Abstract
Background
Rational medicine use during breastfeeding remains a challenge, especially in low- and middle-income countries. Self-medication, incomplete dosing, inappropriate drug selection practices, poor adherence, and adverse drug reactions contribute to antimicrobial resistance. Although the Information-Motivation-Behavioral Skills (IMB) model has successfully influenced health behaviors, its application to promoting safe medicine practices among breastfeeding women in urban settings is limited. This study assessed the association of an IMB-based educational intervention with knowledge, attitudes, and practices related to medication use.
Methods
A quasi-experimental study with intervention and control groups was conducted among 184 breastfeeding women (≥18 years) with infants ≤12 months attending six urban clinics in Kampala, Uganda. Participants were recruited using systematic random sampling and allocated to intervention and control groups. The intervention comprised 12 weekly IMB-based health education sessions delivered through lectures, group discussions, videos, and printed materials. The control group received standard health education. Outcomes were measured at baseline and three months using structured Likert-scale questionnaires. Changes over time and between groups were analyzed using repeated-measures ANOVA and mixed-effects regression models.
Results
Baseline characteristics were comparable between groups. At three months, greater improvements in knowledge were observed in the intervention group compared with the control group (89% vs. 80%; MD =8.3; 95% CI: 6.0–10.6), motivation (81% vs. 70%; MD =11.5; 95% CI: 8.6–14.4), and behavior (91% vs. 81%; MD = 9.9; 95% CI: 7.3–12.5), all p < 0.001. Mixed-effects regression analyses showed that changes over time differed significantly between the intervention and control groups, with greater improvements in knowledge observed in the intervention group compared with the control group (β = 7.34; p < 0.001), motivation (β = 11.49; p < 0.001), and behavior (β = 5.67; p = 0.012).
Conclusion
The IMB-based intervention improved women’s knowledge, motivation, and practices, supporting the integration of such programs into routine health education.
Keywords: Breastfeeding, breastmilk, lactation, infants, medication, model, intervention
Introduction
Breastfeeding is a fundamental public health strategy that substantially enhances child survival, nutrition, and neurodevelopment, while also offering long term health benefits for mothers 1. It serves as a critical pillar of primary health care, particularly in low- and middle-income countries (LMICs), where suboptimal infant feeding practices contribute significantly to preventable morbidity and mortality among children under five years of age 2. Exclusive breastfeeding during the first six months of life, followed by continued breastfeeding alongside age appropriate complementary feeding, can markedly reduce the risk of infectious diseases, malnutrition, and developmental delays 3. In recognition of these benefits, the World Health Organization strongly advocates for breastfeeding as a universal practice. However, despite widespread policy support and public health campaigns, numerous structural and individual level barriers continue to undermine breastfeeding practices. One increasingly recognized yet under addressed challenge to sustained breastfeeding is the uncertainty surrounding medicine use during lactation. Many women require pharmacological treatment for acute or chronic conditions, yet often lack access to reliable, evidence-based information on drug safety and appropriate use while breastfeeding. This information gap can result in maternal anxiety, early weaning, nonadherence to prescribed therapies, and/or inappropriate self-medication 4.
In Uganda, self-prescription and incomplete dosing of medications among breastfeeding women are common. Recent findings from a study in Kampala indicated that nearly 90% of breastfeeding mothers acquired medications through informal channels, with 80% failing to complete prescribed courses of medication, leading to unsafe storage, reuse, and improper disposal 5. Such practices raise concerns not only about drug resistance and toxicity in the mother but also unintended exposure of infants to potentially harmful substances via breast milk. Despite these risks, existing public health programs in Uganda and other low-resource settings have rarely integrated structured interventions to promote informed and safe medicine use during breastfeeding.
Numerous studies have emphasized the lack of consistent, evidence-based communication between healthcare providers and breastfeeding women regarding medication safety 6, 7. While drug information databases and prescribing tools are available, providers often underutilize them, and access to such resources remains limited in many community health settings. Also, research has shown that healthcare workers themselves sometimes lack adequate training or confidence to counsel breastfeeding women on medicine use, leading to inconsistent or overly cautious advice, such as blanket recommendations to discontinue breastfeeding 8.
Behavioral models can offer valuable frameworks to address these challenges through targeted educational interventions. The Information-Motivation-Behavioral Skills (IMB) model, originally developed to guide health promotion strategies in HIV prevention, has been increasingly applied in maternal health and medication adherence 9, 10. The IMB model posits that behavior change is driven by the interplay of three core components: accurate information, personal and social motivation, and the behavioral skills necessary to act on that information. When applied to breastfeeding and medicine use, the model suggests that women must first be well informed about medicine safety, be motivated to make healthy choices for themselves and their infants, and possess the practical skills to apply that knowledge effectively.
Evidence from other domains supports the value of IMB-based interventions in influencing health behaviors. For example, IMB-guided programs have been successful in improving antiretroviral adherence among postpartum women 10, enhancing self-care practices in diabetes management 11, and supporting contraceptive use among adolescent mothers 12. However, there is a paucity of published studies applying the IMB model to the specific issue of medicine use during breastfeeding, particularly in African settings. The unique sociocultural, economic, and health system factors influencing mothers' decision-making around medication use in these settings necessitate appropriate interventions that are both theory-driven and community-informed.
Urban health settings face a mix of challenges, including high patient volumes, limited provider time for counseling, and widespread misinformation circulating through peer networks and informal drug shops 13, 14. Against this setting, educational interventions that are participatory, tailored to local realities, and grounded in behavior change theory offer a promising path for promoting safer medicine use practices among breastfeeding women. The quasi-experimental study was designed to evaluate the effectiveness of a health educational intervention guided by the IMB model in improving the knowledge, attitudes, and practices (KAP) related to medicine use among breastfeeding women in Kampala. By assessing the outcomes of this intervention, the study seeks to generate evidence that can inform future health education programs, guide clinical counseling practices, and contribute to policy recommendations aimed at promoting rational medicine use during breastfeeding.
Materials and methods
Study design and setting
The study employed a two-group quasi-experimental design to examine changes in breastfeeding women's knowledge, attitudes, and practices related to medicine use following an Information–Motivation–Behavioural Skills (IMB) model-based health education intervention. The study was considered quasi-experimental because health facilities were allocated to the intervention and control groups using a non-random approach based on geographical distance to minimise contamination between study arms.
