide in ambulatory care a month ago however he never started to take the drug. Patient education was performed by pharmacist.
Metformin	10 (2.3%)	Adverse drug event occurring Woman, in her early 70s admitted for carotid artery stenosis. Patient with gastrointestinal side effects of metformin immediate release tablets in history mistakenly prescribed this drug form even if she takes extended release metformin. Pharmacist noticed the physician before drug administration.
Bisoprolol	10 (2.3%)	Effect of drug treatment not optimal Woman, in her late 60s admitted for carotid artery stenosis. Patient proclaims to be taking ½ of the 10 mg tablet of bisoprolol that is mistakenly lowered ½ of the 5 mg tablet on admission. Pharmacist warned the clinician.

ATC Anatomical Therapeutic Chemical Classification, DRPs Drug-related problems

Secondary outcomes: physicians’ acceptance of pharmaceutical care recommendations

From all 399 DRPs identified, an intervention was proposed in 390 cases (97.7%), and of all the proposed interventions, the acceptance rate by physicians was 57.4%, and 39.5% of them were implemented into a patient’s therapy (Table 6).

Table 6.

Physicians´ acceptance of pharmacists´ recommendations

DRPs (n = 399)
Intervention proposed No. (%) 390 (97.7%)					Intervention not proposed No. (%) 9 (2.3%)
Intervention accepted 224 (57.4%)			Intervention not accepted 68 (17.5%)	Intervention proposed, acceptance unknown 98 (25.1%)
Intervention implemented 154 (68.8%)	Intervention not implemented 63 (28.1%)	Implementation unknown 7 (3.1%)

DRPs Drug-related problems

A total of 313 causes (Table 7) were associated with accepted interventions, while 88 causes were linked to interventions that were not accepted. The most frequent cause for accepted interventions was 'C9 (other),' accounting for 33.2% (104 cases) of the total. This category included, among other issues, all instances of medication omissions. The next most common cause was "C3 (dose selection)" at 20.4% (64 cases), followed by "C1 (drug selection)" at 16.9% (53 cases). Among the interventions not accepted, "C9 (other)" also predominated, representing 42.0% (37 cases) of the total. "C1 (drug selection)" was the second most common cause at 33.0% (29 cases), followed by "C3 (dose selection)" at 12.5% (11 cases).

Table 7.

The most and the least accepted interventions based on causes of the DRPs

Causes of DRPs	Intervention accepted (total No. of causes = 313)	Intervention not accepted (total No. of causes = 88)
C1 (drug selection)	53 (16.9%)	29 (33.0%)
C2 (drug form)	7 (2.2%)	0 (0.0%)
C3 (dose selection)	64 (20.4%)	11 (12.5%)
C4 (treatment duration)	2 (0.6%)	0 (0.0%)
C5 (dispensing)	0 (0.0%)	2 (2.3%)
C6 (drug use process)	0 (0.0%)	0 (0.0%)
C7 (patient related)	33 (10.5%)	1 (1.1%)
C8 (patient transfer related)	50 (16.0%)	8 (9.1%)
C9 (other)	104 (33.2%)	37 (42.0%)

DRPs Drug-related problems

Secondary outcomes: patient’s understanding of their pharmacotherapy

At admission, the patient’s understanding of their pharmacotherapy was assessed using a three-point scale. A majority of the patients (53; 50.5%) was classified as having a good understanding of pharmacotherapy (score 2-3 points), while 26 patients (24.8%) had average scores ranging from 1-2 and likewise 26 patients had a poor understanding of pharmacotherapy (score 0-1 point).

Discussion

Age and demographic profile of the study population

Among the participants, 80 patients (76.2%) were aged 65 years or older. This is notably higher than the 47% of patients aged 65 or older reported by Hohn et al. [3], indicating a greater representation of older adults in our sample. Furthermore, with only 17.85% of Slovakia's general population being over 65 years old [43], our study cohort is considerably older than the national average, reflecting the age profile typical of vascular surgery patients.

