A quantitative cross-sectional study was conducted among hypertensive patients to identify the factors associated with treatment adherence among hypertensive patients and their associated predictive factors in a state hospital in Kavango East Region, Namibia. Quantitative research is a systematic, objective process that uses numeric data obtained from a subgroup in order to generalise the findings to the population being studied (Maree 2016). This approach uses randomised samples, allowing one to reach a higher sample size, and one can collect information quickly when using this approach (Bhandari 2020). Hence, the approach was used to quantify the factors associated with non-adherence to antihypertensive medication at a state hospital in Kavango East Region, Namibia.

The study was conducted at the outpatient department of a selected regional hospital located in the northeast of Namibia between March 2022 and April 2022. This government hospital has 300 beds and caters to hospital and clinic referrals from the Kavango East, Kavango West, and Zambezi regions. According to the Outpatient Department Monthly Summary Report form of 2021, 401 patients were treated for hypertension (new cases and follow-up).

The researcher used the population of 401 patients who were treated for hypertension as informed by the Outpatient Department Monthly Summary Report form of 2021. The sample size of this study was determined by using a standard formula; sample size = N/(1 + Nxa ^ 2) for a cross-sectional study (De Vos et al. 2017), where:

The calculation will be performed according to Sekaran and Bougie (2016) as below:

Hence, the sample size for this population was 200.

The sample of patients on hypertensive treatment, who satisfied the inclusion criteria, was selected. Consecutive sampling was used, whereby the researcher selected the research subjects based on their availability and willingness to participate in a study. The researcher recruited individuals who met specific criteria one after another until the desired sample size of 200 was achieved. The sampling technique was used because it was difficult to access a large pool of potential participants, and the researcher required a specific population that was readily available. The researcher waited for hypertensive patients who went for treatment or check-ups at the health centre. Data were collected as the patients came for a period of two months. This method is reliably convenient and efficient in terms of time and cost. Participants in this study had to be 18 years or above and on hypertension medication, and they had to be willing to participate and able to give a written informed consent. Participants in this study were excluded if they were less than 18 years old, not on hypertension medication, or not willing to participate.

A semi-structured self-administered questionnaire was designed and used to collect data on the demographic characteristics and the individual-related practices of hypertensive patients. The questionnaire used in this study was developed based on previous research on factors associated with non-adherence to antihypertensive medication, research questions, and research objectives. The instrument was translated into a local language (Rukwangali) and then retranslated back into the original language (English). This practice is referred to as forward and backward translation (Tsang, Royse & Terkawi 2017). This exercise was performed to accommodate participants but fundamentally also to ensure the content validity of the instrument and/or survey, so that ‘the entire theoretical construct the questionnaire is designed to assess’ (Tsang et al. 2017:86) will measure the same in a different context. Variables included in the study were grouped into demographic characteristics, and individual-based practices associated with non-adherence to hypertensive treatment were relevant as they directly addressed the study objectives of understanding the factors influencing medication adherence among hypertensive patients. Demographic characteristics such as age, marital status, religion, education, employment status, gender, and number of children were essential for identifying socioeconomic and cultural factors that may impact adherence. Individual practices related to medication timing, discontinuation, forgetfulness, side effects, and understanding of hypertension were crucial for assessing personal behaviours and perceptions affecting treatment adherence. The predictors were demographic characteristics and individual-based practices of the participants. These variables collectively allowed the study to comprehensively evaluate the predictors of adherence. The questionnaire had 38 items, which were divided into two sections: section A had 20 items on demographic characteristics, and section B had 18 items on medication adherence and reasons for non-adherence. The variables included in the study, which were grouped into demographic characteristics and individual-based practices associated with non-adherence to hypertensive treatment, were relevant as they directly addressed the study objectives of understanding the factors influencing medication adherence among hypertensive patients. Demographic characteristics such as age, marital status, religion, education, employment status, gender, and number of children were essential for identifying socioeconomic and cultural factors that may impact adherence. Individual practices related to medication timing, discontinuation, forgetfulness, side effects, and understanding of hypertension were crucial for assessing personal behaviours and perceptions affecting treatment adherence. The predictors were the demographic characteristics and individual-based practices of the participants. These variables collectively allowed the study to comprehensively evaluate the predictors of adherence.

The questionnaire was pretested on four random participants to ensure the internal validity and reliability of the assessment tool in this study and to confirm whether it was able to collect data in order to meet the objectives of the study. Feedback from a pilot test did not result in any adjustments to question wording and format. Pilot study responses were not included in the main sample. The reliability test using Cronbach’s Alpha was calculated on 18 items (Section B) about perceived experiences, behaviours, and barriers (reasons) for non-adherence among hypertensive patients and had an alpha value of 0.690. The rest of the items were under Section A, which focused on factual information such as age, sex, occupation, and duration of diagnosis, which do not require consistency testing (DeVellis & Thorpe 2021; Gliem & Gliem 2003; Tavakol & Dennick 2011). For reliability, a Cronbach’s alpha greater than 0.70 is acceptable for Likert scale questions (Bujang, Omar & Baharum 2018) to determine internal consistency and stability. The instrument was also presented to the research supervisor to evaluate its content validity.

