Behavioral Sentinel Surveillance Survey in Nigeria.pdf

Population Council Population Council

Knowledge Commons Knowledge Commons

6-1-2023

Behavioral sentinel surveillance survey in Nigeria: Endline Behavioral sentinel surveillance survey in Nigeria: Endline technical report technical report

Paul Hutchinson

Adolor Aisiri Population Council

Udochisom Anaba

Elizabeth Omoluabi

Akanni Akinyemi

See next page for additional authors

Follow this and additional works at: https://knowledgecommons.popcouncil.org/focus_sexual-health-repro-choice

How does access to this work benefit you? Click here to let us know! How does access to this work benefit you? Click here to let us know!

Recommended Citation Recommended Citation Hutchinson, Paul, Adolor Asiri, Udochisom Anaba, Elizabeth Omoluabi, Akanni Akinyemi, U.C. Ifunanya Ozoadibe, and Dele Abegunde. 2023. "Behavioral sentinel surveillance survey in Nigeria: Endline technical report," Breakthrough RESEARCH Endline Technical Report. Washington, DC: Population Council.

This Report is brought to you for free and open access by the Population Council.

https://knowledgecommons.popcouncil.org/ https://knowledgecommons.popcouncil.org/focus_sexual-health-repro-choice?utm_source=knowledgecommons.popcouncil.org%2Ffocus_sexual-health-repro-choice%2F199&utm_medium=PDF&utm_campaign=PDFCoverPages https://knowledgecommons.popcouncil.org/focus_sexual-health-repro-choice?utm_source=knowledgecommons.popcouncil.org%2Ffocus_sexual-health-repro-choice%2F199&utm_medium=PDF&utm_campaign=PDFCoverPages https://pcouncil.wufoo.com/forms/open-access-to-population-council-research-k63lpf30kmo50n

Authors Authors Paul Hutchinson, Adolor Aisiri, Udochisom Anaba, Elizabeth Omoluabi, Akanni Akinyemi, U. C. Ifunanya Ozoadibe, and Dele Abegunde

This report is available at Knowledge Commons: https://knowledgecommons.popcouncil.org/focus_sexual-health-repro-choice/199 https://knowledgecommons.popcouncil.org/focus_sexual-health-repro-choice/199 https://knowledgecommons.popcouncil.org/focus_sexual-health-repro-choice/199

Behavioral Sentinel Surveillance Survey in Nigeria

JUNE 2023

ENDLINE TECHNICAL REPORT

Centre for Research, Evaluation Resources and Development (CRERD) is a non-profit, non governmental organization concerned with improving the understanding and delivery of population and health services and envi-ronmental development in Nigeria. The organization is an outgrowth of the Operations Research Unit and Network, Obafemi Awolowo University, lIe-lfe, sponsored by The Population Council of New York, U.S.A. for an initial period of four years. CRERD has a network of multidisciplinary researchers all over Nigeria.

As stewards of the first school of public health in the United States, the Tulane University School of Public Health and Tropical Medicine cultivates independent thinkers, innovative leaders, fierce advocates, and accomplished scholars.

From the neighborhoods of New Orleans to communities worldwide, we conduct research and collaborate with our partners to ensure that all of humanity has an equitable opportunity to be healthy and pursue optimal well-being.

We train the problem solvers. Find us on the Front Lines.

Breakthrough RESEARCH is made possible by the generous support of the American people through the United States Agency for International Development (USAID) under the terms of cooperative agreement no.

AID-OAA-A-17-00018. The contents of this document are the sole responsibility of the Breakthrough RESEARCH and Population Council and do not necessarily reflect the views of USAID or the United States Government.

The Population Council confronts critical health and development issues—from stopping the spread of HIV to improving reproductive health and ensuring that young people lead full and productive lives.

Through biomedical, social science and public health research in about 50 countries, the Council works with our partners to deliver solutions that lead to more effective policies, programs, and technologies to improve lives worldwide. Established in 1952 and headquartered in New York, the Council is a nongovernmental, nonprofit organization with an interna-tional board of trustees.

Breakthrough RESEARCH catalyzes social and behavior change (SBC) by conducting state-of-the-art research and evaluation and promoting evidence-based solutions to improve health and development programs around the world. Breakthrough RESEARCH is a consortium led by the Population Council in partnership with Avenir Health, ideas42, Institute for Reproductive Health at Georgetown University, Population Reference Bureau, and Tulane University.

Suggested Citation Hutchinson, P.L., A. Aisiri, U. Anaba, E. Omoluabi, A. Akinyemi, U. C. Ifunanya Ozoadibe, D. Abegunde. 2023. "Behavioral sentinel surveillance survey in Nigeria: Endline technical report," Breakthrough RESEARCH Endline Technical Report. Washington, DC: Population Council.

Contact 4301 Connecticut Avenue NW, Suite 280 | Washington, DC 20008 +1 202 237 9400 | BreakthroughResearch@popcouncil.org breakthroughactionandresearch.org

Credit: AnthonyAbu, GHSC-PSM

©2023 The Population Council. All rights reserved.

Acknowledgments This research report was prepared by Paul Hutchinson, Adolor Aisiri, Udochisom Anaba of Tulane University; Elizabeth Omoluabi, Akanni Ankomah and Dele Abegunde of Population Council. We would like to offer our special thanks to the leadership and teams of supervisors and interviewers at CRERD for their diligent work and attention to detail. We would also like to express our sincere gratitude to Ian Tweedie, Shittu Abdu-Aguye, Mathew Okoh, Stella Babalola, and countless others at Breakthrough ACTION for their input on the evaluation design and invaluable contributions toward the completion of this study. Thanks to Sherry Hutchinson for helpful edits on the report and for assistance with preparing the report. Our gratitude also goes to Foyeke Oyedokun-Adebagbo, INSERT OTHER USAID STAFF IN NIGERIA AND HQ at the United States Agency for International Development for their support and insightful inputs throughout the study.