The majority of participants (73.3%) were male, which is consistent with findings from Schmelzer et al., who reported a 68.8% male prevalence among vascular surgery patients. This aligns with the higher incidence of vascular diseases in men [7].

Regarding obesity, 29 patients (27.6%) had a Body Mass Index (BMI) exceeding 30, highlighting a significant prevalence of obesity in our cohort. National statistics show that approximately 60% of Slovak men and 70% of Slovak women are either overweight or obese. In contrast, about one in four Slovak adults aged 18 to 64 is classified as obese (BMI > 30) [43].

Polypharmacy and DRPs in vascular surgery patients

Several studies have reported that vascular surgery patients frequently suffer from multiple comorbidities and use a substantial number of medications; thus, they are at a higher risk of experiencing DRPs [3, 7, 10, 27]. For instance, Wang et al. found that the pre-intervention group of 137 vascular surgery patients had a median of 3.0 (2.0-5.0) comorbidities and were using a median of 7.0 (5.0-9.5) medications [6]. Similarly, Hohn et al., in their study of 105 vascular surgery patients, reported an average of 8.4 ± 3.8 medications per patient [3].

In our study, we observed a mean of 4.5 ± 1.9 comorbidities and 10.6 ± 3.7 medications at admission, aligning with international trends that confirm polypharmacy is common among vascular surgery patients. Polypharmacy has been associated with lower adherence, a higher occurrence of drug interactions, more frequent hospitalisations, and increased treatment costs [44, 45]. Despite hospitalisation often being described as a driving factor for the increase in the number of prescribed medications [6, 46], we observed a decrease in the number of medications to 10.0 ± 3.9 at discharge.

Our findings regarding the number of medications align with those reported in European settings but differ from studies conducted in other regions [22, 47–49]. For example, a study from Ethiopia reported that cardiovascular patients used fewer medications, with an average of 3.43 ± 1.68 medications per patient [50]. Similarly, Phoemlap et al., in a study involving 351 emergency department patients in Thailand, reported a median of five comorbidities, with about half of the patients taking 10 or more medications [51]. These differences likely reflect variations in healthcare systems, patient demographics, and medication management practices.

Impact of pharmacist-led interventions

Until now, few data regarding the impact of pharmacist-led interventions as a part of hospitalisation in vascular surgery patients have been available [3, 7, 10]. Schmelzer et al. described a multidisciplinary collaboration at a university hospital in Germany, where clinical pharmacists and surgeons worked together. Their findings demonstrated that such collaborative care positively impacted therapeutic outcomes and improved the safety of drug therapy for patients with vascular diseases [7].

To the best of our knowledge, PHAROS is the first prospective clinical study in Slovakia to assess the impact of pharmacist-led interventions on the prevalence of drug-related problems not only in vascular surgery patients but generally. Consequently, we are unable to compare our results with those of other Slovak authors. In our study, the average number of DRPs at admission was 2.3 ± 2.1. The number and types of DRPs detected in published studies are not consistent because of the use of different classification methods. Similar studies from our region consist of data from Czech clinical pharmacists, who studied the frequency and type of their interventions over a one-year period in a 900-bed hospital. A rate of interventions was 0.2 per patient, however prescription errors were not counted [52]. Study conducted in a Czech nursing home identified 2.2 DRPs per patient [53]. Compared to similar studies focused on vascular surgery patients, Hohn et al. identified 1.3 DRPs per patient and Martínez Lopéz et al. reported 1.7 DRPs per patient. In contrast, our study reveals a higher figure [3, 10]. Consistent with other research findings, the involvement of hospital pharmacists in the management of pharmacotherapy significantly reduced this number at the time of discharge to 1.6 ± 1.8 (p < 0.001). The effect size of this intervention is intermediate (Cohen's d = 0.36), suggesting that the intervention could have a clinical impact, confirming the positive effects of pharmacist-led intervention as part of hospitalisation [6, 54, 55].