Data were collected by using a questionnaire, which was piloted on ten participants, who did not form part of the main study. The research tool was self-administered and handed out in person to the participants. The questionnaire was paper-based; therefore, respondents wrote their answers directly on the paper, and the filled questionnaires were returned to the data collector. The completed questionnaires were kept under lock and key, and the captured data were stored on a password-protected computer.

Collected data were captured into Microsoft Excel 2016 and secured with a password to ensure data integrity. At the end of the study period, the data were cleaned and checked for consistency and imported to the IBM Statistical Package for Social Sciences (SPSS) version 26.0 for analysis (Jarrar et al. 2025). Quantitative variables were effectively summarised by using frequencies and percentages. The results are presented in frequency tables with percentages (see Table 1, Table 2A, Table 2B, Table 3, Table 4 and Table 5). The Chi-square test was used as a bivariate statistical tool to identify which patient characteristics, clinical factors, behavioural practices, and perceived barriers were significantly associated with medication adherence. The binary logistic regression was used to determine the best independent predictors of medication adherence to ensure robust and reliable findings, thereby thoroughly addressing the research questions regarding demographic and behavioural factors influencing medication adherence. The dependent variable was timely medication intake, while independent variables included demographic characteristics and individual practices identified in the questionnaire. All p-values ≤ 0.005 were considered statistically significant (see Table 5, Table 6 and Table 7). The combination of descriptive statistics, Chi-square tests, and binary logistic regression provided a comprehensive approach to analysing the data, allowing for both a detailed description of the sample and a robust examination of the predictors of treatment adherence. This data analysis method ensured that the research questions were addressed thoroughly and that the findings were statistically sound and actionable.

An application for full ethical approval was made to the University of Namibia School of Nursing Research Ethics Committee and ethics consent was received on 18 October 2021. The ethics approval number is SoN 87/2021. An application for full ethical approval was also made to the Ministry of Health and Social Services Health Ethics Committee and ethics consent was received on 27 October 2021. The ethics approval number is 17/3/3/AVN. Additionally, the gatekeepers of the hospital provided further approval and access to participants. Written informed consent was obtained from the participants before the questionnaires were administered. Participants had full rights to refuse to sign the consent or withdraw anytime from the study without penalties. Prior to signing the consent, participants were informed about the purpose, objective, data collection procedure, advantages, and disadvantages of the study. During this study, codes were assigned and used on behalf of participants’ real names to ensure anonymity, and the data collected were solely used for the purpose of the study. Participants had complete privacy and confidentiality of the information they were going to give throughout the interview. Moreover, the completed questionnaires were kept under lock and key to ensure confidentiality, and only the researcher and supervisor had access to the questionnaires. All eligible participants were given equal information about the study and an equal chance to voluntarily participate therein.

The demographic characteristics in Table 1 show that the majority of patients were older adults, with almost half of them being 60 years and above (90 out of 200; 45%). Women made up more than half of the sample (114; 57%), while men represented a smaller share (86; 43%). More than half of the participants were married (102; 51%), while 25% (50) were divorced or widowed, and 24% (48) were single. In terms of education, 20 patients (10%) had no formal schooling, 72 (36%) had only primary education, 70 (35%) completed secondary school, and 38 (19%) had tertiary qualifications. Employment was also skewed, with more than two-thirds unemployed (136; 68%) and only 64 (32%) reporting employment, while 59% reported a family history of hypertension. Importantly, both lower education (χ² = 9.21, p = 0.027) and unemployment (χ² = 6.98, p = 0.031) were significantly associated with poor adherence (Table 6), suggesting that limited literacy and financial hardship can undermine consistent medication-taking.

Table 2A and Table 2B highlights the clinical and behavioural profiles. A large number of patients had been living with hypertension for at least 5 years (82; 41%), while 62 (31%) had lived with it for 2–4 years and 56 (28%) for less than 2 years. Only 94 patients (47%) had controlled BP at the time of the study, while the rest (106; 53%) were uncontrolled. Nearly two-thirds (128; 64%) were taking two or more antihypertensive drugs, while 62 patients (31%) had diabetes as a comorbidity, and another 28 (14%) reported other conditions such as renal problems. Duration of hypertension was significantly associated with adherence (χ² = 8.76, p = 0.013) (Table 6), and logistic regression showed that patients with hypertension for ≥ 5 years had higher odds of adherence (adjusted odds ratio [AOR] = 1.75, 95% confidence interval [CI]: 1.01–3.10, p = 0.047) as shown in Table 7. Diabetes, on the other hand, was associated with lower adherence in the unadjusted model (odds ratio [OR] = 0.56, 95% CI: 0.32–0.94, p = 0.029) (Table 7), although this was not significant in the adjusted model.