Behavioral Sentinel Surveillance Survey in Nigeria:

Endline technical report

ENDLINE TECHNICAL REPORT JUNE 2023

Paul Hutchinson1 Adolor Asiri1,2 Udo Anaba1

Elizabeth Omoluabi3 Akanni Akinyemi3 U. C. Ifunanya Ozoadibe3

Dele Abegunde2

1Tulane University 2Breakthrough RESEARCH/Nigeria 3Centre for Research, Evaluation Resources and Development (CRERD)

BR E A K THROUGH R ESE A RCH | JUNE 2023 I

Acronyms ACT Artemisinin-based combination therapy ANC Antenatal care ARI Acute respiratory infections BCG Bacille Calmette-Guérin BSS Behavioral Sentinel Surveillance CHW Community Health Worker CI Confidence intervals CRERD Center for Research, Evaluation Resources, and Development CV Community volunteer DID Difference-in-differences DTP3 Diphtheria-tetanus-pertussis IPTp Intermittent preventive treatment for malaria during pregnancy ITN Insecticide treated net JHU CCP Johns Hopkins University Center for Communication Programs LGA Local government area LLIN Long-lasting insecticidal net MMR Maternal mortality ratio MNCH+N Maternal, newborn, and child health plus nutrition NDHS Nigeria Demographic and Health Survey NHREC National Health Research Ethics Committee NMEP National Malaria Elimination Program NMIS Nigeria Malaria Indicator Survey NPC National Population Commission ORS Oral rehydration solutions PHC Primary health care PMI President’s Malaria Initiative PP Percentage point PPMV Patent and proprietary medicine vendor RDT Rapid diagnostic test SBC Social and behavior change SBCC Social and behavior change communication SMC Seasonal malaria chemoprevention U5MR Under-5 mortality rate USAID United States Agency for International Development

I I BEH AV IOR A L SENTINEL SURV EILL A NCE SURV E Y IN NIGER I A: ENDLINE

Table of Contents Acronyms.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii

Executive.summary

Background Family planning in Nigeria MNCH+N in Nigeria Malaria in Nigeria Breakthrough ACTION in Nigeria: Overview and approach BSS rationale and objectives

Methodology Study location Study design Data collection and questionnaires Data entry and management Study sample and response rates

Data.analysis

Chapter.1..Demographics Key findings

Chapter.2..Impact.of.integrated.versus.vertical.SBC Key findings

Chapter.3..Trends.for.socioeconomic.groups Key finding: Wealth Key finding: Education

Chapter.4..Malaria Key findings

4.1 Malaria prevention

4 .2 Malaria in pregnancy 4 .3 Fever care and treatment

4.4 Seasonal malaria chemoprevention

4.5 Malaria ideations

Chapter.5..Family.planning Key findings

5.1 Contraception: current use, knowledge, and access issues

5.2 Contraception: reasons for use/non-use and intentions

5.3 Contraception: beliefs, perceptions, and self-efficacy

BR E A K THROUGH R ESE A RCH | JUNE 2023 I I I

Chapter.6..Antenatal,.delivery,.and.newborn.care 6 .1 Antenatal care 6 .2 Delivery care

6.3 Newborn care

Chapter.7..Nutrition.and.breastfeeding Breastfeeding key findings Nutrition key findings

7.1 Child nutrition and breastfeeding

7.2 Breastfeeding ideations

Chapter.8..Child.health

8.1 Vaccination

8.2 Acute respiratory infections

8.3 Diarrheal diseases

8.4 Child health ideations

Chapter.9..Gender Key findings

9.1. Gender roles

Chapter.10..Breakthrough.ACTION/Nigeria.program.exposure Key findings 10 .3 Albishirin Ku!

Chapter.11..Examining.the.interplay.of.norms,.attitudes,.beliefs.with.priority.behaviors

11.1 Modern contraceptive use

11.2 Intermittent preventive treatment of malaria during pregnancy

11.3 Exclusive Breastfeeding

12..Conclusion.and.recommendations Impact—integrated versus vertical SBCC programs Modern contraception Antenatal care and delivery Newborn care Infant Nutrition and Breastfeeding Childhood vaccination Childhood illnesses—Treatment Exposure to Albishirin Ku!

Appendices

References

I V BEH AV IOR A L SENTINEL SURV EILL A NCE SURV E Y IN NIGER I A: ENDLINE

List of Tables Table A-1. Key objectives of the life stage approach for integrated SBC programming areas

Table A-2 Study overview

Table A-3 2022 BSS Sample Design

Table A-4 Priority indicators and minimal detectable differences

Table A-5 Study sample and response rates

Table 1.1.1. Study sample

Table 1.1.2. Female respondent characteristics

Table 1.1.3. Female respondent characteristics

Table 1.1.4. Female respondent occupations

Table 2.1. Priority Indicators and Difference-in-Differences Estimates, Kebbi v. Zamfara

Table 2.2 Priority Indicators and Difference-in-Differences Estimates, Sokoto v. Zamfara

Table 3.1.1 Modern Contraceptive Use, by Wealth Quintile

Table 3.1.2 Intermittent Preventive Treatment Against Malaria, by Wealth Quintile

Table 3.1.3 or More ANC visits, by Wealth Quintile

Table 3.1.4 Facility Delivery, by Wealth Quintile

Table 3.1.5 Exclusive Breastfeeding, by Wealth Quintile

Table 3.2 Trends in indicators, by women’s education levels

Table 4.1.1. LLIN availability and use in household

Table 4.2.1. IPTp during last pregnancy

Table 4.2.2. Reasons for no IPTp for malaria during last pregnancy

Table 4.2.3 What are some risks to the baby if a woman gets malaria during pregnancy?

Table 4.3.1a. Care and treatment for pediatric fevers

Table 4.3.1b. Care and treatment for pediatric fevers

Table 4.3.2. Reasons for testing the febrile child for malaria

Table 4.3.3. Reasons for not testing the febrile child for malaria

Table 4.3.4. Care-seeking for pediatric fevers

Table 4.3.5. Reasons for choosing source of care for pediatric fevers

Table 4.3.6. Reasons for not seeking care for pediatric fevers

Table 4.3.7. Anti-malarial treatment for pediatric fevers

Table 4.4.1. SMC awareness and uptake

Table 4.4.2. Reasons for not taking SMC medicines to prevent malaria

Table 4.5.1. Reported causes of malaria in children

Table 4.5.3. Malaria-related beliefs and attitudes: Kebbi v Zamfara

Table 5.1.1. Current use of modern contraceptive methods

Table 5.1.2. Have you ever heard of the following contraception methods?

