Satisfaction of Cancer Patients Treated with Oral Anticancer Medications regarding Dispensing by Community Pharmacists: A Cross-Sectional Study

,


Introduction
Community pharmacists ensure that the quality and safety of medication dispensing are guaranteed at all times by limiting as far as possible the risks associated with an error in delivery, prescription, drug interactions or undetected contraindications, inadequate dosages, or noncompliance with treatment [1]. Te longer the patient feels satisfed with the stage of delivery of their treatment, the better their adherence to the treatment, resulting in improved compliance [2].
Traditionally, in oncology, chemotherapy treatments are administered intravenously by trained staf and are rarely managed by the patient at home. However, in recent decades, the number of oral anticancer medications (OAMs) has increased, whether for conventional cytotoxic, hormonal, or targeted therapies, representing a shift of paradigm in cancer care [3]. OAMs put the patients at the center of their own therapeutic management, and in an oncological context, patients' understanding, knowledge, adherence, and compliance have become particularly important to achieve an optimal risk-beneft ratio [4]. From the patient's perspective, OAMs raise numerous safety concerns such as adherence (nonadherence rates ranging from 0 to 54%), storage and handling specifcities, absorption infuenced by the patient's diet (half of OAMs) and gastric pH (ffth of OAMs), drug and food interactions related to the induction or inhibition of cytochrome P450 enzymes (two third of OAMs), and the prevention, detection, and treatment of adverse efects [5].
In France, most OAMs are delivered in community pharmacies, and a limited number of OAMs can be delivered by hospital pharmacists (out-patient dispensing of hospitalreserved drugs) [6]. Consequently, the roles of pharmacists in dispensing OAMs and managing cancer patients have increased in recent decades [7]. In 2020 in France, out of a total medication expenditure of 5.9 billion euros, 3.12 billion euros (52.9%) were related to the delivery of anticancer medications in pharmacies, including 2.03 billion euros (65%) for targeted therapies [8]. Despite this evolving role, little is known about the patient's satisfaction regarding their pharmaceutical management. Tis study aimed to assess the satisfaction of cancer patients treated with OAMs regarding the dispensing of the latter in community pharmacies and explore the patient-pharmacist relationship in cancer pharmaceutical management.

Study Design.
Te main objective of this multicenter and cross-sectional study was to assess the overall satisfaction of cancer patients with the quality of OAMs dispensing in community pharmacies. Te secondary objectives were to assess the relationship between patient satisfaction with (1) the quality and type of information on OAMs given by the pharmacist, general practitioner (GP), and oncologist; (2) medication adherence; (3) oncological characteristics; (4) patient characteristics; and (5) symptoms and health-related quality of life (HRQoL).
Te study protocol was designed to conform to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines for reporting observational studies [9]. Te study has been registered on the ClinicalTrials.gov Identifer NCT03961789 (May 23, 2019). Te study was anonymous, approved by the local ethics committee (No.2017/CE08, May 04, 2017, Comité de Protection des Personnes sud-est 6, IRB: 00008526), and in accordance with the Declaration of Helsinki. After being informed about the aim of the study, the participants' informed consent was obtained through the survey.

Setting.
Tis multicenter study was coordinated by the University Hospital of Clermont-Ferrand (France). All the participants were recruited through the oncology departments of hematology, urology, pulmonology, gastroenterology, and dermatology in three hospital centers (University Hospital of Clermont-Ferrand, Hospital of Le Puy-en-Velay, and Hospital of Mende, France) and from community pharmacies (fve pharmacies in Clermont-Ferrand, France, known to manage cancer patients). Tis study (patient inclusion and data collection) was conducted between June 2017 and July 2018. Cancer patients answered once to the questionnaire, with no follow-up.

Participants.
Participants could be included in the survey if they were cancer patients receiving OAMs delivered in a community pharmacy for the treatment of cancer. Exclusion criteria were defned as follows: patients who were unable to understand or respond to questionnaires and age <18 years.
Tis study was conducted as a survey by contacting each eligible patient after being fully informed by the oncologist or the pharmacist. After acceptance, the patients receive the questionnaire by post with a stamped-addressed envelope for the return of the completed questionnaire.