Common types of DRPs and their management

Underprescribed statins

The most common type of DRPs at hospital admission was an untreated symptom or indication. The number of this type of DRP decreased after the pharmacist interventions. The most common case of an untreated indication was the failure to prescribe statin in patients with lower extremity artery disease. Statins are recommended in all patients with lower extremity artery disease as a part of the best medical therapy [1]. We observed non-compliance with guidelines despite statins being associated with lower rates of mortality and major adverse cardiovascular and cerebrovascular events in patients with lower extremity artery disease [56]. Hohn et al. found the most frequent DRP inaccurate medication history with platelet aggregation inhibitors being the most prevalent drug class involved in DRP, while statins were the fourth most common medication involved in DRP [3]. The under-prescription of statins revealed in our study correlates with findings from various studies on patients with lower extremity artery disease [57, 58]. Singh et al. identify one of the potential clinician-related barriers to statin prescription in patients with PAD as the fear of adverse effects from statin therapy that could mimic the symptoms of the primary disease [59].

Overprescribed proton pump inhibitors

Hospitalisation led to an increase in unnecessary drug treatment, specifically prescriptions of proton pump inhibitors (PPIs) at hospital discharge in patients without a proper indication. PPIs, although being a part of anti-ulcer prophylaxis, have no place in patients with low risk for gastrointestinal bleeding after discharge from a hospital [60]. While PPIs are generally considered safe and well-tolerated, their potential adverse effects should not be overlooked. These include diarrhoea [61], hypomagnesemia [62, 63], impaired vitamin B12 absorption [64], Clostridium difficile infection [65], hip fractures [66], and pneumonia [67]. Older patients with multiple comorbidities are particularly at an increased risk for these conditions.

Despite the availability of clear deprescribing guidelines, such as those by Targownik et al., PPIs overuse remains a global issue, underscoring the need for routine audits to assess the appropriateness of their use [60]. Liu et al. found that 50% of patients were affected by inappropriately prescribed PPIs [68], and Safer noted that PPIs rank among the top four overprescribed medications in the United States [69]. Inappropriate PPI prescriptions or delays in their discontinuation can exacerbate polypharmacy, increasing the risks of non-adherence, prescribing cascades, medication errors, drug interactions, and hospitalisations [70–72].

Professional societies recommend algorithms for deprescribing PPIs, which involve their gradual discontinuation, to mitigate these risks and ensure their use is limited to appropriate indications [63, 64]. Regular evaluation of PPI prescriptions, particularly at the point of hospital discharge, is crucial to align practice with these guidelines and reduce unnecessary treatment burdens.

Role of medication reconciliation

The most common cause of DRPs at admission was unintentional omission of a drug that was regularly prescribed to the patient pre-admission, classified in our results according to PCNE classification as “Other”. This finding is in line with the results of other studies that found drug omissions to be the most frequent DRP identified at hospital admission [64, 73, 74]. Such discrepancies in medical records might result in a lack of symptoms or disease control being potentially harmful to patients. MedRec is a common and required process in many countries to ensure the quality of health care delivery. Ensuring the continuity of pharmacotherapy through MedRec was the main goal of The Joint Commission (USA) as early as 2005 and by implementing the requirement the following year [75]. Since the same year, Accreditation Canada Required Organisational Practice has required MedRec on admission, discharge from hospitalisation, transferring a patient to another hospital or department [76]. Subsequently, in 2006, the WHO launched The High 5 s Project, which aims to ensure greater patient safety in the provision of healthcare. One of the five points is MedRec [25]. Since 2013, The Australian Commission on Safety and Quality in Health Care has included MedRec as one of the quality criteria for healthcare [77]. In March 2015, MedRec became a key process in the recommendation of pharmacotherapeutic optimization published by the British NICE [26]. The European Association of Hospital Pharmacists considers MedRec to be one of the key areas that must be fulfilled to ensure high-quality hospital pharmacy services [11]. Currently, there is no legislative requirement in Slovakia to perform medication reconciliation or medication review. These results further underline the importance of the MedRec and MedRev performed by the hospital pharmacists as a part of safer transition of care.

