Patient-Drug Related Factors Associated with Nonadherence to Chronic Treatment in Patients Attending a Primary Care Setting in South Africa

Abstract

Background: Medication nonadherence among patients with chronic diseases represents a major challenge in healthcare systems worldwide and is associated with poor clinical outcomes, increased hospitalizations, and higher healthcare costs. Patient-drug related factors such as knowledge of treatment, beliefs about medication, and the experience of side effects may significantly influence adherence behaviour. Methods: A cross-sectional quantitative study was conducted among 80 patients receiving treatment for chronic conditions at a primary healthcare facility in South Africa. Data were collected through face-to-face interviews using a standardized questionnaire that assessed demographic characteristics and patient-drug-related factors potentially associated with medication adherence. Statistical analysis was performed using IBM SPSS Version 30.0.0.0 (172). Descriptive statistics were used to summarize participant characteristics, while inferential analyses, including chi-square tests and Fisher’s exact tests, were applied to determine associations between demographic variables, patient-drug related factors, and medication nonadherence. Results: The majority of participants were female, aged between 41 and 50 years, single, unemployed, and had completed secondary education. Most participants lived in rural areas, and HIV/AIDS was the most commonly reported chronic condition. Significant associations with medication nonadherence were identified for the experience of medication side effects and inadequate knowledge about treatment. These factors demonstrated moderate effect sizes and suggest that both clinical and educational aspects of treatment may influence adherence behaviour. Conclusions: Patient-drug related factors, particularly medication side effects and insufficient knowledge regarding treatment, play a significant role in medication nonadherence among patients with chronic conditions in primary care settings. Interventions aimed at improving patient education, counselling regarding medication side effects, and strengthening patient-provider communication may help improve adherence and treatment outcomes.

1. Introduction

Nonadherence to chronic diseases is the failure to follow prescribed treatment regimens, which leads to poor health outcomes, increased hospitalizations, and high healthcare costs. Patel et al. (2025) [1] state that nonadherence is affected by patient-related components (for example, knowledge and motivation), therapy-related factors (such as side effects), and health system-related issues (such as cost and access). Improving adherence is critical for managing chronic conditions and involves tailored treatment plans, strong patient-provider communication, and social support.

Chapman et al. (2025) [2] say that nonadherence means not following a prescribed treatment plan for a chronic condition, which includes taking medication, following dietary guidelines, or attending appointments. The World Health Organization (WHO) (2003) [3] estimates that about 50% of patients with chronic illnesses in developed countries did not comply with their medication plan in 2002, and rates have not improved significantly since then.

Patient-drug related factors associated with nonadherence to medication for chronic conditions include patient beliefs about the disease and its treatment, forgetfulness, negative attitudes, a lack of knowledge about the condition, and experience with side effects (Cardenas et al., 2021) [4]. According to Kassaw et al. (2024) [5], the complexity of the medication regimen itself, such as taking multiple medications, can also contribute to patients’ difficulty adhering to their treatment plan.

A chronic condition is a long-lasting health issue that typically lasts for at least a year, may get worse over time, and often requires ongoing medical attention (WHO, 2025) [6]. Unlike acute illnesses that come and go, chronic conditions are persistent and can limit daily activities, though they can often only be managed, rather than cured. Examples of chronic conditions include heart disease, diabetes, arthritis, and cancer. Chronic conditions, also known as noncommunicable diseases (NCDs), tend to be of long duration and are the result of a combination of physiological, genetic, behavioral, and environmental factors.

From an educational point of view, Al-Qerem et al. (2021) [7] say that patient-drug related factors with nonadherence to chronic diseases include a lack of knowledge about the disease and treatment, misunderstanding of the medication regimen, and negative attitudes or beliefs about the effectiveness of the medication. Factors like a desire for more information, forgetting doses due to complexity, and a lack of understanding of side effects also contribute to nonadherence.

Nonadherence to chronic disease treatment is a worldwide challenge to both patients and healthcare providers in their management. Therefore, the following key points were considered in this research gap: patient and drug-related factors related to nonadherence to chronic medication.

1.1. Statement of the Problem

Nonadherence to chronic medication is a significant problem that negatively affects health outcomes, as patients frequently fail to follow their prescribed regimens due to a variety of patient-related factors. As stated by Religioni et al. (2025) [8], these factors include a lack of understanding about the condition and the medication, negative beliefs about the effectiveness of the medication, psychological issues like depression and anxiety, and practical barriers such as forgetfulness, competing life demands, or misperceptions about the need for long-term medication. This gap between prescribed and actual treatment can lead to disease progression, increased complications, and a reduced quality of life, indicating a critical need to identify and address the specific patient-related barriers to adherence.

Nonadherence to chronic treatment can result in more frequent hospitalizations. Horvat et al. (2024) [9] conclude that identifying the specific patient-related barriers is necessary for developing tailored interventions that can ameliorate compliance effectively and ultimately lead to better treatment outcomes.

1.2. Rationale of This Study

As concluded by Kvarnström et al. (2021) [10], a study on patient-drug related factors and chronic disease adherence is vital because it helps identify barriers to treatment, improves patient outcomes, and reduces healthcare costs. Understanding patient beliefs, knowledge, and regimen complexity is necessary for devising intended strategies to improve adherence and, consequently, patient health, longevity, and quality of life. No other study on patient-drug related factors related to nonadherence to chronic medication has been carried out in this setting.

