STRENGTHS AND LIMITATIONS OF THIS STUDY

The main strength of this study is the use of primary data and inclusion of all hospitals located in North Gojjam Zone in the study.

Due to the cross-sectional nature of the data, it is not possible to establish causality or observe changes in adherence over time for some predictor variables.

Recall bias may have impacted the accuracy of adherence reporting, representing another limitation of the study.

Introduction

Millions of children were affected by the HIV/AIDS pandemic worldwide.1 With proper care, this can be reduced to below 5% in breastfeeding women and below 2% in non-breastfeeding women.2 Adherence is the extent to which an individual’s actions such as taking medication, maintaining a diet or making healthy lifestyle changes conform to the recommendations provided by a healthcare professional.3 Optimal adherence to antiretroviral (ARV) is essential for maintaining the mother’s health, minimising the risk of medication resistance and preventing HIV transmission to children.4

In 2010, the WHO updated its guidelines to prevent HIV transmission from mother to child, aiming to eliminate new paediatric HIV infections. The guidelines proposed two options. Option A involves administering nevirapine to infants throughout the breastfeeding period and starting zidovudine at 14 weeks of pregnancy, continuing the treatment throughout the pregnancy.5

Option B entails administering nevirapine to newborns for the first 4–6 weeks of life and providing pregnant women with triple ARV prophylaxis from the 14th week of pregnancy until the end of the breastfeeding period. Lifelong antiretroviral therapy (ART) is recommended for all women in WHO clinical stages 3 or 4, or with a CD4 cell count of ≤350 cells/mm³. ARV prophylaxis is expected to reduce the risk of HIV transmission to breastfeeding newborns from 35% to 5%.5

The WHO has recommended lifelong triple ART for all HIV-positive pregnant and breastfeeding women, regardless of their CD4 count or clinical stage. This approach is referred to as option B+.6 Ethiopia has implemented national strategies with the goal of increasing adherence rates to 95% or higher.7 Higher levels of adherence to ARV medication are associated with reduced viral load, a lower risk of transmission, improved immunological response and better clinical outcomes.8 In contrast, non-compliance with option B+and Prevention of Mother-to-Child Transmission (PMTCT) ARV medication during pregnancy and lactation can lead to immunosuppression, the development of viral resistance and an increased risk of HIV transmission from mother to child.9

Offering option B+ along with ART adherence to all HIV-positive pregnant and breastfeeding mothers has two main objectives: improving the mother’s health and preventing HIV transmission from mother to child.10 This strategy could contribute to achieving the goal of ending the HIV/AIDS pandemic by 2030.11

Option B+ is more advantageous for the PMTCT approach because it helps overcome barriers such as limited access to CD4 testing and achieves higher treatment coverage.6 This strategy ensures that HIV-positive women who begin treatment immediately on diagnosis experience further reductions in PMTCT rates. Simplifying treatment regimens can help mothers adhere more easily and consistently.12 Annually, 1.4 million HIV-infected women give birth worldwide, often in areas where PMTCT services are limited.13 In children, mother-to-child transmission is the primary route of HIV infection. It is estimated that, among breastfeeding infants not receiving ART therapy, 20%–23% of transmissions occur during pregnancy, 30%–35% during labour and delivery and 25%–45% during the postnatal period.14

Approximately 25% of pregnant women exhibit poor adherence to ART, with this rate increasing in the postpartum period. For women who continue to receive care, adhering to ART remains a challenge. England has faced significant difficulties affecting treatment adherence during both pregnancy and the postpartum period.15

Several studies conducted in Africa show that adherence rates are below 95%, which is considered inadequate. In Mpumalanga Province, South Africa, the adherence rate to ART among HIV-positive pregnant women was 69%16 and only 40% of women in Zimbabwe continued taking ART 1 year after giving birth.17

Obstacles to adherence to option B+drugs include a lack of awareness, insufficient partner support and disclosure, fear of stigma and discrimination, poor clinical practices, gaps in provider expertise and training, limited access to treatments, and negative attitudes of health workers.18 19