The study was conducted in six healthcare clinics in Kampala, Uganda, including the Infectious Diseases Institute (IDI) and Kampala Capital City Authority (KCCA) clinics. These facilities provide a broad range of maternal and child health services, including immunisation, outpatient and inpatient care, growth monitoring, family planning, and specialised services for people living with HIV. The IDI, a research institution affiliated with Makerere University College of Health Sciences, supports over 200,000 people living with HIV through its Health Systems Strengthening Programme, while the KCCA clinics serve a large urban population and provide comprehensive maternal and child health services and routine health education.
Six study sites were purposively selected, with three allocated to the intervention arm and three to the control arm. Allocation was based primarily on the physical distance between clinics to minimise contamination. Clinics located within a 5 km radius of the Infectious Diseases Institute (IDI) were grouped in one study arm, while those located more than 7 km away were assigned to the opposite arm. Accordingly, the IDI, Kisenyi, and Komamboga clinics were assigned to the control arm, while Kiswa, Kitebi, and Kawaala clinics were assigned to the intervention arm.
Eligibility and study selection of participants
The study included breastfeeding women aged 18 years and older with infants aged 12 months or younger at baseline. Participants were recruited between September 2024 and January 2025. Within each health facility, systematic random sampling was used to select eligible participants. Based on average clinic attendance, every 5th eligible woman was approached for participation. The first participant was selected using a simple random lottery method from the first five eligible women at each clinic.
The sample size was determined based on previous studies, where the proportion of breastfeeding women with the outcome after intervention ranged from 50% in the control group to 72% in the intervention group. A total sample size of 184 participants (92 per arm) was calculated to provide 80% power at a two-sided α = 0.05 to detect a standardized mean difference (Cohen’s d) of approximately 0.42 between groups, assuming a 20% loss to follow-up. The calculation assumed individual-level independence and did not account for clustering at the clinic level.
Data collection and study procedures
The survey questionnaire was developed by adapting and modifying items from previously published surveys 15, 16 to align with the study objectives. The questionnaire was self-administered at baseline and at three months to assess differences over time associated with the intervention and was designed to be completed within 20 minutes. The questionnaire was professionally translated into Luganda, the most commonly spoken language in the selected study sites. To ensure validity, the tool was reviewed by peer mothers to assess content relevance and clarity, and it was pretested among breastfeeding women to ensure face validity and appropriateness of the questions. Data from the pilot study were used to refine the questionnaire and were not included in the final analysis. Importantly, the participants involved in the pilot testing were different from those recruited for the main study; none of the pilot participants were included in the final study sample. The questionnaire gathered information on demographic characteristics (e.g., maternal age, education level, parity, and infants’ age).
The primary outcome variables were based on the IMB Skills model and included information (knowledge of rational medicine use), motivation (attitudes and beliefs, including myths), and behavioral skills (reported medicine use practices among women). The outcomes were measured using a set of items or questions using a Likert-type scale. The outcomes were derived by summing up the scores divided by the expected maximum score (50) to form a percentage score per individual. The percent score was used as a dependent variable in our analysis. Negatively worded questions were reverse-coded before analysis.
Information construct measured the knowledge of rational medicine use during breastfeeding was assessed using statements rated on a five-point Likert scale (0–5), with total scores summed for each participant; higher scores indicated safer, more informed practices. The items evaluated participants’ understanding of medicine safety, sources of reliable information, and potential effects of medications, alcohol, and smoking on the breastfed infant. Statements included awareness of medicines that may affect the baby, knowledge of where to find information, attitudes toward taking medicines without healthcare provider advice, the need to inform prescribers about medicine use, use of multiple drugs per consultation, discussing medicine use with family or friends, and use of herbal medicines during breastfeeding, as indicated in Table 6.
Table 6. Summarizing the items under the IMB model constructs.
| IMB Construct | Domain | Example items summarized | Response scale |
|---|---|---|---|
| Information (Knowledge) | Knowledge about the use of medicine during breastfeeding | Knowledge of where to find reliable information; having enough information to use medicines safely; attitudes toward self-medication and herbal remedies; discussing medicines with family/friends; preference for multiple prescriptions per consultation; beliefs about alcohol/smoking effects | 5-point Likert (Strongly Disagree – Strongly Agree) |
| Motivation (Attitudes & Beliefs) | Attitudes and beliefs regarding safe medicine use | Perceived safety or harm of medicines; beliefs about stopping or disrupting breastfeeding; consulting healthcare providers and family/friends; use of natural remedies; protective use of medicines for the infant. The belief that Vaccines are unsafe during breastfeeding | 5-point Likert (Strongly Disagree – Strongly Agree) |
| Behavior (Practices) | Medicine use practices during breastfeeding | Following dosage instructions, consulting healthcare providers before using medicines (OTC or prescribed); storing medicines safely; keeping medication records; avoiding leftover medicines; avoiding sharing medicines with the infant; adjusting breastfeeding schedule; seeking additional information | 5-point frequency scale (Always – Never) |
Motivation construct assessed the attitudes and beliefs regarding medicine use during breastfeeding, with responses captured on a five-point Likert scale (0–5). Higher scores reflected more favourable attitudes and beliefs. The items addressed perceptions of medicine safety and harm, the need for professional consultation, reliance on natural remedies, the influence of family or friends on decision-making, and attitudes toward modifying or stopping breastfeeding while taking medicines. Sample statements included beliefs that all medicines can be used safely or are harmful during breastfeeding, whether breastmilk is unsafe while on medication, the ability to stop or disrupt breastfeeding to complete treatment, the use of natural remedies without medical advice, and the perceived necessity of consulting family or healthcare providers before taking medicines.
