Background

Adherence to treatment is essential for complete cure. Many oral anticancer regimens have been approved for treatment with equivalent efficacy to parenteral regimens. However, the question of adherence arises. Suboptimal adherence is a barrier to effective use of oral anticancer drugs.1 Nonadherence also affects the patient–physician relationship and has a negative effect on the patient’s views about physicians and services.2

Oncologists assume that cancer patients will take their medications as prescribed.34 Therefore, this study will help us understand the prevalence of nonadherence. Poor adherence leads to unfavorable outcome and decreases the 5-year event-free survival of the patients. It will decrease the likelihood of achieving complete cytogenetic response.5 Nonadherence also leads to unwanted diagnostic and treatment procedures, causing health problems.2

Aims and Objectives

The primary objectives of the study were to determine the adherence rates of oral anticancer regimens for different types of cancer in an urban setting and to determine the various factors affecting adherence.

Materials and Methods

The study is cross-sectional analytical consisting of a single group. All fit patients, (patients who were able to fill the questionnaire on their own or able to answer the questions when asked) > 18 years of age, and on oral anticancer drugs for cancer who have taken the drug for at least 1 month, with or without concomitant intravenous (IV) anticancer drugs, were included in the study. Convenient sampling was done. All consecutive patients fulfilling inclusion criteria attending the Medical Oncology Outpatient Department clinic from June 2017 to September 2017 were taken up for the study. The basic sociodemographic details about the cancer diagnosis and the treatment were collected using a pro forma. A standardized Tamil questionnaire consisting of a set of questions to identify the factors affecting adherence was used. To determine adherence, the Morisky–Green–Levine (MGL) adherence scale6 was used.

The Institute Ethics Committee approval was obtained. The questionnaire and the MGL scale were translated into Tamil and the Tamil form was retranslated to English by another person and checked for standardization.

Observation and Results

The study included 152 patients fulfilling the inclusion criteria. It was found that 73% of the patients were adherent and 27% of the patients were nonadherent. The mean age of the study population was 49.03 ± 13.4 years. Approximately 52.6% of the patients were > 50 years of age, with equal proportions of males and females, and 93.4% of the patients had a caretaker. About 30% of the patients were uneducated and 45% of the patients were unemployed. Nearly 54.6% of patients were on treatment for < 12 months’ duration (median of 9 months with an interquartile range of 4–42 months) and 34.2% of patients did not take any oral supportive medications. About 71.7% of the patients took their medications themselves and 41.4% of patients experienced adverse drug reactions (ADRs). About 63.7% of patients had solid malignancies and the rest were hematological malignancies (Table 1).

About 97.4% were aware that they had cancer, and 12.5% of patients were aware of the diagnosis, treatment, and outcome. About 61.2% received majority of the information before treatment initiation, 93.4% of the patients received majority of knowledge from physicians, and 42.8% of the patients were aware about the outcomes of nonadherence. Only 9.2% of patients stopped drugs intermittently without consulting the physicians. Almost 98% did not use any methods to avoid forgetting to take medications. About 23% of patients experienced difficulty in remembering to take medications on time, and 11.8% of patients considered taking medications as an inconvenience. A major proportion of patients (67.1%) visited their physicians once in every 3 to 4 weeks. Only 16.5% of patients visited doctors besides scheduled visits, with the most common reason being pain due to illness, followed by side effects, fever, and doubts regarding drug intake in decreasing order of magnitude. About 92.8% of patients were satisfied with their physicians.

Adherence rates in gastric cancer and breast cancer were 60% and 68%, respectively. Chronic myeloid leukemia (CML) and colorectal cancer had a better drug adherence rate of 75%. A majority of the patients (75.6%) who were nonadherent mentioned forgetfulness as a reason for being nonadherent followed by carelessness (Table 2).

On bivariate and multivariate analysis, we observed that patients with < 1 year of treatment, adverse effects, taking oral anticancer drugs for < 21 days per month, coadministered with IV anticancer drugs, and patients with more than once daily dosing had significantly poor adherence. Bivariate analysis of other variables such as age, sex, caretaker, occupation, income, hospital distance, comorbidities, type and stage of cancer, chemotherapy regimen, dose of drugs, oral supportive drugs, knowledge about disease, and treatment were insignificant. On multivariate analysis, only ADR showed a trend toward correlation with nonadherence.

Discussion

In spite of increased use of oral anticancer drugs in the recent times, the number of studies addressing the issue of adherence is very low. The study assessed the adherence of oral anticancer medication in a tertiary cancer government hospital. We identified that 27% were nonadherent. Our study looked at adherence rates both in solid and hematological tumors. In a study by daCosta et al, which was done to assess the patient preferences and treatment adherence among, 34.8% of the women diagnosed with metastatic breast cancer were nonadherent. Patients receiving hormonal therapy reported the highest level of nonadherence.6 Forgetfulness was higher in our population when compared with a study in noncancer patients.7 In our study, 75.4% of CML patients on imatinib or hydroxyurea were adherent when compared with the study by Marin et al, which reported median adherence measured by microelectromechanical systems (MEMS) was 97.6%.8 Another study done on adherence among CML concluded that 32.7% of participants were highly adherent, medium adherence in 46.5%, and low adherence in 20.7% of the study population.9 The possible reasons for increased nonadherence rate need to be examined in further studies. The adherence rate in CML patients as determined by the questionnaire was 74.5% which is correlating with the study done by Noens et al in which they found questionnaire-based adherence to be between 67% and 97%.10

In our study, the presence or absence of a caretaker was insignificant in bivariate analysis, but a systematic review done to determine adherence to oral anticancer drugs found out that patients who were living alone had poor adherence. The study also concluded that lower educational status and patients with no family history of cancer had poor adherence.11

In a study by Timmers et al, the adherence rate in patients on anticancer drugs experiencing ADR was 33%, which was low when compared with our study.12 Veronesi et al13 and Demissie et al14 reported a nonadherence rate of 26.7% and 15%, respectively, to hormonal therapy in breast cancer, which was lower than our study (32%).

Conclusion

Drug adherence is one of the key factors for treatment failure in cancer patients. ADRs have been found as an independent variable, resulting in decreased drug adherence. Duration of treatment, coadministration with IV chemotherapeutic drugs, duration of oral anticancer drug intake in a month, and frequency of drug intake (anticancer drug) were significant factors which affect oral anticancer drug adherence. Understanding and identification of early adverse events and interventions will help decrease treatment failure. Knowledge about the disease, understanding the drug prescription, and side effects were lacking in our study group.