Inadequate adherence to option B+PMTCT ARV medication during lactation can lead to immunosuppressive responses, the development of viral resistance and an increased risk of HIV transmission from mother to child.9

Ethiopia adopted the WHO’s policy to increase ART adherence to 95% or more to enhance overall health and reduce viral loads. However, research conducted in the South Wollo Zone of Northeast Ethiopia found that only 87.9% of pregnant and breastfeeding mothers adhered to option B+PMTCT ART medication.20 21 Studies on adherence to option B+PMTCT medications in Ethiopia have been atypical due to the exclusion of mothers who breastfeed for up to 18 months. There has been no research specifically focused on Ethiopian breastfeeding mothers for the full 18-month period. Furthermore, no research on adherence to level option B+PMTCT medication has been done in the North Gojjam Zone. Low adherence to option B+PMTCT medication during the breastfeeding phase raises the risk of HIV transmission from mother to child. Therefore, this study aims to assess the level of adherence to option B+ with partner involvement and identify associated factors in governmental hospitals in the North Gojjam Zone of the Amhara Region.

Methods and materialsStudy design, period and area

A multicentre, institution-based cross-sectional study was carried out from 1 March 2022 to 30 April 2022, in government hospitals across the North Gojjam Zone. The North Gojjam Zone is 1 of the 15 zones in the Amhara Region, with Bahir Dar as its administrative centre, located 562 km from Addis Ababa. According to the 2023 population projection, North Gojjam Zone has a population of 1 539 874. The zone includes 4 governmental hospitals and 57 health centres, all providing 24-hour services. We selected four primary hospitals for the study: Adet Primary Hospital (APH) in Yilmana Densa Woreda, Merawi Primary Hospital (MPH) in Mecha Woreda, Duribetie Primary Hospital (DPH) in Debub Achefer Woreda and Liben Primary Hospital (LPH) in Semen Achefer Woreda.

Source population

All HIV-positive breast feeding women who had PMCT follow-up in government hospitals in the North Gojjam Zone.

Study population

All breast feeding women who had PMCT follow-up in governmental hospitals of North Gojjam Zone during the study period.

Inclusion criteria

The study included HIV-positive women who were breast feeding. This encompassed exclusive breast feeding for the first 6 months and continued breast feeding up to 18 months. Participants were required to have had at least 1 month of follow-up in government hospitals in the North Gojjam Zone for PMTCT services.

Exclusion criteria

HIV-positive women who breastfed from birth up to 18 months and were undergoing PMTCT follow-up were excluded from the study if they exclusively used formula feeding for their infants or if they were critically ill, such as being unconscious or requiring emergency treatment.

Sample size determination

The sample size was calculated based on factors identified in previous studies conducted in Tigray,22Hadiya Zone23 and public health facilities of Hawassa city,20 with the largest sample size being 228 (table 1).

Table 1

Calculated sample size by using associated factors

Study variablesIndependent variables

Sociodemographic characteristics

Place of residence.

Level of education.

Occupational status.

Marital status.

Knowledge of HIV-positive breastfeeding women on option B drug-related factor.

HIV status disclosure-related factors

HIV AIDS-related and treatment-related factors

Counselling.

WHO clinical category.

Side effects of ARV drugs.

CD4 count and WHO clinical stages.

The same day of diagnosis and initiation of ART.

Time of option B+initiation.

Male partner involvement-related factor.

Operational definition of variablesLevel of adherence

The adherence of mothers to option B+PMTCT medication was categorised as either good or poor based on their responses. Adherence was assessed using a self-reporting method. HIV-positive mothers who were breast feeding for up to 18 months and answered ‘no’ to all four adherence questions were classified as having good adherence. Conversely, those who answered ‘yes’ to at least one of the four questions were considered non-adherent.24

Knowledge

The level of knowledge about option B+PMTCT was assessed based on the total number of correct answers to the knowledge questions. Women enrolled in the option B+PMTCT programme were classified as having high, moderate or low knowledge according to their scores: ≥80% for high knowledge, 60%–79% for moderate knowledge and <60% for low knowledge, based on their responses to the eight knowledge questions.22