The behavior skills construct assessed women’s practices regarding medicine use, with responses recorded on a five-point Likert scale (0–5; always to never). Behavioral skills were used when referring to the IMB theoretical construct, whereas behavioral practices were used to describe the measured study outcome. Higher scores indicated safer and more appropriate practices. The items addressed behaviors related to medicine adherence, consultation, information-seeking, storage, and breastfeeding management. Sample statements included using herbal or over-the-counter medicines without consulting a healthcare provider, sharing drugs with the infant, using leftover medicines, following prescribed dosage instructions, adjusting breastfeeding schedules based on medication use, discussing breastfeeding status with healthcare providers, seeking additional information from reliable sources, storing medicines safely, and keeping a record of medications taken while breastfeeding.
The health education modules were developed from the following sources: Uganda Policy guidelines on Infant and Young Child Feeding 17, 18, Lactation Management Self-Study Modules 19, the National Medicines Policy 20, breastfeeding and maternal medication, medication safety and 21, 22 drug safety in lactation 23, 24. Videos and flyers were designed to supplement the health education sessions. After each session, participants were given a time frame of 10 minutes to ask questions. The intervention group received an enhanced, IMB-guided educational program, which included additional components not provided to the control group, such as flyers and videos on medication safety tips, structured group discussions, experience sharing, and practical exercises. These additional elements were designed to reinforce knowledge, motivate behavioral change, and build skills for safe medicine use, ensuring that any observed differences in outcomes could be attributed to the enhanced intervention. Before implementation, the study team underwent training on the content and use of these materials, ensuring consistent delivery of the intervention.
Intervention group
Participants in the intervention group received a structured, IMB-based health education programme designed specifically to improve knowledge, attitudes, and practices regarding medicine use during breastfeeding. The intervention was delivered through standardized and structured health education sessions, group discussions, and experience sharing. A written script was used to ensure consistency of content, covering topics such as the definition of medicines, rationale for prescribed medicine use, safety considerations, side effects, and safe medicine storage. The motivation component included facilitated and structured group discussions, brainstorming, and sharing of personal experiences related to medicine use. The behavioural component included video-based demonstrations on medicine safety and reflections on lessons learned. The intervention was delivered over 12 weeks, with monthly sessions lasting approximately 60 minutes.
Control group
Participants in the control group received routine health education delivered within the study setting. This consisted of standard health education messages to pregnant and breastfeeding women during clinic visits. The content included general maternal and child health information, such as the benefits of breastfeeding, breastfeeding practices, addressing common breastfeeding myths, and guidance on when to seek health services. The control group did not receive the structured IMB-based medicine safety modules, written scripts, or video-based materials developed for the intervention group. The study duration was 12 weeks, with monthly sessions similar to those of the intervention group.
Statistical analysis
Descriptive statistics were used to summarize the data, with frequencies and percentages for categorical variables and mean (± standard deviation, SD) or median (interquartile range, IQR) for continuous variables. Differences between the intervention and control arms at baseline were assessed using the chi-square test for categorical variables and the Kruskal–Wallis test for continuous variables.
To examine the association between the intervention and study outcomes, a quasi-experimental pre-test–post-test control group design with a matched baseline comparison group was employed.
Observations were measured repeatedly over time on the same participants. Analysis of variance (ANOVA) and mixed-effects regression modelling were used to estimate differences in outcomes over time between groups. Regression analyses were performed to assess changes in outcomes over time between groups. The models included main terms for time and group, as well as a time-by-group interaction term to assess whether changes in knowledge, attitudes, and behaviors differed between the intervention and control groups. Baseline behavioral scores were included as covariates in the regression and repeated-measures models to adjust for initial differences between groups.
As an estimate of the magnitude and direction of the intervention-related differences, effect sizes (ES) using Cohen's d were calculated, using the formula ((M post, E – M pre, E)/SD pre, E – (M post, C – M pre, C)/SD pre, C) for an independent groups’ pre-test–post-test design. Effect sizes for main and interaction effects are reported as partial η2 (Eta-Squared), adjusted for covariates. Data were analyzed using Stata version 18 (Stata Corp, College Station, TX, 2023).
Results
The study included 184 participants, with 92 mothers in each of the intervention and control groups. Baseline maternal characteristics, including age, number of children, education, and occupation, were generally comparable between groups. Infant characteristics, including median age, were also similar across groups. Statistical comparisons did not show evidence of meaningful differences between groups at baseline. Detailed distributions of these characteristics are presented in Table 1. At two months, one participant in the intervention group and one participant in the control group were lost to follow-up. By the three-month follow-up, no additional loss to follow-up occurred in the intervention group, whereas four additional participants were lost in the control group. Overall, one participant in the intervention group (1.1%) and five participants in the control group (5.4%) were lost to follow-up, corresponding to an overall attrition rate of 3.3% (6/184), as shown in Figure 1, which indicates the study flow chart of participants.
Table 1. Baseline Characteristics of the study participants by intervention/control status.
| Characteristic | Control (n=92) |
Intervention (n=92) |
p-value |
|---|---|---|---|
| Maternal age, median (IQR) | 28 (24.0 – 33.0) | 27.0 (23.0 – 31.5) | 0.554 |
| Number of children, median (IQR) | 2 (1.0 – 3.0) | 2.0 (1.0 – 3.0) | 0.682 |
| Highest level of education | 0.288 | ||
| College/University | 8 (8.7) | 13 (14.1) | |
| Secondary | 51 (55.4) | 41 (44.6) | |
| Primary | 33 (35.9) | 38 (41.3) | |
| Occupation | 0.764 | ||
| Civil servant | 2 (2.2) | 3 (3.3) | |
| Business person | 40 (43.5) | 35 (38.0) | |
| Peasant | 7 (7.6) | 5 (5.4) | |
| Housewife | 43 (46.7) | 49 (53.3) | |
| Infant characteristics |
Figure 1. Study flow chart of participants.