Variables.
Te primary outcome measured was the patient's satisfaction with the dispensing of their OAMs by a community pharmacy. Te participants responded to the question "Are you generally satisfed with the quality of dispensing of your oral anticancer treatment by a community pharmacy?" by assessing their score with a visual analogue scale (VAS; 0: not at all satisfed, 100: fully satisfed).
Te secondary outcomes were the assessment by the patient of the quality of the information dispensed by the oncologist, the pharmacist, and the GP about their OAMs with a VAS score (0: not at all satisfed, 100: fully satisfed) and the type of information dispensed (method of administration, management of adverse efects, and drug interactions). Te patient's adherence to their medication was assessed using the 8-item Morisky Medication Adherence Scale (MMAS-8) [10], to determine their adherence level. If the total score was less than 4, the patient was considered nonadherent. Tere were two types of questions in the MMAS-8: items 1, 4, 5, and 8 were questions for unintentional nonadherence and items 2, 3, 6, and 7 were for intentional nonadherence. Other information recorded was the type of cancer and their prescribed OAMs from a list including all the OAMs available in pharmacies in France. Te length in weeks since the patient has taken their medication was also requested, as well as the number of medications taken per day. Te patient's characteristics were also recorded such as age, gender, professional status, marital status, and the education level. Te symptoms and HRQoL were recorded with the QLQ-C30 questionnaire, including fve functional aspects (physical, role, emotional, cognitive, and social), eight symptoms (fatigue, nausea/ vomiting, pain, dyspnea, insomnia, appetite loss, constipation, and diarrhea), fnancial difculties, and global quality of life [11].
2.5. Data Sources/Measurement. All the data were obtained from a paper questionnaire completed by the patients. All the study data were collected and managed using REDCap electronic data capture tools (Vanderbilt University, Tennessee, US) hosted at CHU Clermont-Ferrand [12].

Study Size.
Tis cross-sectional study assessed the satisfaction of patients treated with OAMs regarding the dispensing of the latter in community pharmacies. Te sample size has been estimated to describe satisfaction scores with a satisfactory accuracy. For a standard deviation of satisfaction scores expected at 25 (out of 100) and an accuracy close to 5 (out of 100), at least 90 patients were needed for a two-sided type I error at 5% [13].

Statistical Methods.
Continuous data were expressed as the median and interquartile range. Te normality was assessed using the Shapiro-Wilk test. Patients' satisfaction for the dispensing of their OAMs was compared between independent groups (information on methods of administration, management of adverse efects, and drug interactions) using the Student's t-test or the Mann-Whitney U test when the assumptions of the t-test were not met. Te homoscedasticity of the data was assessed using the Fisher-Snedecor test. Te results were expressed using Hedge's efect-size (ES) and 95% confdence interval (95% CI) and were interpreted according to the rules of thumb reported by Cohen [14] (small ES � 0.2, medium ES � 0.5, and large ES � 0.8). Categorical data (satisfaction categorized according to statistical distribution, i.e., satisfaction score-≥ 90/100) were compared between the groups using the chisquared test or Fisher's exact test. To analyze the relationships between continuous parameters, Pearson and Spearman correlation coefcients (rho) were estimated according to the statistical distribution of variables and by applying Sidak's type I error correction (negligible < 0.2, weak 0.2 to 0.4, moderate 0.4 to 0.7, and strong > 0.7) [15]. To determine the factors associated with patients' satisfaction scores for the dispensing of their OAMs by pharmacists, multivariable analysis was performed using multiple linear regression, including patients' characteristics (age and the level of education) and the following covariates: source of questionnaire transmission (oncology departments or pharmacies), type of OAMs (targeted therapy or hormonal therapy), and information received (method of administration, adverse efects, and drug interactions). Particular attention was paid to the study of multicollinearity and to the interactions between covariates: (1) studying the relationships between the covariables and (2) evaluating the impact of adding or deleting variables on a multivariable model. Te results are expressed as regression coefcients and 95% CI, and forest plots were used to present the results. Concerning paired data comparisons (i.e., information on methods of administration, the management of adverse efects, and drug interactions between pharmacists, GPs, and oncologists), random-efects models were performed to take into account between-and within-patient variability. Sidak's type I error correction was also applied for two by two multiple comparisons. Statistical analyses were performed using Stata 15 (StataCorp, US). All the tests were two-sided, with a type I error set at 5%. In accordance with the literature [16][17][18], we reported all individual p values without systematically applying any mathematical correction to the aforementioned tests comparing groups. Specifc attention was given to the magnitude of diferences (i.e., ES and rho) and clinical relevance.