Male partner involvement

The level of male partner involvement in the option B+PMTCT programme was assessed using a composite measure based on the total number of correct responses to 10 questions regarding the partner’s support for his wife. Scores ranged from 0 to 10, with a score of 7 or higher classified as ‘good’ involvement, scores of 4–6 as ‘moderate’ involvement and scores of 3 or lower as ‘low’ involvement in supporting their wives in the programme.22

Disclosure

Mothers whose male partner or family members were informed of their HIV test results.7

Sampling technique and procedure

The study included all four hospitals in the North Gojjam Zone. Samples were distributed proportionately among the health institutions: MPH (52), APH (67), DPH (50) and LPH (59). The PMTCT cohort registration book served as the sampling frame. A lottery-style random sampling method was used to select research participants. Participants were interviewed at the government hospital sites in the North Gojjam Zone during their follow-up visits.

Data collection tools and procedure

Data were collected through face-to-face interviews using a standardised, structured questionnaire adapted from a previously validated instrument.22 24 To ensure consistency in the questions, the questionnaire was first written in English, then translated into Amharic, and subsequently, translated back into English by language specialists. The tool is divided into six sections: the first section covers the sociodemographic characteristics of the respondents; the second section addresses HIV status disclosure; the third section contains questions related to HIV disease and treatment; the fourth section focuses on male partner involvement; the fifth section assesses knowledge of option B+PMTCT and the final section evaluates adherence levels.

To assess patients’ medical conditions, including the date of HIV diagnosis, initiation of therapy, CD4 count at admission and WHO clinical stage, respondents were interviewed, and patient records were reviewed. Adherence to option B+PMTCT was evaluated using a self-reporting method, with measurement tools adapted from South African practices designed for resource-limited settings. HIV-positive nursing mothers were considered to have good adherence if they answered ‘no’ to each of the four adherence questions. A response of ‘yes’ to any of the questions indicated non-adherence. The adherence questions included:

Do you sometimes find it difficult to remember to take your medication?

When you feel better, do you sometimes stop taking your medication?

Many patients have trouble taking their ARV doses as prescribed; did you miss any ARV doses in the last 3 days?

If you feel worse when you take the medicine, do you sometimes stop taking it?

Data collection was carried out by nine midwives with diplomas, while five midwives with bachelor’s degrees supervised the distribution of the structured questionnaire.

Detailed tools are provided in online supplemental file 1.

Data quality control

Data collectors and supervisors received 1 day of training from the principal investigator. This training covered the study’s purpose, data collection methods, confidentiality, respondents’ rights, informed consent, study objectives, interview techniques and how to complete the questionnaire. Supervision was provided to ensure proper data collection practices. Before the main data collection began, a pretest was conducted on 5% of the sample size (14 individuals) at Dangila Primary Hospital. Validated tools were employed, and their reliability was assessed using Cronbach’s alpha coefficient, which was found to be 0.82 for this study. The collected data were reviewed for completeness, clarity and consistency, with immediate corrective actions taken to address any issues identified by both data collectors and supervisors.

Data analysis and presentation

After coding, cleaning and importing the data into EPI-Data V.3.1, it was exported to SPSS V.26 for further analysis. Descriptive statistics, including frequency tables and graphs, were used to summarise the data.

Bivariate analysis was conducted to examine the relationship between each explanatory variable and the dependent variable. Multivariable logistic regression analysis was employed to account for confounding variables that were significantly correlated with the outcome variable in the bivariate regression.

In bivariate analyses, the Wald test and the likelihood ratio test were used to evaluate assumptions and model fit. In multivariate analyses, assumptions and model fit were assessed using the Wald test, likelihood ratio test, score test and Hosmer-Lemeshow test. Additionally, diagnostics such as variance inflation factor (VIF) and tolerance were employed to evaluate the effects of multiple predictors and address multicollinearity, ensuring the stability and reliability of the model.