Table 5.1.3. Where did you obtain the [current method] the last time?

Table 5.1.4. Difficulties faced obtaining [current method] the last time?

Table 5.1.5. Content of contraceptive discussions with health provider

Table 5.2.1. Reasons for using a contraceptive method

Table 5.2.2. Reasons for not using a contraceptive method

BR E A K THROUGH R ESE A RCH | JUNE 2023 V

Table 5.2.3. Intention to use/continue to use contraception in the next 6 months

Table 5.3.2. Perceived contraception benefits for mothers

Table 5.3.2. Perceived contraception benefits for future children

Table 5.3.3a. Contraception beliefs, attitudes, and perceived social norms—Kebbi v Zamfara

Table 5.3.3b. Contraception beliefs, attitudes, and perceived social norms—Sokoto v Zamfara

Table 5.3.5. Spousal communication about contraception

Table 5.3.6. Whose decision is it to use contraception?

Table 6.1a. Antenatal Care Visits and Difference-in-Differences Estimates, Kebbi v. Zamfara

Table 6.1b. Antenatal Care Visits and Difference-in-Differences Estimates, Sokoto v. Zamfara

Table 6.1.1. Pregnancy intentions during last pregnancy

Table 6.1.2. Source and timing of ANC during last pregnancy

Table 6.1.3. Whom did you see for the first ANC visit during your last pregnancy

Table 6.1.4. ANC Content during last pregnancy

Table 6.1.5. Medicines or vaccinations received during last pregnancy

Table 6.1.6. Birth planning during last pregnancy

Table 6.1.7. Knowledge of ANC timing and visits

Table 6.1.8. Reasons for not attending ANC during last pregnancy

Table 6.1.9. Influencers of Antenatal Care Visits

Table 6.1.10. Self-efficacy for ANC

Table 6.1.11. Perceived maternal benefits of ANC

Table 6.1.12. What are some danger signs during pregnancy?

Table 6.2.1. Delivery location and respectful care during the last pregnancy

Table 6.2.2. Who assisted in the delivery of your child during the last pregnancy?

Table 6.2.3. Reasons for not delivering in a health facility during the last pregnancy

Table 6.2.4. Who else influences decision about facility delivery?

Table 6.2.5. Spousal discussion about delivery care

Table 6.2.6. Pregnancy-related beliefs, attitudes and values, Kebbi v Zamfara

Table 6.2.7. Pregnancy-related beliefs, attitudes and values, Sokoto v Zamfara

Table 6.2.8. Self-efficacy and intentions for facility birth during the next pregnancy

Table 6.3.1. Newborn care immediately after delivery: Facility deliveries

Table 6.3.1b. Newborn care immediately after delivery: Home Deliveries

Table 6.3.2. Postpartum care for mothers after delivery

Table 6.3.3. What are some ways to protect a newborn immediately after birth?

Table 6.3.4. What are some ways to help a mother recover well after birth?

Table 7.1.1. Breastfeeding practices immediately after birth

Table 7.1.2. Breastfeeding practices among infants aged 0-5 months

Table 7.1.3. Reasons for choosing not to ever breastfeed your child

Table 7.1.4. Reasons for giving non-breastmilk liquids to a newborn in the first three days after birth

Table 7.1.5. Breastfeeding and nutrition among children aged 6–23 months

Table 7.2.1. Perceived benefits of exclusive breastfeeding for the mother

Table 7.2.2. Perceived benefits of exclusive breastfeeding for the baby

Table 7.2.3. Breastfeeding knowledge, beliefs, intentions, and self-efficacy

Table 7.2.4. Breastfeeding knowledge, beliefs, intentions, and self-efficacy

Table 7.2.5. Breastfeeding knowledge, beliefs, intentions, and self-efficacy

V I BEH AV IOR A L SENTINEL SURV EILL A NCE SURV E Y IN NIGER I A: ENDLINE

Table 7.2.6. Who else influences decisions about breastfeeding?

Table 7.2.7. What are some ways to recognize poor nutrition in a child?

Table 8.1.1 Vaccination practices for children 12–23 months

Table 8.1.2 Vaccinated children 12–23 months by type of vaccine

Table 8.1.3. Reasons for not vaccinating a child

Table 8.1.4. Who else influences decisions about getting a child vaccinated?

Table 8.1.5. Vaccine knowledge, perceptions, self-efficacy, and intentions

Table 8.1.6 Vaccine knowledge, perceptions, self-efficacy, and intentions

Table 8.1.7. Vaccine knowledge, perceptions, self-efficacy, and intentions

Table 8.2.1. ARI symptoms

Table 8.2.2. ARI care-seeking by source

Table 8.2.3. ARI treatment by medication type

Table 8.2.4. Reasons for choosing source of care for ARI

Table 8.2.5. Reasons for not seeking care for ARI

Table 8.2.6. Reported signs and symptoms of pediatric pneumonia

Table 8.2.7. Reported ways to prevent pediatric pneumonia

Table 8.2.8. Perceived treatment effectiveness against pneumonia

Table 8.2.9. Pneumonia-related beliefs, intentions, and self-efficacy

Table 8.2.10. Pneumonia-related beliefs, intentions and self-efficacy

Table 8.3.1. Diarrhea prevalence and care-seeking for pediatric diarrhea

Table 8.3.2. Reasons for choosing source of care for pediatric diarrhea

Table 8.3.3. Reasons for not seeking care for pediatric diarrhea

Table 8.3.4. Fluids or treatments given for pediatric diarrhea

Table 8.3.5. Reported ways to prevent a child from getting diarrhea

Table 8.3.6. Treatments perceived as effective against pediatric diarrhea

Table 8.3.7. Reported ways ORS may help a child with diarrhea

Table 8.3.8. Reported ways zinc may help a child with diarrhea

Table 8.3.9. Diarrhea-related beliefs, intentions, and self-efficacy

Table 8.3.10. Diarrhea-related beliefs, intentions, and self-efficacy

Table 8.4.1. Perceptions of health services quality and self-efficacy in convincing partner to seek care for sick children