Knowledge, Attitude, and Behavioral outcomes by intervention/control groups at baseline and completion
At baseline, the intervention and control groups had comparable knowledge and attitude scores regarding medicine use during breastfeeding, with mean knowledge scores of 71.1 ± 11.3 versus 70.4 ± 11.4 (mean difference 0.7; 95% CI: -2.6 to 4.0; p = 0.668) and mean attitude scores of 61.9 ± 8.6 versus 62.2 ± 8.9 (mean difference -0.3; 95% CI: -2.9 to 2.2; p = 0.814), indicating similar baseline understanding in both groups. However, behavior scores were significantly higher in the intervention group (68.6 ± 15.0) compared to the control group (64.3 ± 10.5), with a mean difference of 4.3 points (95% CI: 0.6 to 8.1; p = 0.024). At the end of the three-month follow-up period, the intervention group demonstrated significantly higher knowledge, attitude, and behavior scores than the control group. The mean knowledge score was 88.7 ± 6.4 in the intervention group versus 80.4 ± 8.9 in the control group, with a mean difference of 8.3 (95% CI: 6.0 to 10.6; p < 0.001). Similarly, the mean attitude score was 81.2 ± 11.1 in the intervention group compared to 69.7 ± 8.1 in the control group, with a mean difference of 11.5 (95% CI: 8.6 to 14.4; p < 0.001) Table 2 and Table 3, which display the Baseline and endline knowledge, Attitude, and Behavioral outcomes. As shown in Figure 2, both groups started with similar baseline scores for knowledge (70–72%) and motivation (62%) regarding the use of medicine during breastfeeding. By the end line, the intervention group showed greater improvements than the control group, with knowledge increasing to 89% versus 80% and motivation rising to 81% versus 70%.
Table 2. Baseline Knowledge, Attitude, and Behavioural outcomes by intervention/control groups.
| Outcomes | Baseline | Mean difference (95% CI) |
p-value | |
|---|---|---|---|---|
| Intervention (n=92) |
Control (n=92) |
|||
| Mean Knowledge score (SD) | 71 (11.3) | 70 (11.4) | 0.7 (-2.6, 4.0) | 0.668 |
| Mean Attitude score (SD) | 62 (8.6) | 62 (8.9) | -0.3 (-2.9, 2.2) | 0.814 |
Table 3. Post-intervention Knowledge, Attitude, and Behavioural Outcomes by intervention/control groups.
| Outcomes | Post intervention | Mean difference (95% CI) |
p-value | |
|---|---|---|---|---|
| Intervention (n=91) |
Control (n=86) |
|||
| Mean Knowledge score (SD) | 89 (6.4) | 80 (8.9) | 8 (6.0, 10.6) | <0.001 |
| Mean Attitude score (SD) | 81 (11.1) | 70 (8.1) | 12 (8.6, 14.4) | <0.001 |
Figure 2. Information, motivation and behavior construct to assess medicine use during breastfeeding.
Main outcomes associated with the intervention
Regression analysis indicated significant differences over time between groups in knowledge, attitudes, and behavior associated with the intervention. For knowledge, scores increased significantly over time among all participants (β = 10.00; 95% CI: 7.16–12.84; p < 0.001), with no baseline difference between groups (β = 1.04; 95% CI: -1.76–3.84; p = 0.467), Table 4. The time by group interaction, accounting for baseline scores, indicated that the intervention group experienced greater gains than the control group (β = 7.34; 95% CI: 3.37–11.33; p < 0.001). Specific knowledge items that contributed most to this improvement included participants’ awareness that some medicines may affect their baby while breastfeeding, knowing where to find information about medicines and breastfeeding, and a preference for having three or more drugs prescribed per consultation.
Table 4. Regression model estimating the association between the intervention and study outcomes.
| Characteristics | Coefficient | 95% CI | P-value |
|---|---|---|---|
| Knowledge Score | |||
| Time | |||
| Baseline | Ref | ||
| Post intervention | 10.00 | 7.16 – 12.84 | <0.001 |
| Group | |||
| Control | Ref | ||
| Intervention | 1.04 | -1.76 – 3.84 | 0.467 |
| Time * group | 7.34 | 3.37 – 11.33 | <0.001 |
| Attitude Score | |||
| Time | |||
| Baseline | Ref | ||
| Post intervention | 7.49 | 4.81 – 10.17 | <0.001 |
| Group | |||
| Control | Ref | ||
| Intervention | 0.03 | -2.61 – 2.67 | 0.985 |
| Time * group | 11.49 | 7.73 – 15.24 | <0.001 |
| Behavioral score | |||
| Time | |||
| Baseline | Ref | ||
| Post intervention | 16.89 | 13.73 – 20.05 | <0.001 |
| Group | |||
| Control | Ref | ||
| Intervention | 4.44 | 1.29 – 7.59 | 0.006 |
| Time * group | 5.67 | 1.23 – 10.09 | 0.012 |
Note: Models were adjusted for baseline characteristics (maternal age, education, and occupation); however, none were significant.
For attitudes (motivation), scores also improved over time (β = 7.49; 95% CI: 4.81–10.17; p < 0.001), with no baseline difference between groups (β = 0.03; 95% CI: -2.61–2.67; p = 0.985). A significant interaction between time and group was observed, with a greater increase in attitudes in the intervention group compared with the control group (β = 11.49; 95% CI: 7.73–15.24; p < 0.001). Key motivation items driving this improvement included consulting spouses, parents, friends, or relatives before taking any medicines while breastfeeding, and challenging the belief that all medicines can be used without concern during breastfeeding.
For behavior, scores increased significantly over time (β = 16.89; 95% CI: 13.73–20.05; p < 0.001), with the intervention group having a higher baseline score (β = 4.44; 95% CI: 1.29–7.59; p = 0.006). The significant time by group interaction (β = 5.67; 95% CI: 1.23–10.09; p = 0.012) indicated that the intervention group demonstrated greater improvement in safe medicine-use practices compared to controls. Notable changes in behavior included safer medicine storage, keeping records of medicines taken, reducing the use of leftover medicines, and avoiding sharing drugs with the infant. Although baseline behavioral differences were observed between groups, these were adjusted for in the regression models, and differences over time between groups remained statistically significant. Given the three co-primary outcomes, a Bonferroni correction was applied to adjust for multiple comparisons, setting the significance threshold at α = 0.017 (0.05/3); all observed between-group differences remained statistically significant.
Group-by-time interaction on study outcomes
The repeated-measures ANOVA indicated significant differences in knowledge, attitude, and behavior scores over time between groups. For knowledge, significant differences between groups were observed (F = 17.18, p = 0.001, η2 = 0.086) and time (F = 177.91, p < 0.001, η2 = 0.504), with a significant time × group interaction (F = 11.07, p = 0.001, η2 = 0.059). The intervention group showed significantly greater improvements over time compared to the control group. Participants became more aware of where to find reliable information, preferred fewer drugs per consultation, reported having enough information to use medicines safely while breastfeeding, and recognized the importance of informing the prescriber about all medicines they were using.