Sample Description.
One hundred and four patients agreed to take part in this study and responded to the questionnaire, of which 91 patients were included and analyzed ( Figure 1). Included patients were mostly women (62.6%) with a median age of 69 (58, 76) years old, in a couple (69.2%), retired (65.8%), and with a middle school certifcate (36.3%). Te most commonly represented cancers were breast (25.3%), lung (24.2%), and hematological malignancies (19.8%). Te duration of OAMs was approximately 38 (11, 125.5) weeks. Te majority of patients had visited only one pharmacy during the last 6 months (60.4%). Most of the participants obtained their treatments by themselves from pharmacies (82.4%). Te number of daily medications was 3 (2, 5). Te questionnaire was mostly provided by oncology departments (84.6%), with the remaining from pharmacies (15.4%) ( Table 1). Among the prescribed OAMs, patients mostly took targeted therapies (67.3%), followed by hormone therapies (23.5%) and cytotoxic ones (9.2%) (Supplementary materials) (available here).

Satisfaction of Patients regarding Community Pharmacy
Dispensing. VAS satisfaction scores regarding pharmacy dispensing were about 89 (68, 100) (rated from 0 to 100), and 49.5% (45) of participants had a satisfaction score ≥90/100. Te satisfaction scores were not related to gender, marital status, or professional activity. However, the satisfaction scores were signifcantly related to the patients' educational level. Nongraduated patients had higher satisfaction scores compared to patients in other categories (nongraduated: 100 Tere was a signifcant relationship between the number of pharmacies visited during the last 6 months and the satisfaction scores. Patients who visited one or two pharmacies had higher satisfaction scores than patients who visited three or more pharmacies (one pharmacy: 86 (66, 100) vs. two pharmacies: 93 (85, 100) vs. three or more 73 (49, 79), p � 0.013). Te satisfaction scores were not diferent among the fve main malignancies (breast, lung, hematological, melanoma, and kidney; p � 0.28) and among the three pharmacological classes of OAMs (p � 0.49). Finally, the satisfaction scores did not difer based on the source of questionnaire transmission (oncology departments vs. pharmacies) (p � 0.60) ( Table 1).
Te quality assessed by the patients and the type of information provided by pharmacists, GPs, and oncologists are presented in Table 2. Te VAS scores for assessing the quality of the information provided by the pharmacists were not diferent from those of GPs (p � 0.28) but were lower than those of oncologists (p < 0.001, ES: −0.91 95% CI (−1.23, −0.61)). Te VAS scores for assessing the quality of the information provided by the GPs were also lower than those of oncologists (p < 0.001, ES: −0.86 95% CI (−1.16, −0.55)) ( Table 2). Te VAS scores for assessing the quality of the information provided by pharmacists were moderately correlated with those of GPs (rho: 0.53, p < 0.001) and weakly correlated with those of oncologists (rho: 0.29, p � 0.007), but the VAS scores of the information provided by GPs were not correlated with those of oncologists (rho: 0.19, p � 0.09). All the VAS scores for assessing the quality of the information provided by pharmacists, GPs, and oncologists were correlated to the VAS scores of patients' satisfaction regarding pharmacy dispensing (rho: 0.63, 0.37, and 0.40, respectively, and p < 0.05 for all). Te type of information provided by health professionals (method of administration, management of adverse efects, and drug interactions) was also assessed by the patients. Regarding the method of administration and the management of adverse efects, fewer pharmacists and GPs provided information on these topics than oncologists (p < 0.001 for all) ( Table 2). Between pharmacists and GPs, we found no diferences concerning the method of administration (p � 0.053) and management of adverse efects (p � 0.61) ( Table 2). Regarding information on drug interactions, no diference was recorded between pharmacists and GPs (p � 0.64), and both proportions were lower than oncologist one (p < 0.001, for both) (