Table 2A and Table 2B further reveals worrying trends in the behavioural practices. More than half of the patients admitted that they ‘always’ miss doses (124; 62%) or forget to take their medicine (118; 59%). A considerable proportion reported stopping their medication when they felt well (86; 43%) or when they experienced side effects (92; 46%). Over half (110; 55%) even said that they changed their dosage on their own, while almost half (96; 48%) stopped treatment when their medicine ran out. The Chi-square test in Table 6 confirmed that forgetfulness (χ² = 11.92, p = 0.001) was strongly linked to non-adherence, and logistic regression in Table 7 showed that patients who reported forgetfulness were far less likely to adhere (AOR = 0.39, 95% CI: 0.20–0.76, p = 0.006). This finding makes forgetfulness the single strongest behavioural barrier to adherence in this study.

Table 3 explains in more detail the reasons for non-adherence. The leading cause was forgetfulness, reported by 112 patients (56%). Side effects were the second-most common reason (88; 44%), followed by financial constraints (74; 37%), lack of transport (64; 32%), and lack of knowledge about hypertension (52; 26%). Nearly a quarter (48; 24%) mentioned long waiting times, while 36 patients (18%) turned to traditional medicine, and 30 (15%) reported poor family support. Financial constraints were statistically significant (χ² = 9.55, p = 0.002), and patients facing financial barriers had much lower odds of adherence (AOR = 0.40, 95% CI: 0.21–0.82, p = 0.009) (Table 7). Similarly, side effects (χ² = 7.34, p = 0.007) were a major deterrent (Table 6), with patients who reported side effects showing less than half the odds of adhering to medication (AOR = 0.42, 95% CI: 0.21–0.89, p = 0.011) (Table 7). Transport barriers also mattered (χ² = 6.02, p = 0.014), cutting adherence odds by more than half (AOR = 0.44, 95% CI: 0.22–0.88, p = 0.019). These findings reinforce how practical obstacles and the unpleasant realities of treatment can lead patients to default.

Health system and patient-related issues in Table 4 confirm these struggles. Just over half of the patients (104; 52%) said that long waiting times discouraged them from using services, while 92 (46%) reported that drug stock-outs limited adherence. Almost as many (86; 43%) felt that communication with health providers was poor. Encouragingly, however, 128 patients (64%) felt that family support helped them take their medicine. This finding demonstrates that the health system poses real access barriers, but social support can play an important protective role in adherence.

The statistical tests in Table 5, Table 6, and Table 7 put all of these factors into perspective. Among demographic and socioeconomic factors, education, employment, and marital status showed important associations, although only employment remained significant after adjustment (AOR = 0.52, p = 0.041). Among clinical factors, longer duration of hypertension was associated with better adherence, while diabetes comorbidity weakened adherence in the unadjusted analysis. Among behavioural and patient-reported factors, the most powerful predictors were forgetfulness (AOR = 0.39, p = 0.006), side effects (AOR = 0.42, p = 0.011), financial constraints (AOR = 0.40, p = 0.009), and transport barriers (AOR = 0.44, p = 0.019).

When the findings are considered together, the study paints a picture of patients who are willing to treat their condition but are hindered by low socioeconomic status, daily struggles with remembering and tolerating medication, and systemic barriers such as transport and long queues at facilities. Employment and longer experience with hypertension increased adherence, showing that stability and adaptation make a difference. By contrast, forgetfulness, side effects, financial strain, and transport difficulties cut adherence dramatically, sometimes halving the odds. The results emphasise that in order to improve adherence, a multifaceted approach is required. This approach includes strengthened patient education, designed reminder systems to counter forgetfulness, counselling offered and medication adjustments to address side effects, and reduced financial and logistical barriers that patients face. Family support also emerges as a powerful protective factor, which suggests that when households and communities are involved, treatment adherence could be made more sustainable.

The study was cross-sectional, so causal direction cannot be proven; for example, people who are more adherent may be more likely to remain employed, or conversely, unemployment may worsen adherence. Self-reported adherence and reasons for non-adherence are vulnerable to recall and social-desirability biases; nonetheless, the convergence of self-report with objective clinical markers such as uncontrolled BP (106; 53%) supports the overall validity of the signal. The sample is drawn from a single regional hospital, and patterns may differ in urban centres or other Namibian regions; however, the similarity between our findings and other Namibia studies (Nakwafila et al. 2022) and LMIC reviews suggests that many of the determinants identified are not unique to Kavango East. Future research should test bundled interventions (for example, reminders plus decentralised dispensing plus CHW follow-up) in implementation trials that measure both adherence and clinical outcomes such as BP control and cardiovascular events.