Acceptance and integration of pharmacist recommendations

Physicians, vascular surgeons and internists who participated in the study, accepted more than 57% of the recommendations provided by the hospital pharmacists. This rate varied throughout the patient care continuum, with a higher acceptance rate observed upon admission (66.1%) compared with discharge (41.1%). The lower acceptance rate at discharge can be attributed to recommendations related to the discontinuation of long-term or chronically prescribed medications, which were typically directed towards outpatient physicians. Consequently, the status of acceptance for these recommendations remained unknown (44.2%). As there is no access for pharmacists to primary care electronic health records, we were unable to evaluate the acceptance rate of the interventions given at hospital discharge. We expect the overall acceptance rate to be higher than stated in our results. Literature data of acceptance rate differs significantly. Hohn et al. describe the acceptance of 75% while Rychlícková et al. and Martinéz López et al. state the acceptance of interventions as high as 100% [3, 10, 52].We hypothesised the differences in acceptance rates may be due to several factors. First, our study involved a more diverse range of healthcare professionals, including those who may not have been familiar with the role of pharmacists in medication management. This lack of familiarity could have led to some reluctance in accepting pharmacist recommendations. Second, our study setting may have had different clinical workflows and protocols that made it more challenging for healthcare providers to implement suggested changes. Finally, the timing of recommendations, particularly those made at discharge, may have affected acceptance rates, as these often-required coordination with outpatient care providers, whose acceptance status remained unknown.

Patient understanding and education

Despite numerous studies pointing out the importance of patient understanding for optimal clinical outcomes [78, 79], a low level of patient knowledge about their medications is encountered in clinical practice [18, 80]. However, in our study, up to 50.5% of patients were identified as having a good understanding, while only 24.7% were classified as having a poor understanding. Patients’ understanding is also closely associated with higher adherence and persistence [4, 79], which can lead to a lower incidence of detected DRPs [81]. Our results showed that at admission, up to 15.1% of detected DRPs were patient-related, meaning they often did not correctly understand the indications, dosages and purposes of their prescribed medications. Recognising that patient education has been repeatedly identified as an effective tool for improving patients’ understanding [19, 20], we provided pharmacist-led education to each patient prior to discharge, with a particular focus on newly prescribed medications.

Future directions and recommendations

Currently, the integration of modern technologies and artificial intelligence has proven to be a promising approach for optimising healthcare services. Huang et al. highlighted the profound impact of these innovations on the daily practice of clinical pharmacists [82]. Compared to artificial intelligence, pharmacists have demonstrated increased expertise not only in the field of patient medication education but also in prescription review, recognition of adverse drug reaction and evaluation of their causality. This underscores their key role in direct communication with patients and improving their health outcomes.

Barriers and strategies to implementing pharmacist-led interventions

Resistance to change

Healthcare professionals may resist pharmacist-led interventions, viewing them as an added burden. Implement targeted training that highlights the benefits and cost-effectiveness of pharmacist-led interventions, and provide evidence of improved patient outcomes.

Integration challenges

Incorporating pharmacist-led interventions into existing systems can be difficult due to workflow constraints and lack of established protocols. Developing clear protocols for pharmacist involvement and integrating these into electronic health records can demonstrate their feasibility and benefits.

Workforce shortage

The shortage of physicians and nurses exacerbates the challenge, leading to overburdened staff and limited patient care time. Utilise pharmacists for tasks like medication management and patient education to relieve some of the burden on physicians and nurses.

Implications for future research

In the PHAROS study, we conducted a pilot investigation to evaluate the role of pharmacists in optimising pharmacotherapy for vascular surgery patients. As the first prospective evaluation of pharmaceutical interventions in this patient population in Slovakia, our pilot study aimed to quantify the positive impact of pharmacist involvement on patient outcomes. Preliminary results indicate significant improvements in care when pharmacists are part of the treatment process. A randomized controlled trial analysed the effectiveness of pharmacist-led interventions at a university hospital in South Korea, demonstrating a reduction in DRPs and improved adherence in an inpatient setting [48]. Building on these promising results, our future research should confirm the positive impact of the pharmaceutical care using a prospective, randomised controlled trial design to control for potential biases. By incorporating a control group, a larger sample size, and a post-discharge follow-up, future studies should provide more definitive evidence of the benefits of pharmacist-led interventions. However, when designing such a study, we must consider both ethical and scientific objectives and balance them. These efforts will ultimately contribute to the development of best practices for integrating pharmacists into patient care teams and improving patient outcomes.