Variables with a p value less than 20% in the bivariate analysis were included in the multivariate logistic regression analysis to determine their correlation with the dependent variable. The adjusted OR (aOR) and a p value of less than 0.05, along with a 95% CI, were used to indicate statistically significant relationships between the independent predictors and the outcome variable. The VIF was employed to assess multicollinearity among the independent variables.

The final model’s fit was assessed using the Hosmer-Lemeshow goodness-of-fit test, which yielded a result of 0.5. Since 0.5 is greater than 0.05, the p value indicates no significant discrepancy between the observed and expected event rates, suggesting that the model fits the data reasonably well.

Patient and public involvement

Patients and the public were not engaged in the planning and design of the study.

ResultsSociodemographic characteristics of the respondents

The study included 223 HIV-positive breastfeeding mothers, with a response rate of 97.8%. Among the participants, 136 (61%) were over 29.56 years old, with an SD of ±4.64 years. Of the total respondents, 152 (68.2%) were married, and 176 (78.9%) had received an education (table 2).

Table 2

Sociodemographic characteristics of the respondents adhering to option B+PMTCT in Governmental Hospitals of North Gojjam, Northwestern Ethiopia, 1 March 2022–30 April 2022 (N=223)

Disclosure of HIV status among HIV-positive women

In this study, approximately 197 participants (88.3%) disclosed their HIV status. Of those, 79.2% informed their spouse, 7.1% told friends and 13.2% shared the information with relatives.

HIV AIDS and treatment-related characteristics

Most of the interviewed participants were enrolled in the option B+programme at stage one. At the time of admission, 187 respondents (83.9%) had CD4 counts below 250 cells/mm³. Although most respondents (217 or 93.3%) received counselling about the side effects of ART medication, the average counselling session lasted less than 30 min. The primary challenge of option B+PMTCT drug adherence was fear of stigma and discrimination, reported by 53.9% of participants.

Male partner involvement in ward option B+PMTCT drugs

A significant proportion of respondents (199 or 89.2%) reported that their partner attended PMTCT clinic with them. However, 152 participants (68.2%) did not accompany their spouse to the PMTCT clinic to collect their ARV medication. Despite this, the majority of those interviewed indicated having strong support from their partner (table 3).

Table 3

Male partner involvement in option B+PMTCT drug adherence at governmental hospitals in North Gojjam Zone, Northwest Ethiopia, 1 March 2022–30 April 2022 (N=223)

Knowledge regarding option B+PMTCT drug among HIV-positive women

A total of 92.5% of respondents were aware of option B+PMTCT medication. Among them, 4 (1.8%) had low knowledge, 49 (22%) had moderate knowledge and 170 (76.2%) had high knowledge about option B+medication.

Level of option B+ PMTCT drug adherence

Of the total respondents, 62.8% adhered to option B+PMTCT drugs (figure 1).

Figure 1

Level of option B+PMTCT drug adherence. PMTCT, Prevention of Mother-to-Child Transmission.

Factors associated with adherence to option B+PMTCT drugs

The bivariate logistic regression analysis indicated significant relationships between the respondent’s level of knowledge, educational status, the kind of health facility, the parents’ HIV status, the involvement of their male partner in there, the time it took to get from home to the health facilities and the option B+PMTCT drug adherence. However, the multivariable logistic regression analysis identified male partner support, travel time from home to the facility and the type of health facility as significantly associated with the level of adherence to option B+PMTCT drugs.

Compared with those who travelled 60 min or longer for option B+services, participants who travelled less than 60 min were 6.34 times more likely to adhere to option B+PMTCT medication (AOR=6.34, 95% CI: 2.92 to 13.75).

Individuals with high levels of partner support were 3.03 times more likely to adhere to option B+PMTCT medication compared with those with low levels of partner support (AOR=3.03, 95% CI: 1.12 to 9.13).

Participants in this study who visited a health centre were 83% less likely to adhere to option B+PMTCT medications compared with those who visited a hospital (AOR=0.17, 95% CI: 0.008 to 0.35) (table 4).