Attitude scores showed significant differences between groups (F = 34.64, p < 0.001, η2 = 0.159) and over time (F = 173.66, p < 0.001, η2 = 0.498), with a significant group-by-time interaction (F = 34.45, p < 0.001, η2 = 0.165). Attitudes improved substantially more over time in the intervention group. Participants in the intervention group were more likely to consult healthcare providers than family before taking medicines, recognize that not all medicines are safe without guidance, and reported being less likely to disrupt their breastfeeding schedule when taking medicines.
Behavioral scores showed significant differences between groups (F = 35.24, p < 0.001, η2 = 0.162) and over time (F = 290.75, p < 0.001, η2 = 0.624), with a small but significant group-by-time interaction (F = 5.84, p = 0.017, η2 = 0.032). Safe medicine-use practices improved substantially in the intervention group. Participants demonstrated safer storage of medicines, maintained records of medications taken, reduced the use of leftover medicines, avoided sharing drugs with infants, consulted healthcare providers before using over-the-counter medicines, followed prescribed dosage instructions, and discussed their breastfeeding status with their healthcare provider before taking prescribed medications. The partial eta-squared values (η2) indicate the magnitude of differences associated with the intervention, showing large differences over time across all outcomes, moderate to large between-group differences for attitudes and behaviors, and small to moderate group-by-time interaction differences for knowledge, attitudes, and behaviors, as shown in Table 5.
Table 5. Effect of group to time on study outcomes.
| Outcomes | F | Df | p-value | Partial Eta (η2) |
|---|---|---|---|---|
| Knowledge scores | ||||
| Group | 17.18 | 1 | 0.001 | 0.086 |
| Time | 177.91 | 1 | <0.001 | 0.504 |
| Time*Group | 11.07 | 1 | 0.001 | 0.059 |
| Attitude scores | ||||
| Group | 34.64 | 1 | <0.001 | 0.159 |
| Time | 173.66 | 1 | <0.001 | 0.498 |
| Time*Group | 34.45 | 1 | <0.001 | 0.165 |
| Behavioural scores | ||||
| Group | 35.24 | 1 | <0.001 | 0.162 |
| Time | 290.75 | 1 | <0.001 | 0.624 |
Discussion
This study applied the Information–Motivation–Behavioral Skills (IMB) framework to promote rational medicine use among breastfeeding women and to assess its association with knowledge, attitudes, and practices. At baseline, maternal and infant characteristics were broadly comparable between the intervention and control groups, indicating similarity in demographic and health profiles. Knowledge and attitude scores were also comparable, suggesting similar levels of awareness and beliefs regarding medicine use at study initiation. However, behavioral scores were slightly higher in the intervention group at baseline, reflecting differences in pre-existing safe medicine-use practices, including adherence to dosage instructions, prior consultation with healthcare providers, maintenance of medication records, and disclosure of breastfeeding status during healthcare encounters. Over the three-month follow-up period, the IMB-based educational intervention was associated with improvements in knowledge, attitudes, and safe medicine-use behaviors. Although both groups showed modest improvements over time, greater gains were observed in the intervention group across all outcomes, suggesting the added value of structured, theory-driven education compared with routine health talks.
Participants were predominantly young mothers with modest educational attainment and diverse occupational backgrounds, including homemaking, informal employment, and formal civil service. These characteristics reflect a population of generally low socioeconomic status, which is typical of similar urban settings and may influence access to, interpretation of, and application of health information. Evidence from the literature suggests that higher educational attainment is associated with improved ability to identify medicines that are safe for use during breastfeeding, while women in formal employment may have greater access to health information and health services 25. Although age did not predict baseline scores in this study, evidence from Ethiopia suggests that older mothers may demonstrate greater confidence in medicine-related decision-making, while younger mothers may particularly benefit from structured, guided interventions such as the IMB-based approach used in this study 26. Baseline knowledge and attitude scores were comparable between the intervention and control groups, suggesting similar levels of awareness and beliefs regarding medicine use at study entry. This finding is consistent with the routine provision of maternal and child health education in Uganda, which may contribute to modest improvements in knowledge and attitudes even in the absence of structured, theory-based educational interventions 27. Nonetheless, the greater improvements observed in the intervention group suggest that combining accurate information with motivational components may strengthen understanding of safe medicine-use practices and increase their perceived relevance among breastfeeding women.
Following the intervention, the greatest improvements were observed in knowledge and attitudes, whereas changes in behavior were comparatively more modest. This pattern is consistent with evidence suggesting that improvements in knowledge and attitudes often precede, and do not necessarily translate immediately into, sustained behavior change. Key areas of improvement included increased awareness of the potential effects of medicines on breastfed infants, enhanced understanding of reliable sources of information on medicine use during breastfeeding, greater preference for minimizing unnecessary medication use, improved engagement with healthcare providers, safer medicine storage practices, maintenance of medication records, and reduced use of leftover or shared medicines. These findings are consistent with previous studies demonstrating that theory-driven educational interventions can strengthen maternal knowledge, attitudes, and adherence to recommended health practices, including breastfeeding-related behaviors. For example, Dennis and Mulder reported that IMB-based interventions were associated with improvements in breastfeeding knowledge and attitudes 28, 29. While Fisher et al. demonstrated improved adherence to weight management behaviors among pregnant women following an IMB-based 30. Similar improvements in breastfeeding knowledge and skills have been observed in educational interventions among health professional students 31 and among Jordanian mothers 32. In contrast, a randomized clinical trial conducted in safety-net primary care settings in Washington State found no significant effect of a brief motivational interviewing-based intervention on drug use outcomes compared with enhanced usual care. There were no differences between groups in drug use or addiction severity over 12 months of follow-up 33. The differing results may be due to differences in intervention intensity and duration, as the cited study used a single brief session, whereas our intervention was based on the IMB model and involved repeated health education sessions, group discussions, and experience sharing, which may have provided stronger reinforcement for learning and behavior change than a single brief intervention.
