A hospital based crossectional analytical study design used quantitative methods of data collection to determine the quality of adherence counselling among cases of MDR-TB and DS-TB receiving care at Katutura State Hospital, the national referral TB hospital for Namibia. The target population comprise hospitalized patients treated for DR-TB (Drug-resistant Tuberculosis) and DS-TB (Drug-susceptible Tuberculosis). The study included all the 25 DR-TB patients admitted at the hospital at the time of the study (i.e. April to June 2018). Subsequently, a sample of 25 patients hospitalized with DS-TB were selected using simple randomly sampling. The determination of the sample size was informed by the maximum bed-capacity for DR-TB cases (i.e. 25 beds) and the acquisition of a written informed consent to participate in the study. The study excluded all patients at the private health facilities as well as those at any public health facility that is not Katutura State Hospital.

The data on quality of adherence counselling was collected MDR-TB and DS-TB patients using interviewer-administered questionnaires (Annexure B) that consisted of 26 questions. The tool was piloted on 5 patients with DR-TB and standardized as well as translated in local two languages, Oshiwambo and Afrikaans, during the interviews by the researcher (TS). The data were subsequently entered in Epidata v3.1 for management (checked thoroughly for completion and errors and then stored securely).

The main outcome variables were the level of perceived quality of adherence counselling (i.e. satisfaction with counselling, comprehensiveness, time spent on counselling and the frequency) and associated factors (i.e. patient and health system related factors). Quantitative data were exported to SPSS v 23 for quantitative analysis. The quality of adherence counselling was measured based on indicators (i.e. timing, duration, frequency and satisfaction) as well as patient awareness of good DOT adherence counselling and practices, patient-related factors and hospital-related factors that affect the quality of DOTS Adherence counselling. The factors associated with the quality of DOT adherence counselling were analysed using descriptive statistics and an association was made using chi-squared and t-test. The level of significance for both tests was set at 95% confidence interval with a level of alpha at 0.5.

Ethical clearance was obtained from the research and ethics review board (REC) of the Ministry of Health and Social Services as well as from the Katutura Intermediate Hospital. In addition, each patient freely volunteered to participate in the study and the patients were given a written and verbal detailed explanation of the study (Annexure A). Patient confidentiality was upheld at all times and patient indicators and private details such as their names were not disclosed.

Fifty tuberculosis patients participated in the study, yielding a response rate of 63% (50/80). Of the 50 participants, 60% (n = 30) were male and 77% were aged 45 years or younger. Majority of the participants (66%) had attained at least a secondary education. The number of respondents with multi-drug resistant TB (MDR, n = 25) was equal to those with drug-susceptible TB (DST-TB) patients (Table 1). There were no significant differences between the sociodemographic characteristics of patients with MDR and drug susceptible tuberculosis (p > 0.05) (Table 1).

Of the 50 patients, the majority (56%) felt that their adherence would improve at home as opposed to being in the TB clinic. The main reasons for non-adherence to TB medication and/or DOT appointments were forgetfulness (22%) and seizing treatment upon feeling healthier (12%) (Table 2). In addition, patients perceived improved qualities of DOT adherence counselling to include; individual counselling sessions (74%), an adherence counsellor of the same sex (16%) and/or age category (12%) (Table 2). There were no significant differences in the factors associated with adherence counselling among patients with MDR and DST tuberculosis (p > 0.05) (Table 2).

Of the 50 patients, 92% (n = 46) confirmed that the stipulated schedule for DOT

MDR-TB = Multi-Drug Resistant Tuberculosis, DS-TB = Drug-Susceptible Tuberculosis.

was consistent majority of the time and 76% of the patients found the scheduled time to be convenient. In addition, 80% of the patients were counselled in a language that they understood, however 16% were however not counselled at all since having been admitted (Table 3). In addition, 16% of the patients found the counsellors to be judgmental, while 74% felt that the counsellors created a comfortable and free environment (Table 3). There were significant differences in the hospital factors associated with adherence counselling among patients with DST and MDR tuberculosis (p > 0.05).

Of the total patients that partook in the study, 32% knew what DOT is and while 26% of patients knew what MDR-TB is, only 2% knew how it arises (Table 4). About 88% of the patients were however aware that TB is curable and 96% of them had faith in the possibility of treatment success. Furthermore, 26% percent of the participants felt that a few missed doses or more is insignificant towards precipitating treatment failure (Table 4).

Of the 50 patients, 82% (n = 41) received DOT daily at consistent times, 36% (n = 18) of the patients received adherence counselling at least 1 month ago or earlier whilst 16% (8/50) were never counselled at all. Of the 84% of patients that had received adherence counselling, 74% had counselling sessions that were at least 5 minutes long (Table 4, Figure 1).

The patient-perceived quality of DOT adherence counselling indicators revealed that 34% of the patients (n = 17) felt that DOT was moderately important or unnecessary, however 92% (n = 46) rated the DOT that they received as good or excellent. About 34% of the patients identified aspects that can be modified to improve the DOT quality and similarly, 42% of patients identified areas of adherence counselling where improvements can be made to the quality (Table 4, Figure 2). A significant number of MDR-TB (44%) and DS-TB (57%) patients would not appropriately define what directly observed therapy (DOT) entails and/or aware of whether tuberculosis is curable (Figure 2). Over half of the patients don’t know want to do in case they miss to take a dose of their TB medication (Figure 2).