Table 4

Bivariate and multivariate logistic regression analysis of factors associated with adherence to option B+PMTCT drugs among breast feeding women in governmental hospitals of North Gojjam Zone, Northwest Ethiopia, 1 March 2022–30 April 2022 (N=223)

Discussion

According to our findings, 62.8% of individuals adhered to the option B+PMTCT medication, with a 95% CI ranging from 55.4% to 68.6%.

This finding is consistent with research from studies conducted in America,25Zambia%26 and Kenya,27 which reported adherence rates of 65%, 70.5% and 61.9% to option B+medication, respectively. This could be because the study population, consisting of breastfeeding mothers, was similar across the studies.

Comparing study populations allows for a more direct assessment of adherence behaviours and outcomes. Nursing mothers often encounter similar experiences, challenges and motivations regarding medication adherence in PMTCT programmes, which can lead to consistent adherence rates across different studies.

Moreover, similar study populations help minimise the impact of confounding variables that might affect drug adherence rates. By concentrating on a specific group, such as nursing mothers, researchers can gain deeper insights into the factors influencing adherence within that demographic and make more accurate conclusions about adherence to option B+medication. Thus, using a comparable study population enhances the reliability and comparability of findings across different regions or countries, as shared characteristics within the population can lead to more uniform adherence rates.

However, the adherence level in this study was higher than that reported in study conducted in Uganda (51%).28 This discrepancy could be due to differences in the study populations and designs. This study focused on breastfeeding women up to 18 months and used a cross-sectional design, whereas the studies in Uganda employed a cohort design and included both pregnant and breastfeeding women.

The variance in study design, with our study being cross-sectional and the Uganda study being cohort, can also impact drug adherence rates. Cohort studies track participants over a period, allowing for a more in-depth analysis of changes in adherence behaviour over time. In contrast, cross-sectional studies provide a snapshot of adherence at a single point in time, which may not capture changes or trends in adherence patterns.

Therefore, differences in study population and design can influence drug adherence rates by affecting the motivations, challenges and behaviours of the participants, as well as the depth of understanding provided by the research methodology.

Furthermore, the findings of this study were lower than those of research conducted in East Shewa (82.6%),29 Hadiya Zone (83.7%),23 Southern Ethiopia (81.4%)21, South Wollo Zone (87.9%)21 and Tigray region in Mekelle (87.1%).22 This difference may be due to the study’s focus on breastfeeding women up to 18 months, whereas studies in Southern Ethiopia, Tigray, East Shewa and the Hadiya Zone included both pregnant and breastfeeding women for 6 months. Pregnant women might be more motivated to adhere to option B+PMTCT medication to ensure the birth of HIV-free infants compared with breastfeeding mothers who have already delivered. This study’s results were also lower than those of research carried out in Lagos, Nigeria (80.6%)30 and Uganda in Mugalo Hospital (86%).31 The adherence assessment tool could be the cause of this disparity.

The composition of the study population, such as focusing solely on nursing mothers caring for infants until their HIV status is known, can influence drug adherence rates. Pregnant women, as seen in other studies, may have different motivations and priorities regarding medication adherence compared with nursing mothers. The stage of motherhood, the presence of newborns and other factors specific to nursing mothers could impact their ability to follow medication instructions.

The tool used to assess adherence plays a crucial role in determining adherence rates. Different tools may have varying levels of sensitivity and specificity in capturing adherence behaviours. The adherence assessment tool used in our study differs from those used in studies reporting higher adherence rates, it could lead to discrepancies in the reported prevalence of drug adherence. Factors such as the method of data collection, the frequency of assessments and the criteria for defining adherence can all influence the results obtained.