Symptoms and Quality of Life.
Te scores of the QLQ-C30 questionnaire are presented in Table 3. No correlation was identifed between the QLQ-C30 items and the VAS satisfaction scores. Only the nausea and vomiting scores of the QLQ-C30 questionnaire were lower in patients with satisfaction scores ≥90/100 compared to patients with satisfaction <90/100 (0 (0, 16.7) vs. 16.7 (0, 33.3), p � 0.049, ES: −0.44 95% CI (−0.86, −0.02)). No other diferences were recorded between these two groups of patients (data not shown).
Most of the QLQ-C30 items were not correlated to VAS scores assessing the quality of the information provided by pharmacists, GPs, and oncologists, except for weak correlations for information provided by GPs and constipation, and information provided by oncologists and insomnia and diarrhea (p < 0.05 for all) ( Table 3).
Several items of the QLQ-C30 were weakly correlated with the MMAS-8 scores (total and intentional), with the highest correlations observed for role functioning, fatigue, and appetite loss (rho > 0.3, p < 0.05) ( Table 3).

Multivariate Analysis of Patients' Satisfaction regarding
Community Pharmacy Dispensing. Multivariate analysis of patients' satisfaction, including patients' age, level of education, source of questionnaire transmission (oncology departments or pharmacies), type of OAMs (targeted therapy or hormonal therapy), information on the method of administration, information on adverse efects, and information on drug interactions, revealed a positive relationship with information on the administration method ( Figure 2).

Discussion
Patients' satisfaction scores regarding the quality of OAMs dispensing in community pharmacies were generally high (median 89 out of 100), with about half of them having scores ≥90, but without relationship with most patient characteristics, cancer type, OAM type, medication adherence, or HRQoL. However, patients with lower education levels or those visiting one or two pharmacies reported higher satisfaction scores compared to patients with higher education levels or those visiting three or more pharmacies. Patients reported higher satisfaction scores when pharmacists provided information on the method of administration, management of adverse efects, and drug interactions. Multivariate analysis showed that satisfaction scores were particularly related to the information provided by pharmacists about the method of medication administration. Interestingly, the study found that pharmacists ofered less information on the method of administration and adverse efects of OAMs compared to oncologists. For information about drug interactions, all three health professionals (pharmacists, oncologists, and GPs) provided relatively low levels of information, with no signifcant diference between them.
Te population of the study seemed to be quite representative of the general population and the main characteristics of cancer in France. Te median age at diagnosis was evaluated in 2018 at 67 years old for women and 68 years old for men [19], which corresponds to the median age in this study (66.5 years old). Te higher proportion of women (62.6%) can be explained by the large proportion of breast European Journal of Cancer Care 5 cancer. Te study clearly shows that targeted therapies are the most used and prescribed [20]. Breast and lung cancers are the most represented in this study. However, there is an under-representation of colorectal and prostate cancers in this study compared to French epidemiological data [19]. In this study, oncologists were the main source of information on OAMs, which is not surprising since, for most patients, their oncologist is the primary provider of cancer care and the specialist in this feld [21]. However, we encourage pharmacists to increase the provision of information on OAMs when dispensing medication (method of administration, adverse efects, and drug interactions). Considering that the pharmacist is the last healthcare professional that patients meet before taking their medication, it is essential for pharmacists to provide as much information as the oncologist did.
Regarding education levels, pharmacists must adapt their explanations to each patient. It is well known that patients with a low level of education may also have lower health literacy [22]. Interestingly, patients with lower health literacy tend to have a higher rate of satisfaction with the information provided [23]. Conversely, patients with higher health literacy may be less satisfed with their management, especially in cases of advanced cancer [23].
When it comes to drug interactions, similar results have been reported, with community pharmacists showing a poor level of recognition and management of drug interaction [24][25][26]. Tis is concerning, considering that OAMs are Table 2: Satisfaction with quality of information and type of information provided by pharmacists, GPs, and oncologists. Te quality of information provided by pharmacists, general practitioners (GPs), and oncologists was assessed by the patients with a VAS score (median (interquartile range)). Te type of information (the method of administration, adverse efects, and drug interactions) is presented by the percentage and (number). P values are provided for comparisons between community pharmacists and oncologists, GPs and oncologists, and community pharmacists and GPs.   [27]. In a retrospective nationwide real-world data-based study (N � 11,076), >75% of patients on OAMs had at least one potentially signifcant drug interaction [28].
Interestingly, it has been demonstrated that patients exhibit strong preferences for overall pharmacy quality, including specifc quality in drug interaction management. Patients highly value pharmacists' role in preventing drug interactions, underlining their high expectations in this regard [29].
In the present study, satisfaction scores were not related to HRQoL, whereas it has already been described that satisfaction regarding information is related to HRQoL (global HRQoL, physical well-being, social well-being, emotional well-being, and functional well-being) [30].
Regarding adherence to OAMs, more than 95% of the patients adhered to their medications. Tis number seems high knowing that adherence to long-term therapy for chronic illnesses in developed countries is about 50% [31]. However, other studies obtained results similar to our study, with a high level of medication adherence or compliance to OAMs (compliance to capecitabine: 91% [32], score of adherence to OAMs: 5.4 ± 1 based on a score ranging from 1 to 6 [33], and adherence to OAMs: >80% [34], 78% [35], and 71-73% [36]). Tis high adherence to OAMs can be explained in particular by the gravity of the cancer [32]. It is important to note that there is no gold standard for assessing medication adherence and that several measurement tools or strategies are described in the literature; also, medication adherence can be infuenced by many cofactors (e.g., psychological distress, perception of illness, concern of adverse efects, self-efcacy in medication management and decision making, knowledge of medication, and social support) [37,38]. Importantly, the duration of OAM treatment was relatively short (median of 38 weeks), and longer treatment durations may decrease medication adherence [39]. Most of the items concerning quality of life showed a signifcant link with the MMAS-8, especially the intentional items. We can suppose that an altered HRQoL can be the cause of nonadherence in patients. Indeed, the presence of adverse efects is strongly associated with poorer adherence [40]. However, a pharmacist-led adherence program can increase the proportion of patients adhering to OAMs [41]. Intentional MMAS-8 scores are also weakly correlated to VAS scores assessing the quality of the information provided by oncologists. We presume that good quality information provided by the oncologist to the patient can encourage the latter to better adhere to their treatment.
OAMs prescription and dispensing have changed in France since this study was performed. Multidisciplinary consultation programs involving the oncologist, the hospital pharmacist, and the nurse have been engaged and can decrease the incidence of adverse events [42]. Most importantly, the French regulation of community pharmacy activities has changed (September 30, 2020), with recognition and funding given to the pharmaceutical support of patients on OAMs, aiming "to make the patient autonomous and an actor in their treatment, limit the loss of reference points for these patients, promote monitoring, proper use and compliance with OAMs, inform the patient and obtain adherence for their treatment, help them in the management of treatments, prevent adverse efects, and provide coordinated patient care" [43].

