Summary of the Research Medication non-adherence has a huge economic impact on the patient and the society at large. World Health Organization (WHO) has suggested that noncompliance with medication is a common problem which often leads to compromised health benefits and serious economic consequences in terms of wasted time, money and increased morbidity. Other consequences are waste of medication, disease progression, reduced functional abilities, lower quality of life, and increased use of medical resources such as nursing homes, hospital visits and hospital admissions. Non-adherence is common to patients with chronic diseases, such as hypertension and diabetes, compared to patients with acute conditions. The annual cost of medication non-adherence, which includes hospital and nursing home admissions, increased ambulatory costs, lost productivity and premature deaths, has been estimated to be more than $100 billion per year in the United States. In Canada, hospital expenditures caused by non-adherence was estimated to be more than US$1.6 billion. The potential burden of medication non-adherence outcomes on health care delivery makes it an important public health concern.

Many studies have addressed non-adherence to medication in developed countries. Past studies on non-adherence in Nigeria have identified different rates of non-adherence to medications for disease states such as diabetes mellitus and hypertension. Literature is scanty on studies on cost of non-adherence in Nigeria. This study will be focus on cost implications of non-adherence to treatment among patients of selected disease states (hypertension and type 2 diabetes mellitus).

Research design The study was a quasi-experimental study among T2D patients that were recruited from the two hospitals using questionnaire-guided semi-structured interview. At the baseline, participants with HbA1c ≥7% were classified as uncontrolled/intervention group, while those with HbA1c <7% were considered as control group. Similarly, participants with blood pressure <140/90mmHg were classified as control/adherent while those with blood pressure ≥140/900mmHg were uncontrolled/intervention group.

Methodology A quasi-experimental study among 201-patients with T2D using semi-structured interview. Baseline questionnaire comprised modified 4-items Medication Adherence Questions (MAQ), Perceived Dietary Adherence Questionnaire (PDAQ) and International Physical Activity Questionnaire, to assess participants' adherence to medications, diet and physical activity, respectively. Patients were assigned into control (HbA1c<7%, n=95) and intervention (HbA1c≥7%, n=106) groups. Post-baseline, participants were follow-up for 6-months with educational intervention provided to clarify and resolve identified discrepancies among the intervention group only, while the control group continued to receive the usual care. Costs of management including transportation fare, consultation fee, medications and laboratory investigations were estimated for 6-months pre-baseline and 6-months post-baseline for both groups. Data were summarized using descriptive statistics, while Chi-square, McNemar and paired t-test were used for categorical and continuous variables at p<0.05.

Principal exposure: Glycated haemoglobin HbA1c, was measured for Type 2 DM participants and blood pressure was taken for hypertensive participants Outcome variable: Effect of pharmacist-led intervention on therapy non-adherence, reasons for non-adherence and costs of management.

The study involved adult T2D, hypertensive and those with the two diseases attending the cardiology and endocrinology out-patient clinics of the two tertiary hospitals [University College Hospital (UCH), Ibadan and Federal Medica Centre (FMC), Abeokuta, Nigeria].

Sample size determination

For the prospective phase, an average of 25 patients regularly attend the weekly medical out-patient endocrinology clinic of the two hospitals. This gives a total of 100 patients per month in each hospital, and for the 6-months study period gave an estimated population of 600 T2D per hospital. However, a maximum of 2 to 6 months clinic appointment is usually given to T2D in UCH, and a maximum of 3 months appointment in FMC, the appointment schedule was based on the extent of glycaemic control. Based on this information, an estimated T2D population for 4-months and 3-months was used to calculate the target sample size for UCH and FMC, respectively. Thus, considering the regular attendee of 100 patients per month in each of the hospital, a total of 700 T2D was used as estimated population to guide sample size determination using the Raosoft® sample size calculator at 95% confidence level and 5% margin of error (http://www.raosoft.com/samplesize.html). This gave a target sample size of 249 T2D participants to be enrolled for the study. Similarly, Average of 20 hypertensive patients regularly attend the weekly medical out-patient cardiology clinic of the two hospitals. This gives a total of 80 patients per month in each hospital, and for the 6-months study period gave an estimated population of 320 hypertensive patients per hospital. However, a maximum of 2 to 6 months clinic appointment is usually given to hypertensive patients in UCH, and a maximum of 3 months appointment in FMC, depending on the extent of control of the blood pressure. Based on this information, an estimated hypertensive patients population for 4-months and 3-months was used to calculate the target sample size for UCH and FMC, respectively. Thus, considering the regular attendee of 80 patients per month in each of the hospital, a total of 560 was used as estimated population to guide sample size determination using the Raosoft® sample size calculator at 95% confidence level and 5% margin of error. This gave a target sample size of 229 hypertensive participants to be enrolled.