1.3. Research Aim and Research Questions

The aim of this study was to assess patient-drug-related factors associated with nonadherence to chronic medication in a primary healthcare setting in the Eastern Cape, South Africa.

This was attained through the following objectives, which were to:

• Establish the prevalence of demographics associated with patient-drug factors.
• Determine associations between demographics and nonadherence with chronic medications.
• Identify chronic conditions experienced by participants and nonadherence status.
• Identify patient-drug related factors contributing to nonadherence with chronic medications; and
• Assess drug-related factors associated with nonadherence to chronic medication.

1.4. Literature Review

This section starts with an outline of the literature analysis and its objectives. Ebidor and Ikhide Zarei (2025) [11] say that a literature review is a synthesis and comprehensive analysis of an existing study on a specific topic. It combines the findings and conclusions from various sources to provide a complete understanding of the topic, setting up the basis for the research question and later original research. The literature review assesses sources; it goes beyond a simple, annotated bibliography, giving a nuanced and in-depth analysis of the existing understanding of the topic. This section analyses the current literature related to both global and local perspectives on patient-drug related factors associated with nonadherence to chronic treatment.

Furthermore, this literature review focuses on the related factors of nonadherence of patients with chronic diseases attending a primary healthcare setting. It explores patient-drug factors leading to nonadherence. The association between drug-related factors and nonadherence with chronic conditions is determined. Furthermore, the literature review has contributed to a better understanding of strategies that will be established to prevent these factors, thus avoiding complications in these patients. This will improve the outcome of the treatment of these patients.

The findings of Religioni et al. (2025) [8] reveal that poor compliance worsens the risk of complications, disease progression, and high healthcare costs. Conversely, improved compliance improves disease control, fewer complications, and improved patient quality of life. Patient education, digital health tools, and simplified treatment plans are effective strategies to address hindrances to treatment adherence.

Patient education helps individuals to understand their treatment, while simplified regimens reduce complexity, and digital tools like reminder apps can improve self-efficacy and consistency. Furthermore, Katantha et al. (2025) [12] state that the role of healthcare professionals is underscored as fundamental, with their effective communication, continuous support, and efforts to build patient trust being necessary to improve adherence.

According to Zwide et al. (2025) [13], treatment nonadherence is an important public health concern and is high among mental healthcare users. Approximately 65% of patients with severe mental illness do not comply with their prescribed treatment. Drug nonadherence may worsen mental illness and lead to poorer clinical outcomes, including rehospitalization rates, because of frequent relapses as well as a long time to remission, which may contribute to higher costs of care.

Smith et al. (2020) [14] report that a systematic review found that factors such as a lower quality of life, current and past experiences with symptoms, psychological factors, and explicit suggestions can increase side-effect expectations from medical interventions. These heightened expectations are linked to a greater likelihood of experiencing side effects and decreased adherence to treatment. The review suggests that to reduce unrealistic expectations, communication about medical interventions should include numerical data on side effect incidence.

According to Hisan et al. (2025) [15], cancer, diabetes, hypertension, cardiovascular diseases, and chronic respiratory conditions are noncommunicable diseases (NCDs) that are the leading global causes of morbidity and mortality. With these conditions, depression frequently co-occurs and may significantly reduce medication adherence, thereby aggravating health outcomes. Their narrative review examines the association between depression and medication adherence in patients with NCDs. It also culminates the existing challenges in managing this comorbidity and analyzes potential interventions for enhancing adherence outcomes.

A study performed by Rafhi et al. (2024) [16] reveals that medication use in older adults is increasing; therefore, reducing the risk of suboptimal medicine use is imperative in achieving optimal therapeutic outcomes. Research suggests that factors such as beliefs about medicines and personal beliefs may be related to nonadherence and inappropriate medicine use.

2. Methodology

2.1. Study Design

A cross-sectional study, quantitative in nature, using questionnaires with face-to-face interviews was conducted in 80 patients with chronic conditions who attend a primary healthcare setting in South Africa. These illnesses are HPT, DM, TB, HIV/AIDS, asthma, epilepsy, and CVDs. The design was chosen because quantitative research is a systematic, objective, and an formal process used to explain and examine associations and examine cause and effect interaction among variables. Furthermore, the design was selected to meet the objectives of the study, which were to evaluate patient-drug related factors associated with nonadherence to chronic medications in patients attending a primary healthcare setting in Mthatha in the Eastern Cape, South Africa.

2.2. Study Setting

The research was conducted at Gateway clinic, located in Mthatha in the Eastern Cape (EC), South Africa. The clinic serves nearby areas like Old Payne, Zimbane Valley, South Ridge, and others. Gateway Clinic is the type of facility that is focused on outpatient services, meaning patients can go home after they receive care. It is also a public service Centre, meaning that one does not have to make an appointment to go there.

2.3. Study Population

A population, as defined by Willie (2024) [17], is all the components (including objects, events, and individuals) that satisfy the sample criteria of a project. The study population is composed of chosen patients on chronic treatment attending the healthcare clinic. The patients were people attending the Gateway clinic who had chronic illnesses such as HIV/AIDS, TB, HPT, DM, asthma, epilepsy, and cardiovascular diseases.