Table 8.4.2. Perceptions of health services quality and self-efficacy in convincing partner to seek care for sick children

Table 8.4.3. Care-seeking influencers for sick children

Table 9.1.1. Women’s role in household decision-making

Table 9.1.2. Gender-related beliefs, attitudes, and perceptions

Table 9.1.3. Gender-related beliefs, attitudes, and perceptions

Table 10.1.1. Albishirin Ku! and Breakthrough ACTION/Nigeria exposure

Table 10.2.1. Exposure to any Albishirin Ku! radio program, by Wealth Quintile

Table 10.2.2. Exposure to any Breakthrough ACTION/Nigeria activities, by Wealth Quintile

Table 10.2.3. Exposure to any community dialogue/event, by Wealth Quintile

Table 10.2.4. Exposure to any CV household visit, by wealth quintile

Table 10.3.1. Albishirin Ku! radio programming perceptions

Table 10.3.2. Albishirin Ku! radio exposure and health messages heard on the radio

Table 10.3.3. Malaria jingles or shows heard on the radio

Table 10.3.3. Media Partnership Programming shows heard on the radio

BR E A K THROUGH R ESE A RCH | JUNE 2023 V I I

Table 11.1. Logistic regression: Current use of modern contraception

Table 11.2. Logistic regression: 3+ doses of SP/Fansidar as part of intermittent preventive treatment of malaria during pregnancy

Table 11.3. Logistic regression: Exclusive breastfeeding

V I I I BEH AV IOR A L SENTINEL S URV EILL A NCE SURV E Y IN NIGER I A: ENDLINE

Executive Summary This technical report presents results from the Behavioral Sentinel Surveillance (BSS) endline survey undertaken by Breakthrough RESEARCH/Nigeria in Kebbi, Sokoto, and Zamfara states between October 1 and November 10, 2022. BSS surveys are intended to assess changes in indicators targeted by the integrated social and behavior change (SBC) activities of the USAID-funded Breakthrough ACTION/Nigeria project. The Breakthrough ACTION/Nigeria project, which began in 2019 and is slated to run until 2025, focuses on the health areas of malaria; family planning; and maternal, newborn, and child health plus nutrition (MNCH+N) in Kebbi and Sokoto states, as well as malaria-only SBC activities in Zamfara State. The project uses three primary SBC approaches:

advocacy outreach to opinion leaders and community influencers at the state and local government area (LGA) levels, direct engagement of community members through community dialogues and group meetings, and SBC messaging campaigns through mass media and digital media.

The primary objective of this study is to compare the effectiveness of integrated SBC programming, as imple-mented by Breakthrough ACTION/Nigeria in Kebbi and Sokoto states, with single-focused vertical SBC program-ming, used by Breakthrough ACTION/Nigeria to target malaria outcomes in Zamfara state.

About the BSS The endline BSS follows prior survey waves conducted in 2019, at the inception of Breakthrough ACTION/Nigeria, and again in 2021, at the project’s originally intended midpoint. This report presents baseline-to-endline comparisons of female respondents’ behaviors, atti-tudes, knowledge, self-efficacy, and beliefs related to malaria, family planning and MNCH+N, as well as their knowledge, perceived social norms, and intentions.

The survey employed a multi-stage sampling design to gather information from women of reproductive age on health behaviors and the factors that shape them. Using interviewer-directed, face-to-face interviews, the survey collected data on psychosocial influences across cog-nitive, emotional, and social domains. These influences have been identified by a range of behavioral theories as intermediate determinants of health decision-making and behavioral outcomes. As these theories suggest, an individual’s beliefs, attitudes, perceptions of social norms, and sense of self-efficacy can all play roles in shaping their actions. By addressing these perceptions through Breakthrough ACTION/Nigeria interventions, positive changes in behavior may be encouraged. For this wave of the BSS, a total of 3,144 women between the ages of 15 and 49 years, each with a child under 2 years old, were randomly selected from wards within Breakthrough ACTION/Nigeria program areas. Data were also collected from the women’s husbands, but analysis of couples’ knowledge, attitudes, and norms, as well as couples’ dynamics and decision-making are the subject of a com-panion report and are not addressed in detail here.

To examine the causal impact of integrated SBC relative to vertical SBC, we examined trends in key behavioral indicators in the integrated states of Kebbi and Sokoto relative to those in Zamfara using both unadjusted and adjusted difference-in-differences (DID) models. DID models, frequently used in evaluation research, compare the difference in an outcome in one group before and after receiving treatment (i.e., baseline to endline) with the difference in that same outcome over the same period in a control group. In this case, Kebbi and Sokoto states, the treatment states, received integrated pro-gramming, while Zamfara, the comparison state, received vertical malaria-only SBC. The difference between these two differences (i.e., the difference-in-differences) is the treatment effect, reflecting whether integrated or vertical SBC has a greater effect on an outcome. Positive (and statistically significant) estimates suggest that the integrated states had greater improvement in outcomes than Zamfara, while negative estimates indicate the opposite.

We improved the validity of the causal interpretation of integrated versus vertical SBC using adjusted DID models that control for potential confounding factors that could also be contributing to changes over time (or lack thereof). Multivariate DID models controlled for differences in women’s characteristics (e.g., age, level of education, parity), husband’s characteristics (e.g., level of education), and household characteristics (e.g., asset-based wealth).