The study’s findings indicated that individuals with high levels of partner support were three times more likely to adhere to option B+PMTCT medication compared with those with low levels of support. This is consistent with research from Tanzania,32 Lilongwe in Malawi33 and Mugalo Hospital in Uganda,31 which also supports these findings. This could be because partners can help women adhere more closely to option B+PMTCT medication by providing financial support, encouraging medication adherence, promoting prenatal care, increasing utilisation of PMTCT services and improving overall PMTCT interventions.7

Compared with those who travelled more or equal to 60 min for option B+service, participants who travelled less than 60 min for option B service were 6.34 times more likely to adhere to option B+PMTCT medication. The results of this investigation corroborated cross-sectional research performed in southern Ethiopia,29 Nepal34 and WHO.10 This may be due to gendered power dynamics that affect their ability to access necessary care.35 This may also be since participants who travel shorter distances encounter fewer barriers, such as transportation costs, time constraints or physical exertion, making it easier for them to access and adhere to their medication regimen. This convenience can significantly impact their ability to consistently follow their treatment plan. In contrast, women who must travel long distances to reach a health facility may experience fatigue and miss appointments due to social issues. Additionally, fear of stigma and discrimination (53.9%) was identified as a major obstacle to adherence to option B+PMTCT in this study.

Participants in this study who visited a health centre were 83% less likely to adhere to option B+PMTCT medications compared with those who visited a hospital. This finding is supported by a similar investigation conducted in the South Wollo Zone.21 This could be differences in the quality of care provided, accessibility to medications, level of education or knowledge about option B+drug and distance to the health facility. Additionally, in our study, 1.8% of participants had low knowledge and 21.1% were uneducated, which may also affect adherence to option B+medication

Due to logistical and time constraints, this study adopted a quantitative approach, limiting its capacity to thoroughly explore ‘why’ questions. Adherence categorisation may be prone to recall bias, even with efforts to reduce it through methods such as conducting interviews in private settings, employing prompting techniques, using memory aids, requesting concrete examples and encouraging participants to share specific instances related to the topic. Additionally, social desirability bias could impact the findings; however, measures were taken to address this issue. Participants were informed about the study’s purpose, confidentiality protocols and the significance of honest participation, which may have influenced the outcomes.

Conclusion and recommendation

This study demonstrates that, in comparison to the recommended adherence level, the level of adherence to option B+PMTCT medication was low. Option B+PMTCT drug adherence was strongly associated with partner support, the types of health facility and the time taken to reach from home to the health facility.

Health interventions aimed at improving adherence to PMTCT medication should consider strategies to involve and educate partners. Encouraging partner support can potentially lead to better health outcomes for both the mother and child by ensuring consistent adherence to the prescribed treatment protocol. Enhancing partner support involvement for individuals receiving option B+PMTCT medication.

Health systems and policy-makers should prioritise efforts to improve geographical access to PMTCT services. This can include decentralising services, establishing more accessible clinics, improving transportation infrastructure and providing support for transportation costs. By reducing travel barriers, more women can access and adhere to PMTCT medication, thereby improving maternal and child health outcomes in HIV prevention and treatment programmes.

Efforts should focus on ensuring that PMTCT services are available and accessible at higher-level healthcare facilities, such as hospitals. This might involve:

Service integration: Integrating PMTCT services into hospital settings where comprehensive care and support systems are often better established.

Training and resources: Providing adequate training and resources to healthcare providers in health centres to improve their capacity to deliver effective PMTCT services.

Referral systems: Implementing effective referral systems between health centres and hospitals to ensure continuity of care and support for PMTCT participants.

Community engagement: Engaging communities to raise awareness about the importance of accessing PMTCT services at hospitals for better adherence and health outcomes.

Additional support, resources or tailored strategies may be needed to enhance medication adherence for participants who PMTCT follow-up in the health centre.

Subsequent research should be conducted using a longitudinal approach in conjunction with a qualitative study design.

Data availability statement

All data relevant to the study are included in the article or uploaded as online supplemental information.

Ethics statementsPatient consent for publication

Consent obtained directly from patient(s).

Ethics approval

This study involves human participants, and ethical approval was obtained from Debre Tabor University’s Research and Ethical Review Committee; the reference number for the approval letter was DTU 0841/2021. Additionally, a letter of support was obtained from the North Gojjam Zone Health Office. Participants provided informed consent before participating in the study.

Acknowledgments

We would like to thank Debre Tabor University for its ethical letter and financial support. Our gratitude is also great to all data collectors and study participants. We are also glad to North Gojjam public health institutions for writing a permission letter.