2.3.1. Inclusion Criteria

The study included:

• All patients diagnosed with a chronic condition and who were getting their chronic medication from the Gateway clinic for the period June 2022 to August 2022; and
• All patients 18 years and older, irrespective of gender, with chronic conditions were included.

2.3.2. Exclusion Criteria

As described by Patino and Ferreira (2018) [18], exclusion criteria are characteristics or factors that disqualify potential participants from a study, even if they meet the inclusion criteria. These criteria are established to ensure the safety and ethical treatment of participants, improve the scientific rigor by reducing confounding variables, and define the specific characteristics of the study population. Therefore, the following patients were excluded from the study:

• All the patients who have not been collecting their medications outside the defined period, June to August 2022.
• Patients reluctant to take part in the study; and
• Patients who were too severely ill to answer interview questions or did not have a care assistant, and patients who do not have chronic conditions.

2.4. Sampling Strategy

A simple random sampling method was used to choose patients with the following chronic diseases: HPT, DM, TB, HIV/AIDS, asthma, epilepsy, and CVDs. The selection of patients for this study was unbiased and was based on the ideal benefits of this study.

Clinical samples may be unbiased due to the following factors:

Broad-Based Attendance: If a clinic serves a wide geographical area and is the primary care provider for a diverse local population (e.g., a public community health center), the demographic, racial, and socioeconomic characteristics of patients may reflect the surrounding community.

Nature of the Condition: If the disease or condition being studied is not strongly tied to socioeconomic status, education, or specific demographics, the sample may not be inherently biased.

Universal Health Access: In regions with universal healthcare or where the study is conducted within a large, inclusive public hospital system, economic barriers to entry are lower, resulting in a more representative sample than, for example, a private specialty clinic.

Sample Size and Strategy

I hereby acknowledge that the sample size, which is 80, is convenience-based.

A convenience sampling was used for this, which is a non-probability method that relies on the easiest-to-reach subjects. The patients were consecutively approached during clinic days as they came in for their reviews with the doctors or to refill their prescriptions. The target population was gender-range from 18 to above 50 years, male and female, and other demographics, and on chronic treatment. The detailed, transparent reporting on recruitment is crucial for validating the study and assessing external validity to ensure reproducibility.

I, however, acknowledge the following: This is a limitation in that the sample is not representative of the broader population. Selection bias: the potential for bias, as participants are selected based on availability rather than random chance. The scope of Generalization: the findings from this study can only be generalized to the specific, locally accessible population. This transparency reporting is also included in the limitation section.

2.5. Ethical Considerations

The protocol was authorized by the Faculty of Medicine & Health Sciences Biosafety Human Research Ethics Committee (HREC) of Walter Sisulu University, and the study was conducted in compliance with the Declaration of Helsinki. The research identification number was 080/2020 and was issued on 25 August 2022. Furthermore, permission was sought from WSU Gatekeepers, Tambo Municipality, and KSD under which the clinic functions. Finally, permission was sought from the manager of the clinic where the data was gathered.

The project team informed the participants of the goals and advantages of the study using a participant information sheet prior to data collection. Then the participants were asked to sign a written consent form, which was optional, to participate in the study. It was made clear to the participants that if at any moment they felt like not continuing to take part in the study project, they were free not to proceed. The researcher and data collectors gave a guarantee to the participants of anonymity and confidentiality of the information as stipulated in the participants’ consent form.

According to ethical rules, a written consent is a documented agreement where an individual voluntarily agrees to participate in a study or procedure after fully understanding its nature, purpose, and potential risks and benefits (Shah et al., 2025) [18]. Furthermore, while written informed consent is the standard and generally required procedure for most human research studies, it is not always mandatory and can be optional or waived in specific, ethically approved circumstances.

2.6. Data Collection Tool and Data Acquisition

2.6.1. Data Collection Tool

A previously validated standard questionnaire was used. A questionnaire is a research tool made of a set of questions designed to gather data. In this study, demographic characteristics of the patients were obtained using face-to-face interviews. The questions encompassed both closed-ended questions with pre-determined responses and open-ended questions that allowed respondents to give comprehensive, unscripted answers. All questionnaires were interpreted into the local vernacular called “IsiXhosa”.

2.6.2. Data Collection

The data was collected from 80 patients at the Gateway clinic, obtaining their socio-demographics like gender, age, marital status, educational background, employment status, salary earned, form of settlement, means of transport, number of people in the household, and the chronic conditions from which they are suffering.

Medication nonadherence in a study of this kind that evaluates patient-related factors is typically measured using validated self-reported questionnaires (e.g., MMAS-8, MARS-5. The Morisky Medication Adherence Scale (MMAS-4 or MMAS-8) is a validated, copyrighted, self-report questionnaire used to assess patient adherence to chronic medication regimes (e.g., hypertension, diabetes) (Nassar et al. 2022) [19]. It identifies reasons for nonadherence, with scores indicating high (8), moderate (score 6 to <8) or low (<6) adherence. It is easy-to-administer questionnaires that classify adherence based on total scores.