The most important tests of integrated versus vertical SBC are in the area of malaria where both approaches

BR E A K THROUGH R ESE A RCH | JUNE 2023 1

seek to change the same indicators, albeit with their different approaches. On other indicators (e.g., family planning, pregnancy-related care excluding intermittent preventive treatment for malaria during pregnancy [IPTp], breastfeeding, child vaccinations and treatment of acute respiratory infection [ARI] and diarrhea), comparisons between the integrated and vertical states primarily test the effectiveness of integrated SBC versus no SBC, since Zamfara would have no SBC programming targeting these indicators.

Because this was a large-scale evaluation covering wide geographic areas, there were inevitably real-world factors that could have differentially affected each state—the timing of mass distribution of long-lasting insecticide treated nets (LLINs), varying levels of insecu-rity, challenges with the distribution of drugs and other commodities—as well as potential spillover of Albishirin Ku! radio programming from Sokoto to Zamfara, that would have potentially affected non-malaria outcomes in the latter state. Nevertheless, the paucity of at-scale tests of these two common approaches to SBC program-ming provided more than sufficient justification for this research endeavor.

Key Findings

Impact In aggregate, our results suggest that integrated SBC interventions may be more effective in promoting posi-tive health outcomes across various domains, including malaria prevention, family planning, pregnancy-related care, and treatment of childhood illnesses, compared to vertical programming approaches.

For malaria, Breakthrough ACTION/Nigeria uses intensive programming focused solely in this health area in Zamfara to target malaria outcomes, but in the integrated states the program incorporates malaria mes-saging within branded programming. The key behaviors targeted by Breakthrough ACTION/Nigeria for malaria are LLIN use and care by vulnerable groups (pregnant women and children under age two), IPTp, malaria testing for children with fever, prompt treatment of positive malaria cases with artemisinin-based combination therapy (ACT), and children’s receipt and completion of seasonal malaria chemoprophylaxis provided by community health workers.

During the study period, differential timing of mass LLIN distribution across the three states introduced the potential for diminishing the effects of malaria preven-tion SBC efforts by Breakthrough ACTION/Nigeria in the integrated SBC states relative to the vertical SBC state.

In Zamfara, where a mass LLIN distribution occurred in 2020, the percentage of households owning at least one net increased by 12 pp, from 74% at baseline to 88% at endline. Net use, however, increased by a considerably larger amount. In households with at least one LLIN, the percentage of pregnant women sleeping under an LLIN nearly doubled from 39% at baseline to 77% at endline, while the percentage of children under the age of 2 sleeping under an LLIN increased by 35 pp from 54% to 89%. These sizable increases in LLIN use suggest that the gains are likely due to a combination of improved access to LLINs and an increase in motivation among households in Zamfara, potentially (but not conclusively) driven by effective interventions by Breakthrough ACTION/Nigeria.

Absent mass distribution campaign since 2018 in Kebbia and since 2017 in Sokoto, the percentage of households with at least one LLIN decreased considerably in Kebbi (from 78% at baseline to 46% at endline) and in Sokoto (from 80% at baseline to 61% at endline. In spite of this, lower LLIN ownership did not translate into sharp declines among the priority groups. Among households with at least one LLIN, use by pregnant women increased in both Kebbi (29% at baseline, 51% at endline) and Sokoto (31% at baseline, 41% at endline). Similarly, among households with at least one LLIN, use by children under age 2 improved from 38% to 68% in Kebbi and from 47% to 66% in Sokoto.

The integrated programming of Breakthrough ACTION/ Nigeria seeks to improve the uptake of modern con-traception by targeting norms, attitudes, and beliefs related to birth spacing and contraceptive use. The primary family planning indicator, current use of modern contraception, indeed improved more in the integrated state of Kebbi relative to Zamfara, but this result was not statistically significant in either unadjusted or adjusted regression models (Adjusted DID=9.8 percentage points (pp), p=.392; unadjusted DID=2.8 pp, p=.547).

Contraceptive use indicators in Sokoto showed slight declines, leading to negative DID estimates relative to Zamfara, but these too were not statistically significant.

aThe decline in Kebbi was partially mitigated by a mass distribution campaign of LLINs that began toward the end of the data collection period, with one in five households in Kebbi receiving a net at that time.

2 BEH AV IOR A L SENTINEL SURV EILL A NCE SURV E Y IN NIGER I A: ENDLINE

While pregnancy-related outcomes improved in both Kebbi and Sokoto, improvements were larger in Zamfara, where Breakthrough ACTION/Nigeria worked to improve uptake of IPTp. For example, the percentage of women in Kebbi who made four or more ANC visits increased by 8 pp, from 24% at baseline to 32% at endline. However, this same indicator increased by a larger 12 pp in Zamfara, from 26% to 38%. A similar pattern was observed for facility deliveries as well. The percentage of births occurring in a health facility increased by 7 pp in Kebbi (15% at baseline, 23% at endline) but by a larger amount, 11 pp, in Zamfara (16% at baseline, 27% at endline). This pattern may reflect differences in the characteristics of the respondents in the two states, with those in Zamfara having greater economic resources and higher education levels. After controlling for these differences in regression analysis, we observed more pronounced improvements in Kebbi relative to Zamfara, as indicated by adjusted DID estimates of 3.5 pp for ANC4+, 4.0 pp. for facility deliver-ies, and 10.4 pp for IPTp, although these results are not statistically significant.

The integrated states generally performed better than the vertical state in terms of ensuring appropriate treat-ment for common child illnesses. For example, treatment of diarrhea with oral rehydration solutions (ORS)/zinc improved by 12 pp in Kebbi (from 18% to 30%) and by 11 pp in Sokoto (from 12% to 23%), as compared with a decline of 4 pp (from 27% to 23%) in Zamfara. In adjusted DID models, this yielded statistically significant treatment effects of 24 pp for Kebbi relative to Zamfara and of 16 pp for Sokoto relative to Zamfara. Similarly, treatment with antibiotics for children with ARI symptoms increased by 22 pp (from 41% to 63%) in Kebbi and by 12 pp (from 26% to 38%) in Sokoto, while this indicator declined by 24 pp (from 53% to 28%) in Zamfara during the same period.

Hence, adjusted treatment effects were 45 pp for Kebbi and 35 pp for Sokoto and were statistically significant.