Instrument descriptions The baseline questionnaire comprised sections A to F. Section A captured demographic data, section B contained modified International Physical Activity Questionnaire short-form (IPAQ-SF) (IPAQ-SF, 2002) and individual physical activity was calculated as metabolic equivalent task (MET) per week (http://www.ipaq.ki.se). IPAQ -SF instrument measures duration, frequency, intensity of physical activity in leisure-time, work, and doing household tasks in the past seven days. This short version consists of seven items including time spent in walking, vigorous activity, moderate activity, and sedentary activity weekly. Metabolic Equivalent Tasks (MET) values reported by patients were computed. To interpret the modified IPAQ-SF, according to WHO, an adult is expected to have minimum of 600 Metabolic Equivalents (MET) physical activities weekly. Participants were classified as non-adherent and adherent to exercise if he/she has less than 600 and above 600METs per week respectively.

Section C consisted of modified Perceived Dietary Adherence Questionnaire (PDAQ) developed by Assad et al., (2015), to measure patients' perception of dietary adherence. For these items, higher scores reflected higher level of adherence (Assad et al., 2015). However, items four and nine, indicated unhealthy choices (foods high in sugar or fat). For these items, higher scores reflected lower adherence. In computing the total PDAQ score, the scores for these two items were inverted. Although based on a weekly timeframe, it was anticipated that the PDAQ would reflect usual dietary patterns based on knowledge that most people consume similar foods from week to week (Assad et al., 2015). In this study, PDAQ was interpreted by classifying participants as adherent to dietary recommendations if they scored 51 (80%) and above, out of the 63 maximum obtainable score, while those who scored less than 51 (80%) were classified as non-adherent to dietary recommendations. Section D of the instrument contained the 4-items Medication Adherence Questions (MAQ) while section E explored reason(s) for non-adherence to medications, physical activity and dietary recommendations. Section F of the instrument contained the components of educational intervention to resolve non-adherence problems that were identified from the responses to questions in MAQ, IPAQ-SF and PDAQ among patients in the intervention groups. Where necessary, at least one to two clinic appointment was rescheduled to follow-up of the participants in the intervention group. The 6-months post-baseline evaluation employed the same questions in sections B, C, D and E to re-assessed participants in both control and intervention groups, to ascertain the possible change in adherence status to recommended medications, physical activity and dietary management.

Validation and Pre-test of Data Collection Instrument

The questionnaire was assessed for content validity by an endocrinologist and cardiologist in UCH and three pharmacists in the academia in University of Ibadan who were knowledgeable about diabetes mellitus and hypertension. A pre-test was caaried out among twenty-five T2D and hypertensive patients in UCH who were subsequently excluded from the main study. The pre-test was aimed at ascertaining the appropriateness of the data collection procedure as well as the study design. Feedback from the pre-test was used to modified. Some questions on physical activity which were rephrased as open-ended questions. In addition, some food items in the PDAQ that are not native food of the studied population were replaced with their indigenous food with similar calorie in accordance with official recommendations (Atkinson et al., 2008; Evert et al., 2014) for better understanding of the partiicpants. Internal consistency of the question-items in PDAQ and MAQ were determined using Cronbach alpha test with values of 0.87 and 0.76, respectively.

Participants enrollment procedure

Only eligible patients were approached while waiting to consult the attending physicians on endocrinology clinic day of each hospital. At baseline, a total of 249 T2D and 262 hypertensive patients were approached, while 227 (91.2%) T2D and 248 (94.6%) hypertensive patients consented to partake in the study from both hospitals within the study period. Every participant had individualized baseline, and once the 6-months duration of enrolment into the study is completed, then individual participation is terminated. At baseline, participants were administered the questionnaire, while glycaemic status (HbA1c) and blood pressure (BP) were also assessed. However, at the end of the study period, a total of 201 (88.5%) T2D and 200 hypertensive patients completed the study and were those considered for data analysis, 26 (11.5%) T2D and 48 (19.3%) hypertensive patients were lost to follow-up. Of the 201 and 200 who competed the study, a total of 95 T2D patients had HbA1c <7%, 120 hypertensive patients had blood pressure <140/90mmHg and were classified as control group, while 106 T2D patients had HbA1c ≥7% and 80 hypertensive patients had blood pressure ≥140/90mmHg and were classified as uncontrolled/intervention group using the ADA target for diabetes control (ADA, 2018) and according to JNC-7 (Chobanian et al., 2003) respectively. Finger pin-prick blood assay of glycosylated haemoglobin (HbA1c) was done using the clover A1C automated analyser ® (PTS Diagnostics, IN, USA) point-of-care kit, while BP measurements using Omron® digital monitor were taken on two separate occasions at few minutes interval and the average was recorded. Patients' BP and HbA1c were also rechecked after 6-months of interaction to assess the extent of change in the diabetes-specific clinical parameters and hypertension-clinical parameters.