MMAS-8, the Morisky Medication Adherence Scale (MMAS-4 or MMAS-8) is a validated, copyrighted, self-report questionnaire used to assess patient adherence to chronic medication regimes (e.g., hypertension, diabetes). It identifies reasons for non-adherence, with scores indicating high (8), moderate (6 to <8), or low (<6) adherence. MMAS-8 is the enhanced 8-item version, which is more commonly used to assess adherence in chronic conditions.

The 8-item Morisky Medication Adherence Scale (MMAS-8) is a self-reported, validated instrument used to measure medication-taking behavior, with total scores ranging from 0 to 8. It uses a combination of dichotomous (Yes/No) and Likert scale questions to categorize patients into low, medium, or high adherence levels. Item 8 is a 5-point Likert scale measuring the frequency of difficulty remembering to take medications. Scores are assigned as follows: Never/Rarely = 1, Once in a while = 0.75, Sometimes = 0.5, Usually = 0.25, All the time = 0. Total Score: The sum of all items ranges from 0 to 8, with higher scores indicating higher adherence.

Cut-off Points. Based on the total score, patients were categorized into three levels of adherence: High Adherence: Score of 8. Medium Adherence: Score of 6 to <8. Low Adherence: Score of <6.

Based on the standard, validated MMAS-8 and its implementation in studies, the 8 items (7 yes/no questions and 1 Likert-type question) are part of the validated tool itself, not supplementary items created for specific studies. The tool was translated into isiXhosa, and the present the translation is presented as a practical adaptation.

2.6.3. Validity of the Data Collection Tool (Mo et al., 2023) [20]

The validity of a questionnaire refers to how accurately it measures what it is intended to measure. It is a crucial aspect of questionnaire design, ensuring that the questions effectively capture the intended concepts and that the collected data reflects the reality it aims to represent.

2.6.4. Reliability of the Data Collection Tool (Mo et al., 2023) [20]

The reliability of a questionnaire refers to its consistency in producing the same results when used repeatedly under similar conditions. A reliable questionnaire will yield similar responses when administered multiple times to the same individuals or to separate groups with comparable characteristics. This consistency is necessary for making certain that the validity and trustworthiness of the data collected.

2.7. Data Protection and Management

The information collected was kept by two people; in case one person lost it, it would be available to the other person. Only eleven people had access to the data. To protect the data from being lost, it was saved on Google Drive, where it could be retrieved at any time it was needed. To prevent accessibility from unauthorized persons and maintain the confidentiality of data, it was stored in a computer, protected by a passcode. Physical security measures were set, such as storing copies of research in locked cabinets.

2.8. Data Cleaning and Entry

The data was entered into SPSS. Data cleaning was done through thoroughly scanning data for errors, duplications, and removing all irrelevant and incorrect information. All the faulty questions and consent forms filled in by participants were disposed of. We checked whether all questions were answered and checked the response rate.

2.9. Data Analysis

When all the data were collected, variables were checked for completeness and consistency using a mechanistic analysis of data. The pre-coded data were captured into SPSS. Descriptive statistics were used to narrate patient attributes. Categorical variables were reported using frequencies and/or percentages. The association between participants’ attributes, chronic conditions, and patient-drug factors and nonadherence to chronic medications was determined by the Pearson chi-square test. A p-value of less than 0.05 was regarded as statistically significant. Furthermore, effect sizes in terms of Cramer’s V were also determined to measure the strength of the association (0–1). Thus if, V = 0.00–0.10 it is small, then 0.10–0.30 medium and 0.30–0.05 large, and >0.05 large. Furthermore, variables with very small or zero cell counts, where chi-square assumptions may be violated, Fisher’s Exact Tests were run. The data is presented in the form of tables, followed by a discussion of each graph.

3. Results

3.1. Demographic Characteristics of Patients

A total of eighty questionnaires were collected, giving a response rate of 100%, meaning that every single person who was invited to participate in the research study completed it. This was possible in this study because of the controlled environment where the project was conducted, and the small number of participants, which was eighty. Adherents were 47 (58.75%), while 33 (41.25%) were nonadherent.

Females represented 56.25%, and males 43.75% of the participants included in the study. Males 17 (51.52%) were more nonadherent. Regarding marital status, those without partners were the majority (61.25%), and nonadherent accounted for 22 (66.67%). In terms of age groups, participants belonging to the 41–50 years group comprised the highest percentage (31.25%), followed by the 31–40-year-old group and >50 years group, both accounting for (21.25%) and lastly 16.25% for the 18–30-year-old group. The age group with the highest number (14; 44.44%) of nonadherent was in the 31–40 age group.

While most of the participants (75.00%) had received a secondary/tertiary education, others (25.00%) only received a primary education, with the majority (27; 81.82%) of nonadherent having received secondary/tertiary level of education. More than half (60.00%) of the patients were unemployed and nonadherent, with 24 (72.73%). For income status, the majority (72.50%) earned less than R 5000 per month, and more nonadherent patients accounted for 29 (87.88%).