Conclusion

Our study revealed that vascular surgery patients are burdened with a high number of DRPs at hospital admission, as well as hospital discharge. Pharmaceutical care provided at hospital admission and at hospital discharge reduced the prevalence rates of DRPs in our study setting. The implementation of pharmacist-led interventions upon hospital admission and discharge, followed by patient education, might enhance overall patient safety.

Limitations

The primary limitation of our study is the absence of a control group, which arose due to ethical considerations. This limitation introduces potential biases, such as selection bias and confounding variables, which may affect the validity of our findings. Without a control group, it is challenging to definitively attribute observed outcomes solely to the intervention. The study was designed as a prospective study conducted in a single centre, which may limit the generalizability of the results to broader populations and different healthcare settings. We initially planned to include a post-discharge check-up visit at 4-8 weeks following the patient’s release. Regrettably, owing to organisational limitations, we encountered difficulties in achieving this objective, as outlined in the protocol’s amendment approved by the Ethics Committee. Therefore, we could not assess the acceptance rate of recommendations given for outpatient physicians at hospital discharge, nor could we determine the effect of pharmacist-led education. Lastly, due to regulatory obstacles regarding the access to primary care electronic health records, we were unable to evaluate the acceptance rate of the interventions given at hospital discharge. This limitation prevented us from determining the extent to which our recommendations were implemented in primary care settings, potentially impacting the overall assessment of the intervention's effectiveness.

Despite these limitations, our study provides valuable insights and highlights important areas for future research. Implementing a randomised controlled trial design in future studies would address these limitations by minimising biases and providing more robust and generalizable evidence of the intervention's effectiveness.

Abbreviations

ADR

Adverse drug reaction

ATC

Anatomical Therapeutic Chemical Classification

BMI

Body Mass Index

BPMH

Best possible medication history

DRP

Drug-related problem

ME

Medication error

MedRec

Medication reconciliation

MedRev

Medication review

PCNE

The Pharmaceutical Care Network Europe Association

PPIs

Proton pump inhibitors

SD

Standard deviation

Authors’ contributions

All authors contributed to the study conception and design. Material preparation and data collection were performed by SP, KS, KL,VS and JT. Data analysis was performed by VS, ED, MJ, TT, ZK, LM and LL. The first draft of the manuscript was prepared by SP, KS, KL, VS, JT, ZK, LM, LL and TT and was finalised by all authors. All authors read and approved the final manuscript. TT is responsible for the overall content as the guarantor.

Funding

None declared.

Data availability

Only investigators from the National Institute of Cardiovascular Diseases have access to all patients' data. They maintained an accurate and comprehensive list of biomedical research participants, along with the alphanumeric codes assigned to them, all documented in the "Participant Identification Sheet." Both the completed Participant Identification Sheet and all research data are securely stored. Anonymised personal data are entered into the case report form on the online database MIA DMS and are accessible only to investigators.

The datasets used and analysed during the study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committee of the National Institute of Cardiovascular Diseases (Approval Number 1625/21, 26th May 2021) and the ClinicalTrials.gov (Trial Registration Number NCT04930302, 16th June 2021). All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable. Written informed consent was obtained from all subjects before the initiation of any research-related activities. The consent form used was approved by the Ethics Committee and was presented in the Slovak language, ensuring participants could read and comprehend it. By signing the informed consent, participants agreed to the publication and presentation of their pseudonymized analysed data.

Patients’ personal data were anonymised to ensure confidentiality. The Participant Identification Sheet was securely stored at the study centre along with other documentation. Anonymised data were then entered into the case report form in the online database MIA DMS, which operates in Europe.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.