Study Limitations.
A selection bias may be present with more satisfed patients or more unsatisfed ones. Satisfaction scores tended to be higher for patients recruited in pharmacies compared to patients recruited in oncology departments, but the diference was not statistically signifcant. Te use of a VAS score to assess satisfaction which is subjective could be a source of bias, but this has already been used in several medical conditions (hip arthroplasty [44], epilepsy management [45]). It is noteworthy that when assessing patient satisfaction, in comparison to the Likert scale, VAS seems to be less sensitive to bias from confounding factors and a ceiling efect, and the time taken to answer is shorter [46]. Moreover, this assessment of the patient's satisfaction for the quality of the dispensing of OAMs remains a subjective assessment of the work quality of pharmacist. Finally, the small sample size of patients means that the interpretation of the results should be treated with caution, particularly for secondary outcomes.  Figure 2: Multivariate analysis of patients' satisfaction for the dispensing of their OAMs by community pharmacists. Te multivariate analysis includes age of patients, level of education (reference: no education), source of questionnaire transmission (ref: oncology departments), type of OAMs (reference: no targeted therapy), hormonal therapy (reference: no hormonal therapy), and information provided by the community pharmacist on the method of administration (reference: no information), adverse efects (reference: no information), and the drug interactions (reference: no information). Te results are presented by coefcient and 95% confdence interval (95% CI).

Conclusions
Te satisfaction of patients regarding their OAM dispensing by community pharmacists was high, but the provision of information on OAMs by pharmacists remained low in comparison to oncologists, whereas it should be at least at the same level. Moreover, patients' satisfaction was associated with the provision of OAM information during dispensing. Te good practices of medication dispensing by pharmacists are an important component of patient care, and we encourage pharmacists to provide more medication information to their patients.

Data Availability
Data will be made available upon request to the corresponding author.

Conflicts of Interest
Te authors declare that there are no conficts of interest regarding the publication of this paper.