While these large effects could possibly be attributable to the integrated SBC targeting treatment of child illness in Kebbi and Sokoto, it could also be attributable, at least in part, to frequent stockouts of antibiotics in Zamfara.

Breastfeeding indicators worsened in all three states, with none of the adjusted DID estimates being statisti-cally different from one state to the next. Full childhood immunization rates increased in both Kebbi and Zamfara, but the magnitudes of the changes were small, 4 pp in Zamfara and 3 pp in Kebbi, and both unadjusted and adjusted DID treatment effects were not statistically different from zero.

Malaria IPTp during pregnancy: The percentage of women who reported receiving the recommended three or more (3+) doses of SP/Fansidar—intermittent preventive treatment for malaria during pregnancy (IPTp) - increased by nearly 10 pp from baseline to endline in Kebbi (23% at baseline, 34% at endline) but changed minimally in the other two states—Sokoto (15% at baseline, 15% at endline) and Zamfara (19% at baseline, 21% at endline). This yielded an adjusted DID estimate of 10 pp in Kebbi relative to Zamfara, but this result was not statistically significant.

As in the baseline, the majority of women received their first dose in their second trimester. The three most common reasons cited for not taking IPTp were that it was not necessary because the pregnant woman was not sick, that the woman did not attend ANC, or that preg-nancy outcomes are “Up to God.” In Kebbi, there was an increase in awareness from baseline to endline of all risks to the baby if the mother contracts malaria during pregnancy, including premature birth (26% at baseline, 44% at endline), low birth weight (13% at baseline, 27% at endline), and miscarriage (35% at baseline, 57% at endline). In Sokoto, awareness of risks declined, while the results for Zamfara were mixed.

Pediatric fever prevalence and treatment: Across all three states, a minority of children with fever were given a malaria blood test, although there was a sizable increase in testing in Sokoto (17% at baseline, 25% at endline) that was not matched in the other two states. In Kebbi, testing rates fell (23% at baseline, 15% at endline), while they increased in Zamfara (from 24% at baseline to 30% at endline). None of the estimated adjusted DIDs were statistically significant, however. The most common reason for not testing was that respondents felt the test was unnecessary (Kebbi: 52%, Sokoto: 36%, Zamfara: 23%) or that the test was not offered (Kebbi:

28%, Sokoto: 25%, Zamfara: 33%).

Sokoto showed increases in several treatment-related behaviors, including increases in obtaining medicines for fever (74% at baseline, 91% at endline), treatment with ACT (18% at baseline, 27% at endline), seeking care (54% at baseline, 79% at endline). These latter results yielded DID estimates relative to Zamfara of 14.8 pp (p<.01),

7.6 pp (NS), and 22.8 pp (p<.01) respectively. However, prompt treatment within the first 24-48 hours showed no improvement in Sokoto, decreasing slightly from 65% to 61%. Across all treatment seeking indicators, Zamfara showed no improvement, while Kebbi showed a wors-ening in several indicators, except for prompt treatment

BR E A K THROUGH R ESE A RCH | JUNE 2023 3

of fever within 24-48 hours, which increased from 62% to 81%, yielding an adjusted DID estimate of 22 pp (NS).

There was also an improvement in completion of the full course of ACT in Kebbi, from 62% to 94%, as compared with an improvement from 79% to 95% in Zamfara, and a decline in Sokoto from 82% to 73%.

Seasonal malaria chemoprevention (SMC): At endline, most women were aware of medicine given to children during the rainy season to prevent malaria (Kebbi:

88%, Sokoto: 86%, Zamfara: 79), and most households reported that a CHW had visited their household during the rainy season to distribute chemoprophylaxis to children (Kebbi: 86%, Sokoto: 71%, Zamfara: 64%). For both indicators, prevalence was higher in the integrated states than in the intense malaria-only state. Among households that were visited in all three states, most children (Kebbi: 86%, Sokoto: 71%, Zamfara: 76%) took the medicine provided by the CHW. For those that did not, the most common reasons given were that the child was unavailable during the visit, that the child was not sick, or that they forgot.

Malaria ideations: Even though Zamfara state received more intensive malaria-related SBC, the adjusted DID estimates indicate that malaria ideational factors such as attitudes, norms, and beliefs improved more in Kebbi relative to Zamfara, suggesting that malaria SBC efforts may have been more effective there, although almost none of the adjusted DID estimates were statistically different from zero. For example, in Kebbi, there was a more significant reduction in the perception that a child’s fever is nearly always malaria, with an adjusted DID of 4.3 percentage points. Similarly, beliefs surrounding malaria prevention, like the necessity of children under five and pregnant women sleeping inside nets, increased more in Kebbi (with adjusted DID estimates of 4.9 and 3.6 percentage points respectively). Sokoto, too, improved more than Zamfara in several of the attitudinal and behavior indicators related to malaria, while Zamfara showed greater improvement in other areas. However, the magnitudes of the improvements in ideational indicators in Sokoto relative to Zamfara were generally larger for those indicators where it did better, suggesting overall more effective malaria SBC interventions in Sokoto. Again, however, the adjusted DID estimates were generally not statistically significant.

Family planning Contraceptive use: Modern contraceptive use increased in both Kebbi (from 9% to 16%) and Zamfara (from 16% to 19%) but these increases were more sizable in Kebbi, possibly a result of Breakthrough ACTION/Nigeria efforts.

In contrast, modern contraceptive use declined slightly in Sokoto (from 11% to 9%). The improvements in Kebbi and Zamfara were primarily driven largely by a rise in the use of injectables (Zamfara: from 8.3% to 9.4%, Kebbi: from 5.8% to 6.4%) and implants (Zamfara: from 3.9% to 5.6%, Kebbi: from 2.1% to 5.3%). Additionally, oral contracep-tion usage increased by 2.0 percentage points in Kebbi (from 1.2% to 3.2%). The changes in postpartum modern contraceptive use mirrored those of the full sample of women with a child under two years.