More than 50% (42, 52.50%) stay in urban areas, and 57.58% more nonadherent. Less than half (43.75%) live 5–10 people per household; those who live less than 5 people per household were more nonadherent, accounting for 17 (51.52%). In terms of chronic conditions, most of them, 38 (47.50%), were HIV/AIDS, and the most nonadherent were 20 (60.61%), as reflected in

Table 1. Demographic characteristics of participants and level of adherence and nonadherence.

Gender	Frequency (n)	Percentage (%)	Adher.	AR.	Nonadher.	AR.	X2	p Value	Cramer’s V
Male	35	43.80	18	−0.8	17	1.2	0.131	1.376	0.24
Female	45	56.30	29	−0.5	16	−1.2
Total	80	100.00			33
Marital status							4.083	0.397	0.23
Without Partner	49	61.25	27	−0.8	22	0.8
With Partner	31	38.75	20	−0.5	11	0.5
Total	80	100.00	47		33
Educational level							7.136	0.028	0.29
Primary	20	25.00	15	1.7	5	−1.7
Secondary/Tertiary	60	75.00	32	−2.4	27	2.4
Total	80	100.00	47		33
Age group							7.994	0.046	0.32
18–30	13	16.25	8	0.2	5	−0.2
31–40	21	26.25	7	−2.8	14	2.8
41–50	25	31.25	18	1.6	7	−1.6
>50	21	26.25	14	0.9	7	−0.9
Total	80	100.00	47		33
Employment status							0.390	0.016	0.36
Unemployed	48	60.00	24	−1.9	24	1.9
Employed	32	40.00	23	2.3	9	−2.3
Total	80	100.00	47		35
Income status							12.740	0.002	0.39
≥5000	58	72.50	29	−2.6	29	2.6
5001 ≥ 10,000	22	27.50	18	−1.7	4	1.7
Total	80	100.00	47		33
Residential area							0.580	0.446	0.09
Urban	58	52.50	23	−0.8	19	0.8
Rural	22	47.50	24	0.8	14	−0.8
Total	80	100.00	47		33
People per household							50.997	<0.001	0.73
<5	29	36.25	12	−2.4	17	2.4
5–10	35	43.75	35	6.6	-	−6.6
>10	16	20.00	-	−5.3	16 5.3
Total	80	100.00	47		33
Chronic conditions							11.156	0.084	0.37
HIV/AIDS	38	47.50	18	−2.0	20	2.0
TB	10	12.50	9	2.1	1	−2.1
HPT	13	16.25	8	0.2	5	−2
Asthma	10	12.50	4	−1.3	6	1.3
Others	9	11.25	8	3.3	1	−3.3
Total	80	100.00	47		33

Adher. = Adherence, NonAdher. = Nonadherence, AR = Adjusted Residual, Threshold > 1.96. The bold numbers shows p values that are statistically significant since they are less than 0.05.

.

3.2. Association Between Demographic Characteristics and Nonadherence

Pearson Chi-square tests were used to establish if there is a statistically significant association between demographic characteristics and nonadherence, and Effect size (Cramer’s V) to determine the strength of the association. There was a small statistically significant association between educational level and nonadherence (X² = 7.136, p = 0.028, V = 0.29), with age group (X² = 7.994, p = 0.046, V = 0.32), employment status (X² = 0.390, p = 0.016, V = 0.36), income status (X² = 12.740, p = 0.002, V = 0.39), people per household (X² = 50.997, p = 0.001, V = 0.37) were all statistically significant with medium effect size. The rest of the demographic characteristics were not statistically significant. Adjusted Residual test (AR) was done, and if AR is >1.96 (Threshold), the deviation is statistically significant, as in educational level, employment status, age group, residential area, and no of people per household, as shown in Table 1.

3.3. History of Chronic Conditions with Adherents and Non-Adherents

In terms of the history of chronic diseases, 47.50% had HIV/AIDS, followed by 16.25% having HPT, while both TB and asthma accounted for 12.50% each, 7.50% had DM, 2.50% had epilepsy, and 3.75% had other conditions. The majority (60.61%) of nonadherent were those people with HIV/AIDS, followed by 18.18% with asthma, then 15.15% with HPT, 3.03% for both TB and DM. Neither epilepsy nor heart disease had any nonadherent. A p value of 0.084 was obtained, which is >0.05, indicating that there is no statistically significant association between chronic conditions and nonadherence, and an effect size of 0.37 means there is a moderate practical effect size, as shown in

Table 2. The prevalence of chronic conditions and association between the level of adherence.

Chronic Condition	Total		Adherent		Effect Size
					X2	p	(Cramer’s V)
	(n)	(%)	(n)	(%)	(n)	(%)	11.156	0.084	0.37
HIV/AIDS	38	47.50	18	38.30	20	60.61
Hypertension	13	16.25	8	17.02	5	5.15
Tuberculosis	10	12.50	9	9.15	1	3.03
Asthma	10	12.50	4	8.51	6	18.18
Diabetes	6	7.50	5	10.64	1	3.03
Epilepsy	2	2.50	2	4.26	-	-
Heart disease	1	1.25	1	2.13	-	-
Total	80	100.00	47	100.00	33	100.00

.

3.4. Identify Patient-Related Factors of Patients with Chronic Conditions

Patients were interviewed about the reasons among those who reported skipping their chronic treatment. Thirty-three (41.35%) responded yes, and 47 (58.75%) responded no. Then they were asked for reasons among those who skipped their chronic medications, and gave reasons as indicated in

Table 3. Reasons among those who reported skipping medication (n = 33).