The repeated-measures ANOVA indicated differences in knowledge, attitudes, and safe medicine-use behaviors over time between groups, with the largest improvements observed in knowledge and attitudes. Although behavioral improvements were smaller, they were still notable, suggesting that behavior change may require longer duration, repeated reinforcement, or additional structural support. Effect size estimates further showed strong differences over time across all outcomes, moderate-to-strong between-group differences for attitudes and behaviors, and small-to-moderate group-by-time interaction differences. Baseline differences in behavioral scores were observed and adjusted for in the regression and repeated-measures models to ensure comparisons were based on adjusted estimates. The analysis also accounted for clustering at the clinic level using appropriate statistical methods, reducing potential bias from site-level differences. As all outcomes were self-reported, findings may be subject to recall bias, social desirability bias, and Hawthorne effects, potentially leading to overestimation of reported improvements. However, consistency across outcomes and analytic approaches provides some reassurance, although results should be interpreted with caution. The smaller magnitude of behavior change observed mirrors findings from other IMB-based interventions, including those focused on pregnancy-related weight management, where behavioral modifications typically require more intensive or prolonged exposure 34. Consistent with prior research, both information and motivation significantly influenced safe medicine-use behaviors in this study. Knowledge was associated with behavior both directly and indirectly through behavioral skills, while motivation was associated with behavior both directly and through its association with skills. These findings emphasize the need for interventions that address not only information deficits but also motivational and skill-based determinants of health behavior. The broader literature suggests that theory-guided education often yields greater improvements in knowledge than routine health talks. For example, targeted health education in Ethiopia significantly improved infant feeding knowledge compared to standard antenatal counseling 35.
However, some IMB-based interventions have demonstrated limited improvements in outcomes. A pilot study in South Africa aimed at promoting exclusive breastfeeding among women living with HIV found no significant differences compared to standard care, possibly due to social desirability bias, close researcher–participant relationships, or inadequate intervention design 36. Similarly, an IMB-based intervention among people living with HIV in Iran showed mixed associations with outcomes and significant improvements in information (P = 0.034) and motivation (P = 0.003) for medication adherence, and in information (P = 0.025), motivation (P = 0.001), self-efficacy (P = 0.010), skills (P = 0.011), and dietary adherence (P = 0.011), but no significant improvement in physical activity 37. The absence of improvement in physical activity was attributed to external barriers such as physical limitations and environmental constraints. These findings contrast with the present study, likely because breastfeeding knowledge and skills are more directly responsive to educational interventions than complex behaviors like physical activity.
The findings of this study have implications for maternal health education in low- and middle-income countries, where safe medicine use during breastfeeding is often inadequately addressed within routine health services. The greater improvements observed among participants receiving the IMB-based intervention suggest that incorporating structured, theory-informed educational content into antenatal, postnatal, sexual and reproductive health (SRH), and maternal and child health (MCH) services may enhance the delivery of medicine safety information. The comparatively smaller changes observed in behavioral outcomes highlight the potential importance of complementary strategies, including improved access to pharmaceutical counselling, simplified medicine safety guidance, and consistent messaging from healthcare providers, to support the translation of knowledge and attitudes into practice.
The findings also suggest that intervention design should consider the diverse needs of different sociodemographic groups, including younger mothers, housewives, and women engaged in informal employment. Strengthening the capacity of frontline healthcare workers to apply behavior change frameworks such as the IMB model may further support implementation within routine care settings. However, given the quasi-experimental design, short follow-up period, and absence of economic and clinical outcome evaluations, the findings should be interpreted cautiously. Future research should examine the long-term sustainability of observed changes, evaluate the cost-effectiveness of implementation, and explore applicability in rural and low-literacy settings.
Strengths and limitations
The findings suggest that IMB-based educational interventions may have potential for incorporation into routine maternal and child health services to address persistent gaps in knowledge, attitudes, and practices related to medicine use during breastfeeding. Such approaches may be particularly relevant in settings where self-medication is common and misconceptions regarding medicine safety during breastfeeding remain widespread. The observed findings are consistent with the study hypothesis that mothers receiving the IMB-based intervention would demonstrate greater improvements in knowledge, attitudes, and medicine-use practices than those receiving routine care. The significant time-by-group interaction indicates that the observed differences over time were not solely explained by repeated measurement or temporal trends and were associated with participation in the intervention. This observation is noteworthy because modest improvements were also observed in the control group, likely reflecting exposure to routine health education services; however, the magnitude of change was greater among participants in the intervention group. Future research should assess the long-term sustainability of observed changes, evaluate the cost-effectiveness of implementation, and explore adaptation of the intervention for rural and low-literacy populations.
Among the study limitations was the limited follow-up period of three months, which did not allow assessment of the long-term sustainability of the observed changes in breastfeeding knowledge, attitudes, and skills. Second, behavioural outcomes were self-reported and may therefore be subject to recall and social desirability bias, although the consistency of improvements across knowledge, attitudes, and skills strengthens confidence in the observed trends. Third, the study was conducted in urban and peri-urban settings in Kampala, which may limit generalisability to rural populations with different access to health information and maternal support services. Fourth, although efforts were made to ensure comparability between study groups, the quasi-experimental design without individual randomisation limits causal inference. In particular, the non-random allocation of health facilities based on geographical distance to minimise contamination may have introduced selection bias and cluster-level confounding due to differences in clinic practices, participant characteristics, and routine health education exposure across sites. Allocation based on clinic attendance may have introduced selection bias and cluster-level confounding due to differences in clinic practices, participant characteristics, and routine health education exposure across sites. In addition, unmeasured confounding factors such as prior exposure to similar health messages or peer interactions outside the intervention may have influenced the observed outcomes. Finally, although baseline characteristics were comparable between groups, residual confounding cannot be fully excluded, and therefore, the observed improvements should be interpreted as associations rather than definitive causal effects of the intervention. Despite these limitations, the study reflects real-world implementation conditions and provides useful evidence on the potential effectiveness of IMB-based interventions in routine maternal health settings.