Contraceptive beliefs, attitudes and myths: The situation regarding contraceptive beliefs, attitudes, and myths is complex, with a range of indicators showing improve-ments in both the integrated states of Kebbi and Sokoto, as well as in Zamfara. In Sokoto, there was an increase in women’s approval of contraception for birth spacing from 36% to 44%, while the same indicator remained largely unchanged in Kebbi (48% at baseline and 46% at endline). The percentage of women who believed that it was important for couples to discuss family planning increased slightly in Sokoto, from 63% to 70%, but increased by a larger margin in the malaria-only state of Zamfara, moving from 73% to 82%. There was no change observed in Kebbi. Across all three states, women at end-line were more likely to agree that couples using family planning have a better quality of life, with the increases in Zamfara being higher than those in the other two states.

There was little change, however, in the percentage of women who agreed that using contraception would lead to negative social repercussions, such as being called bad names by others in the community, which remained low in Kebbi (15%) and Zamfara (17%) but considerably higher in Sokoto (34%).

MNCH+N

Antenatal care: In Kebbi, there was an increase in the percentage of women who made their first antenatal care (ANC) visit during their first trimester, rising from 31% at baseline to 36% at endline. However, this per-centage remained unchanged in Sokoto, with 26% at baseline and 25% at endline. In Zamfara, there was a slight decrease, with 21% at baseline and 16% at endline.

The vast majority of pregnant women who went for ANC attended either a government hospital or primary health care, with slight increases in each of the three states (Kebbi: 87% at baseline and 92% at endline; Sokoto: 79%

4 BEH AV IOR A L SENTINEL SURV EILL A NCE SURV E Y IN NIGER I A: ENDLINE

at baseline and 82% at endline; Zamfara: 88% at baseline and 90% at endline).

There was a marked improvement in all three states in the services provided during ANC visits for women who attended. Over 90% of the women who attended had their blood pressure measured, and most had urine (Kebbi: 66% at baseline and 78% at endline; Sokoto: 67% at baseline and 77% at endline; Zamfara: 80% at baseline and 88% at endline) and blood samples taken (Kebbi: 72% at baseline and 91% at endline; Sokoto: 78% at baseline and 76% at endline; Zamfara: 82% at baseline and 92% at endline). The discussion of a child’s due date with a pro-vider also showed improvement, with increases in Kebbi (45% to 73%), Sokoto (40% to 52%), and Zamfara (54% to 84%). However, the distribution of mosquito nets during ANC visits decreased in all states. Over 60% of women at endline reported taking iron supplements during their most recent pregnancy, while over 40% reported taking folic acid. In all three states, there were improvements in both indicators, except for folic acid supplement uptake, which saw only a marginal increase in Zamfara.B

Although attendance of 4+ ANC visits remained low, there was a notable increase in the percentage of women who were aware of the recommended number of ANC visits in all three states, with the largest increase occurring in Kebbi (29% at baseline, 51% at endline) but notable increases in the other two states as well (Sokoto:

36% at baseline, 45% at endline; Zamfara: 50% at baseline, 59% at endline. Among women who chose not to make any ANC visits during their last pregnancy, the most common reasons were a lack of perceived need, spousal opposition, facility distance, and fatalism (“Up to God”). Other provider-side barriers, such as closed facility, mistrust of provider, poor treatment by provider, poor quality of service, or absence of a female provider, were generally cited by fewer than 2% of respondents in Kebbi and Sokoto, but by slightly more in Zamfara but still fewer than 5% of respondents.

The development of a birth plan with the assistance of a healthcare provider does not seem to be a common practice in the study areas. The endline results indicate a continued need for more intensive SBC messaging to improve this practice in the future. At endline, only Zamfara state had more than 50% of women reporting discussions of birth plans with a healthcare provider, which was a significant increase from 35% at baseline. Of the women who had discussions about birth planning, the discussion mostly centered on the delivery location.

ANC decision-making, self-efficacy, and intentions:

Husbands clearly play an important role in the use of antenatal and other pregnancy-related care. Nearly all women in Kebbi (94%) reported that their husband influences their decision to make 4 or more ANC visits.

Women acknowledge the importance of discussing their pregnancies with their husbands, and this recognition increased in all three states over the study period (Kebbi:

91% at baseline and 95% at endline; Sokoto: 88% at baseline and 91% at endline; Zamfara: 95% at baseline and 97% at endline). Commendably, there were also improvements in women’s confidence in initiating a conversation with their husband about ANC (Kebbi: 77% at baseline and 78% at endline; Sokoto: 65% at baseline and 75% at endline; Zamfara: 74% at baseline and 86% at endline).

Breakthrough ACTION/Nigeria SBC efforts may have improved ANC knowledge and beliefs in Kebbi, where there were notable declines in the number of respon-dents agreeing with ANC-related myths. In contrast, myths about ANC remain common in Sokoto and Zamfara states, where beliefs in these myths may have actually increased since the baseline. For instance, the percent-age of women who agreed that only first-time pregnant mothers need ANC decreased from 44% to 38% in Kebbi, while it increased in Zamfara from 34% to 39%. In regression adjusted models, this yielded a DID of -17 pp.

Facility delivery: Across all three states, there were large reductions in the percentage of births assisted by family, friends, and neighbors, with much of this reduction made up for by increased reliance on traditional birth atten-dants in Kebbi and by doctors, midwives, and community health workers (CHWs) in Zamfara. The most common reason for not delivering in a health facility was that it was “not necessary to go,” but this reason declined in importance in all three states: Kebbi (76% at baseline, 45% at endline), Sokoto (74% at baseline, 58% at endline), Zamfara (61% at baseline, 37% at endline).

As with ANC, the majority of women (Kebbi: 93%: Sokoto:

85%; Zamfara: 81%)) reported that their husbands influence their decision to have a facility delivery. Even so, while husbands are critical to decision-making, discussions with husbands about where to deliver are infrequent; only 21% of women in Kebbi, 13% in Sokoto, and 30% in Zamfara reported that they had discussed where to deliver. This absence of discussion is puzzling, as most women—66% in Kebbi, 56% in Sokoto, and 71% in

BR E A K THROUGH R ESE A RCH | JUNE 2023 5

Zamfara—indicate that they are confident that they could initiate such a discussion.