	Reason Among Those Who Reported Skipping Medication	(n)	(%)
1	No knowledge about the importance of taking meds	8	10.00
2	Side effects	6	7.50
3	Forgot	5	6.25
4	Did not have money to go to the clinic	4	5.00
5	Felt better	3	3.75
6	In denial	3	3.75
7	Stigma	2	2.50
8	Did not get medication at the clinic	2	2.50
9	Total	33	41.25
10	Missing	47	58.75
11	Total	80	100.00

.

3.5. Associations Between Patient-Drug Related Factors Associated with Nonadherence to Chronic Treatment

Associations between patient-drug-related factors and nonadherence to chronic treatment were determined using chi-square tests and Effect Sizes. Appropriate statistical testing for sparse data was done using Fisher’s Exact Test. A value of p ≤ 0.05 was considered statistically significant, and Effect Sizes (Cramer’s V) were used to determine the strength of the associations (0–1) as follows: 0.00–0.10: negligible, 0.10–0.30: small/moderate, 0.30–0.50: moderate/large, and 0.50: large as reflected in

Table 4. Association between patient-drug related factors and nonadherence to chronic treatment.

Patient-Drug Related Factor	Adherence	Nonadherence		Pearson Fisher’s Exact Effect Size
No Knowledge of Importance	Yes	No	Total	X2	Test (p Value)	(Cramer’s V)
medication adherence	47	25	75	12.660	<−0.001	0.36
	0	8	5
	47	33	80
Side effects	0	6	6	9.238	0.004	0.34
	47	27	74
	47	33	80
Forgot to take the medicine	0	5	5	7.595	0.010	0.31
	47	28	76
	47	33	80
I felt better	0	3	3	4.439	0.066	0.24
	47	30	77
	47	33	80
I was in denial	0	3	3	4.439	0.066	0.24
	0	30	77
	0	33	80
Had stigma	0	2	2	2.922	0.167	0.19
	47	31	78
	47	33	80
Did not get medication from the clinic	0	2	2	2.922	0.087	0.19
	0	31	78
	0	33	80
Did not have money to go to the clinic	0	4	4	5.997	0.014	0.27
	47	29	76
	47	33	80

The bold shows values of p that are statistically significant.

.

Patient-drug related factors identified to be statistically significant with nonadherence to chronic treatments are: patients having no knowledge of the importance of adhering to treatment with (X² = 12.660, p = < 0.001, V = 0.36) moderate effect size, side effects (X² = 9.238, p = 0.002, V = 0.34) moderate effect size; forgetfulness (X² = 7.595, p = 0.006, V = 0.31) moderate effect size, patients felt better (X² = 4.439, p = 0.035, V = 0.24) small effect size, being in denial (X² = 4.439, p = 0.035, V = 0.24) small effect size, and patients who did not have money for transport to the clinic (X² = 5.997, p = 0.014, V = 0.27) small/moderate association (Table 4).

3.6. Discussion

The aim of this study was to assess patient-drug-related factors associated with nonadherence to chronic medication. Adherents accounted for 47 (58.75%), and 33 (41.25%) nonadherent. In this study, although females 45 (56.30%) were the majority, males were the most nonadherent, accounting for 17 (51.51%). This agrees with research conducted by Schutt-Cerdan et al. (2025) [21], who say that being male (OR = 28.3, 95% CI: 5.3–149.7) was significantly associated with a higher likelihood of nonadherence to treatment. Mokoena et al. (2024) [22] say that studies across various conditions, including diabetes and hypertension, consistently show that men are less likely to follow prescribed treatment plans compared to women.

In terms of the age range, 41–50 was the majority with 25 (31.21%). This age group was also the most nonadherent. However, the age group of 31–40 was more nonadherent. This is supported by Stanly et al. (2025) [23]. Some studies indicate that younger adults often have higher rates of nonadherence, while others show that adherence peaks by middle age (e.g., 41–69 years), before declining in older adults. According to the age of the patients, the results of associations with nonadherence to chronic treatment reveal a p value that is statistically significant (X² = 7.994, p = 0.046, V = 0.29) with a small/moderate effect size.

Results from this study revealed secondary level of education was obtained by most participants, with 51 (53.70%) having a secondary level of education. At the same time, these participants had the most nonadherents with (X² = 7.136, p = 0.028, V = 0.29), which is statistically significant with a small effect size. According to Zhu et al. (2023) [24], their results concluded that most nonadherent participants in their study had a secondary level of education, highlighting the complex and multi-factorial nature of medication nonadherence.

Findings from this study revealed that most participants were unemployed 45 (60%). This agrees with Zwide et al. (2025) [13] who say that medication nonadherence was high, with a significant majority of its participants (93.3%) being unemployed, and it was statistically significant with nonadherence with X² = 10.390 and p = 0.016, V = 0.36 with a medium effect size.

The majority of 58 participants (72.50%) in the study earned R5000 with a significant value of (X² = 12.740, p = 0.002, V = 0.39) with a medium/large effect size. According to Appiah et al. (2023) [25], a specific study on nonadherence to anti-tuberculosis treatment in the Ashanti region in Ghana, in a qualitative study, only 5% of the nonadherent participants in that study earned more ZAR 5000.