The study aimed to determine the quality and factors associated with TB DOTS adherence counselling among hospitalized DR-TB patients and DS-TB patients. We found that the quality of DOT Adherence counselling is sub-optimal in terms of the comprehensiveness and service delivery. The study also found that patients had poor knowledge on basic aspects of DOT and TB treatment and outcomes. A stratified cluster randomized control trial in America also found the inadequacy of the standard DOT adherence counselling as well as improvement in treatment outcomes [19] . Similar studies, also report improvement in treatment outcomes among patients in the adherence-counselling programme [26] [27] . This calls for strengthening quality of adherence counselling in the DOTS programmes particularly in TB programmes in LMIC.

The main driving factors of the quality of DOT adherence counselling identified by this study were patient-related factors and hospital-related factors (Table 2, Table 3). The main patient-related determinants identified by this study were patient preferences, of which the most significant one was the preference of individual counselling sessions as opposed to group counselling sessions (74%, n = 37). This was because the patients felt that they could disclose any potential problems affecting their adherence more freely in individual counselling sessions as well as ask any relevant questions could potentially be personal. In addition, the majority of the patients however did not have any preferences regarding the age and sex of the counsellors (Table 2) and reasoned that any age or sex is acceptable, provided that the counsellors are professional and competent. The main reasons identified for a lack of adherence was forgetfulness and treatment cessation by the patient upon feeling better (Table 2).

Similar studies, such as a study done in India have also identified patient-related factors that affect the emergence of drug resistant TB, the study found that the main barriers to adherence were drug side effects, a perceived lack of provider support, patient financial constraints, and conflicts with the timing of treatment services, alcoholism and social stigma [15] [28] [29] . Another study done in China used multivariate conditional logistics regression models to identify determinants of MDR-TB and/or adherence to medication which included, previous TB treatment; male sex, high school or lower education degree, unemployment, long distance of residence from the health facility; smoking, poor knowledge regarding MDR-TB, social stigma, among others [29] [30] [31] .

These studies are however quite limited because the variables investigated as potential determinants of MDR-TB were only from the patient’s perspective and completely neglected any determinants of MDR-TB that could be as a result of the health care providers, such as a lack of adherence counselling or a poor quality thereof. The results of the patient related factors affecting the quality of DOT adherence counselling in this study reinforce the results that were found in the above mentioned studies. This indicated that there is a compelling need for better management particularly through improving medication adherence guidelines.

The study also investigated patient preferences. We found that most patient preferred to take medication at home (56%, n = 28) rather than in the hospital. Whilst the others argued that being an in-patient was more beneficial because they had access to free meals at the hospital and reduced negative influences such as taking alcohol, that hinder good adherence and treatment completion.

Several studies such as a study done by Cadosch, et al. have investigated the effects of adherence counselling and DOT on treatment outcomes. The results of that study showed that failure to adhere to anti-TB medication directly affects treatment outcomes [22] . Another study done in 2016 looked at the association between DOT and TB treatment outcomes and found that there was no statistical significance [17] . These studies however did not look at any factors that directly influence the quality of DOT adherence. This study did however investigate the hospital-related factors and found that: majority of the patients found that the DOT schedule was consistent and convenient (Table 3). About 80% of the patients were counselled in a language that they understand, however (16%, n = 8/50) were not counselled at all and 74% felt that the counsellors were friendly and not judgmental. These results reflect that with respect to the hospital related factors, the null hypothesis is satisfied and the quality of the DOT counselling according to these factors is optimal.

This study also used specific indicators to directly measure the quality of the TB DOT adherence counselling (Table 4). The indicators were based on the DOT adherence counselling services itself, patient knowledge after DOT counselling and patients directly rating the DOT counselling (Table 4).

Results that directly reflect on the service itself concluded that (94%, n = 47) of the patients received DOT at each or most doses (although some patients revealed that they were allowed to drink the medicines in their rooms without supervision which defeats the purpose of DOT), however only 40% of the patients were counselled at least 3 months ago, 36% were counselled at least 1 month ago and 16% were not counselled at all since being admitted. These results, especially those pertaining to adherence counselling frequency are significantly low and thus provide compelling evidence for the need for improved national standard operating protocols for DOTS and adherence counselling. In addition, results pertaining to the patient-perceived quality of the DOT services were high (Table 4) and acceptable for optimal DOTS quality.

The patients’ knowledge as assessed after the patients received DOTS and adherence counselling was however very poor (Table 4). This clearly rejects the null hypothesis and indicates that the quality of DOT counselling as related to patient knowledge is indeed suboptimal. There are no documented studies that directly measured the quality of DOTS or adherence counselling using indicators and thus the results from the indicators could not be compared to any other studies.

Based on the findings, we conclude that the quality of DOTS counselling for DR-TB as well as DS-TB cases at Katutura Intermediate Hospital is sub-optimal.

The results from the patient-related factors and hospital-related factors reinforced findings from studies previously done that indicate that the quality of DOT adherence counselling in these particular areas is optimal.

The results from the indicators revealed that the patients’ knowledge regarding TB at large is very poor, which in turn reflects that the quality of DOT adherence counselling is sub-optimal. The frequency of adherence counselling was also found to be sup-optimal according to our judgement and the patients’ views and recommendations.

Despite the compelling evidence provided by similar studies on the strong association between good DOT adherence counselling and increased treatment outcomes, there are no existing TB DOT adherence counselling protocols nationwide or from the World health organization that may be used to accurately measure the quality of DOT adherence counselling.

The indicators identified in this study may be used to develop national standard operating for DOTS and adherence counselling which can be used to measure and ensure nationwide good quality DOT adherence counselling which is a crucial determinant of the emergence of drug-resistant tuberculosis.

This research study was done in Windhoek (which is an urban setting) and therefore, the results are not generalizable to the country’s entire population. Furthermore, this study was done on a Namibian population and the results from this study are not necessarily be generalizable for any other population groups.