Conclusion
This study suggests that an IMB-based educational intervention was associated with improvements in mothers’ knowledge, attitudes, and safe medicine-use behaviors during breastfeeding compared with routine health talks commonly delivered in maternal and child health services. The most notable differences were observed in motivational outcomes, followed by knowledge, with more modest but meaningful changes in behavior, a pattern consistent with findings from similar behavior change interventions. These results support the potential value of structured, theory-driven approaches in maternal health education while recognising that sustained behavior change may require additional strategies addressing structural and contextual barriers. Given the quasi-experimental design, self-reported outcomes, baseline behavioral differences, and possible clinic-level confounding, the findings should be interpreted with caution. Future interventions should consider longer follow-up periods, strengthened study designs, and context-specific adaptations to better understand long-term and real-world effectiveness in promoting safe medicine use during breastfeeding.
Acknowledgments
We acknowledge the MILK Study team members, specifically JM, MJ, and JK, the participants, the peer mothers, and the healthcare providers at the participating facilities.
Funding
No funding was received for this report; however, the study activities were funded by the Wellcome Clinical Research Career Development Fellowship (222075/Z/20/Z) and the NIHR Global Health Research Professorship NIHR304266, both awarded to CW. RK is funded by the UK Medical Research Council (MR/V03510X/1, MR/V030434/1).
Declarations
Ethics approval
The study obtained expedited ethical clearance for this study from the institutional review board (IRB) of the Infectious Diseases Institute (SPH-2023-448), Uganda National Council of Science and Technology (HS3166ES).
Informed consent to participate
Approval was sought from all study participants. The research team provided detailed information about the study objectives, procedures, risks, and benefits both verbally and through written materials. Participants were given time to ask questions and consider their participation. Written informed consent was then obtained from those who voluntarily agreed to participate. Throughout the study, confidentiality was strictly maintained to protect participants’ privacy.
Consent for Publications
Not applicable.
Authors Contributors
RN, ASP, RK, LA, CW, RM, and CN contributed to the study's conceptualization and design. RN. ASP designed and edited the videos and health education flyers; NS designed the database for data collection; and RN, PM, and CN contributed to the data analysis. RN drafted the manuscript. PM, RN, LA, and CW edited the draft manuscript. All authors reviewed and approved the manuscript before submission for publication.
Competing Interests.
The authors have no competing interests to declare.
Availability of data and materials
https://doi.org/10.5281/zenodo.16960215
References
- 1.Chade ES, Júnior OR, Souza NMP, da Silva AJdOK, Ferreira LM, Reolon JB, et al. The influence of nutritional status on brain development: Benefits of exclusive breastfeeding. Pediatr Rep. 2024;16(3):724–735. doi: 10.3390/pediatric16030061. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lassi ZS, Rind F, Irfan O, Hadi R, Das JK, Bhutta ZA. Impact of infant and young child feeding (iycf) nutrition interventions on breastfeeding practices, growth and mortality in low-and middle-income countries: Systematic review. Nutrients. 2020;12(3):722. doi: 10.3390/nu12030722. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kramer MS, Kakuma R. Optimal duration of exclusive breastfeeding. Cochrane database of systematic reviews. 2012;(8) doi: 10.1002/14651858.CD003517.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Szucs KA, Miracle DJ, Rosenman MB. Breastfeeding knowledge, attitudes, and practices among providers in a medical home. Breastfeed Med. 2009;4(1):31–42. doi: 10.1089/bfm.2008.0108. [DOI] [PubMed] [Google Scholar]
- 5.Nakijoba R, Kiguba R, Dhikusooka F, Namyenya L, Musaazi J, Atuyambe LM, et al. Prevalence, safety evidence, and determinants of medicine use during breastfeeding among women in kampala, uganda. BMC Women’s Health. 2025 doi: 10.1186/s12905-025-03939-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Khalil H, Lee S. Medication safety challenges in primary care: Nurses’ perspective. J Clin Nurs. 2018;27(9-10):2072–2082. doi: 10.1111/jocn.14353. [DOI] [PubMed] [Google Scholar]
- 7.Leape LL, Berwick DM, Bates DW. What practices will most improve safety? Evidence-based medicine meets patient safety. JAMA. 2002;288(4):501–507. doi: 10.1001/jama.288.4.501. [DOI] [PubMed] [Google Scholar]
- 8.Hussainy SY, Dermele N. Knowledge, attitudes and practices of health professionals and women towards medication use in breastfeeding: A review. Int Breastfeed J. 2011 Aug 26;6:11. doi: 10.1186/1746-4358-6-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Meunier S, Coulombe S, Beaulieu MD, Côté J, Lespérance F, Chiasson J-L, et al. Longitudinal testing of the information-motivation-behavioral skills model of self-care among adults with type 2 diabetes. Patient Educ Couns. 2016;99(11):1830–1836. doi: 10.1016/j.pec.2016.06.011. [DOI] [PubMed] [Google Scholar]
- 10.Fisher JD, Amico KR, Fisher WA, Harman JJ. The information-motivation-behavioral skills model of antiretroviral adherence and its applications. Curr HIV/AIDS Rep. 2008;5(4):193–203. doi: 10.1007/s11904-008-0028-y. [DOI] [PubMed] [Google Scholar]
- 11.Gavgani RM, Poursharifi H, Aliasgarzadeh A. Effectiveness of information-motivation and behavioral skill (imb) model in improving self-care behaviors & hba1c measure in adults with type2 diabetes in iran-tabriz. Procedia Soc Behav Sci. 2010;5:1868–1873. [Google Scholar]
- 12.Fullerton T, Rye B, Meaney GJ, Loomis C. Condom and hormonal contraceptive use by young women: An information-motivation-behavioral skills assessment. Can J Behav Sc. 2013;45(3):196. [Google Scholar]
- 13.De Snyder VNS, Friel S, Fotso JC, Khadr Z, Meresman S, Monge P, et al. Social conditions and urban health inequities: Realities, challenges and opportunities to transform the urban landscape through research and action. J Urban Health. 2011;88(6):1183–1193. doi: 10.1007/s11524-011-9609-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mayora C, Kitutu FE, Kandala NB, Ekirapa-Kiracho E, Peterson SS, Wamani H. Private retail drug shops: What they are, how they operate, and implications for health care delivery in rural uganda. BMC Health Serv Res. 2018;18(1):532. doi: 10.1186/s12913-018-3343-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Al-Sawalha NA, Tahaineh L, Sawalha A, Almomani BA. Medication use in breastfeeding women: A national study. Breastfeeding Med. 2016;11(7):386–391. doi: 10.1089/bfm.2016.0044. [DOI] [PubMed] [Google Scholar]
- 16.van Wattum JJ, Leferink TM, Wilffert B, Ter Horst PG. Antibiotics and lactation: An overview of relative infant doses and a systematic assessment of clinical studies. Basic Clin Pharmacol Toxicol. 2019;124(1):5–17. doi: 10.1111/bcpt.13098. [DOI] [PubMed] [Google Scholar]
- 17.Child Y. Infant and young child feeding. Nutrition. 2011:11–13. [Google Scholar]
- 18.Child Y. Uganda policy guidelines on infant and young child feeding. 2007 [Google Scholar]
- 19.Naylor AJ. Lactation management self-study modules: Level. 2009 [Google Scholar]
- 20.Ministry of Health U. Uganda national medicines policy. 2015.