While many women cited lack of perceived need as a reason for not delivering in a health facility, 73% (Kebbi), 53% (Sokoto), and 64% (Zamfara) of women agreed that a health facility is the best place to deliver a baby, with larger increases in the integrated states relative to Zamfara (adjusted DID Kebbi—Zamfara: 12 pp; adjusted DID Sokoto—Zamfara: 4.5 pp), although these estimates were not statistically significant. Further, increasing majorities of women in Kebbi: (56% at baseline, 63% at endline) and Zamfara (53% at baseline, 66% at endline) reported intentions to deliver their next child at a health facility.

Newborn care: Among women who delivered in a health facility, nearly all newborn practices improved in all three states. The percentage of newborns who were wiped dry after birth increased in Kebbi (89% at baseline, 97% at endline), Sokoto (84% at baseline, 95% at endline), and Zamfara (94% at baseline, 97% at endline). The percent-age of mothers who reported placing the newborn on a bare chest immediately after birth also increased in Kebbi (56% at baseline, 78% at endline) and Zamfara states (65% at baseline, 84% at endline), while Sokoto experienced a decline from 72% at baseline to 60% at the endline. At endline, nearly all babies delivered in facilities had their umbilical cord cut with a new blade (Kebbi:

98%, Sokoto: 96%, Zamfara: 98%), an improvement from baseline (Kebbi: 82%, Sokoto: 94%, Zamfara: 92%). About three quarters of facility-born infants at endline in Kebbi (76%) and Zamfara (77%) had chlorhexidine applied to their stump at the endline, higher than in Sokoto (61%).

Among babies delivered outside of formal health facili-ties, the above practices were less commonly followed.

The percentage of newborns who were wiped dry after birth decreased in Kebbi (48% at baseline, 29% at endline), increased in Sokoto (42% at baseline, 49% at endline), and decreased in Zamfara (43% at baseline, 29% at endline) (Table 6.3.1b). Most babies delivered at home were bathed (78% in Kebbi, 79% in Sokoto, and 72% in Zamfara), and nearly all of these were bathed in the first 24 hours. Placing the newborn on the mother’s bare chest immediately after birth was uncommon, performed for only 30% of newborns in Kebbi, 4% of newborns in Sokoto, and 15% of newborns in Zamfara at endline.

Breastfeeding and nutrition: While more than 95% of children in all three states were breastfed at some point, initiation of breastfeeding within one hour of birth (e.g., immediate breastfeeding) was low at baseline and trended downward at endline: 42% at baseline and 39% at endline in Kebbi, 32% at baseline and 24% at endline in Sokoto, and 46% at baseline and 43% at endline in Zamfara. Further, exclusive breastfeeding among children under 6 months of age was lower in Kebbi (17%), Sokoto (11%) and Zamfara (37%) at endline, as compared with baseline values of 20%, 29% and 46% respectively. This low coverage is in large part due to high rates of giving non-breastmilk liquids to a child in the first three days after birth. This practice increased in all three states, from 77% to 80% in Kebbi, from 70% to 89% in Sokoto;

and from 50% to 66% at endline in Zamfara. Notably, while spousal opposition to breastfeeding declined in Sokoto (baseline: 25%, endline: 15%) and Zamfara (base-line: 30%, endline: 25%), it increased twofold in Kebbi (baseline: 37%, endline: 73%). Although there was high awareness of breastfeeding and its benefits, only about half of respondents in Kebbi (44%) and Sokoto (44%) and almost three in four women in Zamfara (70%) at endline felt confident that they could exclusively breastfeed their child for the first six months, a similar or modest increase from baseline (Kebbi: 39%, Sokoto: 44%, Zamfara: 65%)

Regarding complementary feeding, the majority of respondents reported feeding their child solid or semi-solid food three or more times a day (Kebbi 67%, Sokoto 53%, Zamfara 58%), with an increase observed at the endline (Kebbi 80%, Sokoto 63%, Zamfara 70%). However, the proportion of women feeding their children twice a day decreased (Kebbi 24% to 14%, Sokoto 29% to 23%, Zamfara 29% to 18%).

Grains were the most commonly given food group (Kebbi 85-88%, Sokoto 88-89%, Zamfara 84-86%). Dairy consumption remained low, with minimal improvement (Kebbi 48% to 34%, Sokoto 53% to 43%, Zamfara 57% to 58%). Consumption of beans, nuts, and Vitamin A-rich foods was relatively low across all states. To improve feeding practices, emphasis should be placed on nutri-ent-dense complementary foods. Specific food groups, such as dairy products, beans, nuts, and Vitamin A-rich foods, require attention. Tailored messaging through social and behavior change communication interventions should promote a diverse and culturally appropriate diet. Contrasting feeding frequencies and food group consumption between baseline and endline provide insights for future directions in interventions and areas of improvement.

6 BEH AV IOR A L SENTINEL SURV EILL A NCE SURV E Y IN NIGER I A: ENDLINE

Vaccination: In spite of small increases in vaccination rates from baseline to endline, vaccination coverage remained low in all three states. The percentage of children aged 12–23 months who had received all basic vaccinations (BCG, measles, DPT3, polio3) increased from 4% to 9% in Kebbi and from 8% to 15% in Zamfara but remained unchanged in Sokoto (4%). The most common reason for not receiving all vaccinations was spousal opposition, which increased from 30% to 54% in Kebbi, from 34% to 45% in Sokoto, and from 33% to 47% in Zamfara. Relatedly, an increased percentage of women across Kebbi (73% at baseline, 93% at endline), Sokoto (74% at baseline, 84% at endline) and Zamfara (73% at baseline, 82% at endline) said that the decision to vaccinate a child was influenced by their spouse.

Considering that husbands were the top barriers and also top influencers for childhood vaccination, this reinforces the importance of targeting spousal opposition and engaging men.

From the endline results, it is clear that promoting norms and awareness of the value of routine vaccinations should be a priority area for SBC efforts.