Data collection procedure participants were approached on their clinic days while waiting to meet the physician at each clinic. the structured questionnaire was used to collected baseline information from the participants. Reasons for non-adherence to medication (s) were obtained if any of the questions in the MAQ was answered in affirmative. IPAQ -SF instrument measures duration, frequency, intensity of physical activity in leisure-time, work, and doing household tasks in the past seven days. Participants were classified as non-adherent and adherent to recommended physical activity if he/she has less than 600 and above 600METs per week respectively. Reasons for non-adherence/inability to achieve recommended MET per week were obtained from such participant. Before recommending an increase in physical activity to patients in the intervention group, the participant's current health status was put into consideration based on the information provided by participant on the disease and other comorbid condition(s) been managed for, such as arthritis, heart disease.

Similarly, participants were further evaluated for adherence to dietary recommendation for their disease state using the modified Perceived Dietary Adherence Questionnaire (PDAQ) (Appendix II). The cut off of 51 points of the total of 63 points indicated that the patient have 80% adherence to dietary recommendation. Similarly, participants reason(s) for non-adherence to recommended dietary management was also obtained.

HbA1c/B.P measurement was carried out to assess their clinical outcome at baseline and 6month post-baseline to evaluate their clinical outcome.

Post-baseline individualized education (Intervention) Phase The individualized educational intervention was check using Template for Intervention Description Replication (TIDiER) check list

The individualized intervention was aimed at improving treatment adherence among patients in the intervention group.

The individualized educational intervention was aimed at resolving non-adherence to therapy (medication, diet and physical activity) deficit identified during the face-to-face interview at the baseline. The hospital information leaflets provided for each patient at the hospital clinics during registration which focused on recommended diet and physical activities was employed to further explained the recommended therapy for the patients. Pharmacist who was knowledgeable on the disease state by the virtue of training and several seminars and workshops on diabetes and hypertension management carried out the face-to-face individualize educational intervention. This was done at the clinics while the participants were waiting for their turn to meet with the physician. At least participants were seen twice to ensure their understanding of information provided for them and resolve any for of misunderstanding that might have taken place. During the pretest, the project supervisor assessed the intervention to ensure it was carried out as planned.

Post-intervention phase This phase of the study re-assessed the treatment adherence of all the participants (both control and intervention group). Sections B, C, D, E and F of Appendix II were re-assessed for both control and the intervention group, six months after the educational intervention of the intervention group. This was done to assess if there is any change in the measured parameters from what was obtained at baseline.

Cost estimation The direct medical costs associated with the treatment of diabetes and/or hypertension for six months prior baseline and six months post-baseline were estimated. The direct medical cost items included; cost of all outpatient clinic visit (consultation fee), cost of medications, and cost of laboratory investigations associated with the disease state. Cost of medication(s) was calculated using the daily dose for medications prescribed with reference to the hospital pharmacy department tariff for the study year. Consultation fee is the amount charged by the hospital for a consulting encounter with a physician. Other direct non-medical costs included are; transportation cost to the hospital for patients based on patient's residence and this was estimated using the NURTW tariff for the year of study. Conversion of Nigeria Naira to US Dollar: ₦303 to 1 USD for the year 2017 was based on conversion rate on currency converter (oando.com).

Data Analysis All data was sorted, coded and entered into SPSS version 23.0 for analysis. Descriptive and inferential statistics were employed for analysis. Descriptive a statistics including frequencies, percentages, and means ± standard deviation were used to summarized data.

significance of change in continuous variables, especially the clinical parameters (such as HbA1c, systolic and diastolic blood pressure), physical activity performed weekly, adherence to recommended diet at baseline and six month post-baseline were compared using paired sample t-test and significance was set at P <0.05. Cost of each cost items 6months prior baseline and at 6months post-baseline for each disease state within and between the two facilities were also compared using paired sample t-test. Total cost of management (including all direct cost associated with the disease state) for each group in each disease state before and after the intervention was evaluated with paired sample t-test.

McNemar test was used to evaluate for significant change in responses to MAQ items at baseline and 6-months post-baseline. Chi-square (χ2) was used to test for association between categorical variables (gender, age, educational qualifications and adherence status) at p<0.05.

Inclusion Criteria

• Adult out-patients diagnosed and being managed for T2D or hypertension or the co-morbid diseases.

  • Participants must have been on antidiabetes/or antihypertensive medication(s) for at least six-months prior to their recruitment to the study.

Exclusion Criteria:

• Newly diagnosed T2D and hypertensive patients
• Pregnant women with diabetes or hypertension.
• Unconscious patients and those booked for admission.
• Patients less than 18 years of age.

Participants were allocated based on their HbA1c values and blood pressure in to control and intervention groups. Participants were not aware if they were in intervention group or control group as they were not informed.