- 21.Tosounidou S, Gordon C. Medications in pregnancy and breastfeeding. Best Pract Res Clin Obstet Gynaecol. 2020;64:68–76. doi: 10.1016/j.bpobgyn.2019.10.007. [DOI] [PubMed] [Google Scholar]
- 22.Thobaben M. How to avoid medication errors: Six tips for clients/caregivers. Home Health Care Manage Pract. 2009;21(5):363–365. [Google Scholar]
- 23.Verstegen RH, Ito S. Drugs in lactation. J Obstet Gynaecol Res. 2019;45(3):522–531. doi: 10.1111/jog.13899. [DOI] [PubMed] [Google Scholar]
- 24.Soon HC, Geppetti P, Lupi C, Kho BP. Medication safety. Textbook of patient safety and clinical risk management. 2020:435–453. [Google Scholar]
- 25.Said F-hA, Okafor CJ, Yusuf SA, Ali AA, Abeid KA, Mahmoud SA, et al. Knowledge levels and practice regarding breastfeeding among mothers with children under 2 years admitted in the pediatric ward at mnazi mmoja hospital, zanzibar. J Adv Med Med Res. 2021;33(13):124–133. [Google Scholar]
- 26.Kassahun A, Zewdie A. Decision-making autonomy in maternal health service use and associated factors among women in mettu district, southwest ethiopia: A community-based cross-sectional study. BMJ Open. 2022 May 2;12(5):e059307. doi: 10.1136/bmjopen-2021-059307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Onyinyechi OM, Mohd Nazan AIN, Ismail S. Effectiveness of health education interventions to improve malaria knowledge and insecticide-treated nets usage among populations of sub-saharan africa: Systematic review and meta-analysis. Front Public Health. 2023;11:1217052. doi: 10.3389/fpubh.2023.1217052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Dennis CL. Breastfeeding initiation and duration: A 1990‐2000 literature review. J Obstet Gynecol Neonatal Nurs. 2002;31(1):12–32. doi: 10.1111/j.1552-6909.2002.tb00019.x. [DOI] [PubMed] [Google Scholar]
- 29.Mulder PJ, Johnson TS, Baker LC. Excessive weight loss in breastfed infants during the postpartum hospitalization. J Obstet Gynecol Neonatal Nurs. 2010;39(1):15–26. doi: 10.1111/j.1552-6909.2009.01085.x. [DOI] [PubMed] [Google Scholar]
- 30.Davies A, Mullin S, Chapman S, Barnard K, Bakhbakhi D, Ion R, et al. Interventions to enhance medication adherence in pregnancy-a systematic review. BMC Pregnancy Childbirth. 2023;23(1):135. doi: 10.1186/s12884-022-05218-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Sandhi A, Nguyen CTT, Lin-Lewry M, Lee GT, Kuo SY. Effectiveness of breastfeeding educational interventions to improve breastfeeding knowledge, attitudes, and skills among nursing, midwifery, and medical students: A systematic review and meta-analysis. Nurse Educ Today. 2023 Jul;126:105813. doi: 10.1016/j.nedt.2023.105813. [DOI] [PubMed] [Google Scholar]
- 32.Khasawneh W, Kheirallah K, Mazin M, Abdulnabi S. Knowledge, attitude, motivation and planning of breastfeeding: A cross-sectional study among jordanian women. Int Breastfeed J. 2020;15(1):60. doi: 10.1186/s13006-020-00303-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Roy-Byrne P, Bumgardner K, Krupski A, Dunn C, Ries R, Donovan D, et al. Brief intervention for problem drug use in safety-net primary care settings: A randomized clinical trial. J Jama. 2014;312(5):492–501. doi: 10.1001/jama.2014.7860. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Skouteris H, Hartley-Clark L, McCabe M, Milgrom J, Kent B, Herring S, et al. Preventing excessive gestational weight gain: A systematic review of interventions. J Obesity Reviews. 2010;11(11):757–768. doi: 10.1111/j.1467-789X.2010.00806.x. [DOI] [PubMed] [Google Scholar]
- 35.Admasu J, Egata G, Bassore DG, Feleke FW. Effect of maternal nutrition education on early initiation and exclusive breast-feeding practices in south ethiopia: A cluster randomised control trial. J Nutr Sci. 2022;11:e37. doi: 10.1017/jns.2022.36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Tuthill EL, Butler LM, Pellowski JA, McGrath JM, Cusson RM, Gable RK, et al. Exclusive breast-feeding promotion among hiv-infected women in south africa: An information-motivation-behavioural skills model-based pilot intervention. Public Health Nutr. 2017 Jun;20(8):1481–1490. doi: 10.1017/S1368980016003657. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Ameri M, Movahed E, Farokhzadian J. Effect of information, motivation, and behavioral skills model on adherence to medication, diet, and physical activity in hiv/adis patients: A health promotion strategy. J Educ Health Promot. 2020;9:317. doi: 10.4103/jehp.jehp_188_20. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
https://doi.org/10.5281/zenodo.16960215