The number of people per household has a significant association with nonadherence. This is confirmed by the results of the Pearson chi-square test, with (X² = 50.997, p < 0.001, V = 0.37) in cases of less than 5 people per household. This is backed by the results from a study performed by Katende-Kyenda (2025) [26], that fewer than five people per household was a significant factor that made nonadherence to chronic treatment more likely. This implies that individuals in smaller households have less access to social and family support systems, which are crucial for medication adherence.

Results from this study on the chronic conditions revealed that HIV/AIDS is currently higher, accounting for 18 (20%). At the same time, this is the condition with the highest number of nonadherent. This affirmation is backed by research, performed by Mashele et al. (2025) [27], which consistently identifies high rates of nonadherence to antiretroviral therapy (ART), due to a complex mix of patient, drug, and systemic factors.

Patient-drug related factors associated with nonadherence to chronic conditions in this study were: poor knowledge regarding the significance of adhering to their chronic medications (X² = 12.660, p < 0.001, V = 0.36), which is statistically significant with a medium effect size. As emphasized by Zwide et al. (2025) [13], poor knowledge is a significant barrier, as it directly correlates with a higher likelihood of nonadherence. Patients with good medication knowledge are significantly more likely to comply with their medication plans.

Other factors identified in this study include side effects, with results that are statistically significant (X² = 9.239, p = 0.004, V = 0.34) with a medium effect size. According to Tadesse et al. (2025) [28], patients experiencing medication side effects are another significant factor, with patients who have side effects having higher odds of not adhering to their medication. Participants who were experiencing side effects perceived stigma from psychotropic medication had 2.84 (CI: 2.05, 3.93) and 2.41 (CI: 1.87, 3.10) times more odds of nonadherence, as opposed to participants who did not experience side effects and perceived stigma, respectively.

Four patients (5.00%) also indicated that they defaulted on taking their medications, because they did not have money to go to the clinic (X² = 5.997, p = 0.026). Patients defaulting on their medication due to a lack of money for clinic visits is a significant issue, as stated by Paundi et al. (2024) [29], with the cost of transportation being a major barrier. This financial strain can lead to a complete lapse in medication, resulting in negative health outcomes, such as increased costs, drug resistance, and the progression of illness.

Patient-related factors associated with nonadherence to chronic medication were 3 (3.80%), as some patients said they felt better (X² = 4.439, p = 0.066, V = 0.24). Feeling better is a common patient-related factor associated with nonadherence to chronic medication, as the relief of symptoms can lead individuals to assume incorrectly that they are cured or no longer need the treatment. As emphasized by Taderera (2025) [30], feeling better can indeed lead to medication nonadherence and stopping treatment.

Denial is a well-known patient-related factor related to nonadherence to chronic medication. This is supported by statistically significant results from this study (X² = 4.439, p = 0.066, V = 0.24). Konkor et al. (2025) [31] say that patients in denial may struggle to accept their diagnosis, leading them to minimize the severity of their condition, believe the medication is unnecessary, and avoid engaging with treatment plans altogether.

Patients in this study expressed that they defaulted on their chronic medications due to stigma, though results obtained were not statistically significant (X² = 2.922, p = 0.167). Stigma is a significant patient-related factor that contributes to nonadherence to treatment for chronic conditions. This can be due to internalized self-stigma, which leads to feelings of shame and a reluctance to follow medical advice, or external social stigma, which can cause patients to hide their illness or feel shame when taking medication. This is supported by Al-Rajhi and Alqassim (2025) [32] when they say that stigma impacts adherence through internalized stigma: patients may feel a sense of shame, hopelessness, or isolation, which can lead them to believe that their illness is a personal failing. This can decrease their self-efficacy and lead to a reluctance to follow through with treatment plans. Furthermore, a study concluded by Abdisa et al. (2020) [33] that fear of social stigma means that patients may fear being judged or discriminated against by others if their condition becomes known. This can lead them to avoid taking medication in front of others or to discontinue treatment altogether.

Stigma in diabetes is often termed the “blame and shame disease,” arising from the incorrect belief that individuals are solely responsible for their condition through poor diet, lack of exercise, or laziness. Stigma leads to barriers like:

Avoidance of Self-Management: Fear of being judged leads people to hide their condition. Many skip or delay necessary actions, such as checking blood sugar or injecting insulin in public, to avoid unwanted attention or comments.

Non-Disclosure and Isolation: Patients may avoid disclosing their status to friends, family, or employers, reducing their social support network, and potentially leading to delayed or improper treatment.

Mental Health Impact: Stigma contributes to higher levels of depression, anxiety, diabetes distress, and lower self-esteem, which directly impair a person’s motivation to manage their health.

3.7. Recommendations

Recommendations for a study on patient-drug related factors of chronic treatment non-adherence include:

From the research findings that highlights the novelty in understanding how patient beliefs, perceptions, and knowledge gaps specific to this primary health care setting in a low–middle income country drives non-adherence, as there’s limited prior evidence it is therefore recommended that special attention should be attend to specific clusters of risk factors of non-adherence such as younger age, multiple chronic conditions, side effects and desire for more information of patients on chronic treatment.

3.8. Limitations

The study used a small sample size of 80 participants, which significantly weakens the validity and reliability of the research findings. A sample of 80 is generally considered underpowered for detecting small to moderate effects, leading to several critically detrimental statistical issues.

The convenience sampling strategy that was used is a non-probability method that relies on the easiest-to-reach subjects; it inherently lacks representativeness, meaning it cannot confidently be extrapolated to a larger population. There is also the potential for bias, as participants were selected based on availability rather than random chance.

Other disadvantages include:

Low Statistical Power (Type II Errors): The study is less likely to find a true effect that exists. The probability of incorrectly failing to reject the null hypothesis (a false negative) is high, meaning you may conclude there is no effect when one exists.

Wide Confidence Intervals (Low Precision): Smaller samples produce less precise estimates of population parameters. A small sample results in a wider margin of error, making the findings less reliable and less applicable.

Increased Risk of Sampling Error: The smaller the sample, the higher the likelihood that the sample is not representative of the broader population, leading to skewed results or spurious correlations.

High Sensitivity to Outliers: A few anomalous data points can disproportionately skew the mean and variance in a small sample, destroying the accuracy of the entire study.

Limitations in studies of patient-drug related factors for nonadherence include a reliance on self-reported adherence, which is often inaccurate, and the potential for a ceiling effect, where patients report being highly adherent, making it difficult to detect improvements from interventions.

Another important limitation is that this study employed a cross-sectional study design. The main limitation of cross-sectional studies is their inability to establish causality due to data collection at a single point in time, making it impossible to determine temporal order (cause before effect).

Other significant drawbacks include susceptibility to selection bias, difficulty studying rare disease reliance on participant memory (recall bias), and challenges in controlling for confounding variables, which limit the ability to infer generalizable, predictive conclusions.

Other limitations include a focus on single factors, ignoring the interplay of patient beliefs, drug-related issues, and healthcare system barriers, and a shortage of studies on specific populations, such as those in developing and emerging-income countries.

Additionally, some studies may not differentiate accurately between intentional and unintentional non-adherence, and technologies used to measure adherence can have their own limitations, such as patient acceptability concerns.

3.9. Strengths

Strengths of a study on patient-drug related factors for nonadherence include using validated scales for intentional and unintentional nonadherence, measuring both patient beliefs and practical difficulties, and employing statistical methods to identify significant predictors like poor knowledge or side effects.

The use of a multivariate model to analyze numerous variables simultaneously is a significant strength, as it helps control for confounding factors and produces more reliable results.

3.10. Conclusions

The aim of this study was to assess patient-drug-related factors associated with nonadherence to chronic treatment in a primary healthcare setting, in a low-middle-income setting. Patient-related and drug-related factors are complex and interconnected, significantly contributing to non-adherence in chronic treatment. Results from this study highlight the novelty in understanding how patient beliefs, perceptions, and knowledge gaps specific to this primary health care setting in a low–middle income country drive non-adherence, as there’s limited prior evidence. Therefore, specific clusters of risk factors such as younger age, multiple chronic conditions, and desire for more information create a novel predictive profile. Furthermore, there is novelty in quantifying how experiencing side effects, and not receiving adequate knowledge (p < 0.001), significantly increases nonadherence. Addressing these factors requires targeted interventions that improve patient understanding and enhance drug adherence to chronic medication.

3.11. Suggestions for Future Research

Future research should focus on integrating qualitative methods (patient experiences), exploring novel interventions (health coaching, digital reminders), investigating specific patient-drug interactions (polypharmacy, side effects), understanding socio-environmental factors (food insecurity, substance use), and using conceptual frameworks (like WHO’s) to guide studies, particularly in Low and Middle income Countries (LMICs) for tailored strategies.

The research focus areas should include the following:

Deeper Dive into Beliefs & Perceptions: Explore specific patient beliefs about medication effectiveness, side effects (real vs. perceived), and cultural influences on adherence. Investigate the impact of health literacy and how it interacts with the drug information provided.

Patient-Drug Specifics: Examine nonadherence in patients with multiple chronic conditions (polypharmacy) and complex regimens (e.g., single pill combinations vs. multiple pills). Study the role of specific drug classes (e.g., mental health meds, insulin) and their side effects on adherence.

Interventions-Focused Studies: Test the effectiveness of personalized health coaching and motivational interviewing in primary care settings. Evaluate digital tools (apps, SMS reminders) for improving adherence and health literacy. Research pharmacist-led interventions, focusing on enhanced counselling and follow-up.

Contextual & Socio-Economic Factors: Investigate how food insecurity, housing stability, and access to transport affect medication access and adherence in primary care. Study the interplay between recreational drug/alcohol use and chronic medication adherence in PHC patients.

Methodological Enhancements: Employ mixed methods approaches (qualitative interviews + quantitative surveys) to get a holistic view. Utilize conceptual frameworks (e.g., the WHO’s) to systematically map adherence determinants. Conduct longitudinal studies to track adherence over time and understand factors leading to lapses.

LMIC Context: Call for more research in LMICs, focusing on beliefs, perceptions, and intervention effectiveness, given the unique challenges.