Scaling Up Interventions to Improve Infant and Young Child Feeding: The Role of Frontline Workers in Alive & Thrive Purnima Menon with Rahul Rawat, Kuntal Saha, Phuong Nguyen, Disha Ali, Andrew Kennedy, Adiba Khaled, Parul Tyagi, Lan Tran Mai, Roman Tesfaye & Marie Ruel International Food Policy Research Institute International Congress of Nutrition, Granada, Spain; Sept 18, 2013 Frontline workers and nutrition • Frontline workers – community health workers, community health volunteers, health staff in facilities – are where the rubber hits the road for public health and nutrition interventions. • Health systems literature is expanding on role of frontline workers for delivering life-saving interventions such as immunization • Less is known about how best to engage, motivate and deploy these frontline forces for nutrition behavior change – Challenges: sustained performance for non-tangible interventions, types of capacity strengthening investments needed, roles of incentives, monitoring and performance improvement in scaling up effective FLW contact for nutrition Mostly frontline workers! Elements of Alive & Thrive models, by country Some core elements but variability across country program models in platforms, and extent of emphasis on mass media See Food & Nutrition Bulletin Sept 2013 Supplement for more information! Frontline workers in Alive & Thrive interventions Bangladesh • Existing worker: Shashtya Sebika (frontline volunteer) • New worker: Pushtikormi (skilled nutrition worker) • CONTACT : OUTREACH TO FAMILIES THROUGH HOME VISITS (NGO Platform) Ethiopia • Existing worker: Health Extension Workers • Diverse frontline volunteers • CONTACT : OUTREACH TO FAMILIES THROUGH HOME VISITS, COMMUNITY GROUPS, AT HEALTH POSTS (Government Health System) Vietnam • Existing health staff at commune health centers • Village nutrition workers for demand-creation • CONTACT : FACILITY-BASED THROUGH SOCIAL FRANCHISE APPROACH LINKED TO GOVERNMENT HEALTH SYSTEM Implementation durations and exposures, by country Duration of implementation Bangladesh Exposures (in intervention areas) Communitybased interventions Mass media intervention Communitybased interventions Mass media intervention 3 years 2.5 years 69-98% 61-77% There is variability duration of implementation of Ethiopia 1.5across yearscountry 1program year models in35-73% 8-17% program components and household-level exposure to these components Vietnam 2 years 1.5 -2 years 45% 33-70% (spotspecific) Exposures are ranges capturing household exposure to any of the A&T-supported FLWs or mass media interventions. Exposure measures based on recall/aided recall. Source: Process evaluation surveys, 2013 Insights on A&T-linked frontline workers from baseline surveys • Strong knowledge of BF, but less on skills for EBF; poorer knowledge on complementary feeding, hygiene care, and feeding during illness • Regression analysis of predictors of FLW motivation highlighted the roles of knowledge, training, supportive supervision Bangladesh (SS) Ethiopia (HEW) Vietnam (CHC staff) Motivating factors: Motivating factors: • • Motivating factors: • • • Positive, supportive supervision (high) IYCF knowledge Refresher training with 1-3 months Job duration equal or more 24 months • • Positive, supportive supervision (high) Education (technical/vocational) Supervision visits on specified topics • • Positive, supportive supervision (high) Participated in training within 12 months A&T core interventions in all three countries aim to strengthen these motivational factors BANGLADESH: ENGAGING FLWS FOR DELIVERING INTERVENTIONS THROUGH A LARGE-SCALE NGO PLATFORM IMPLEMENTED BY BRAC BANGLADESH IMPACT EVALUATION DESIGN At scale implementation in 40+ subdistricts 60 rural subdistricts 20 (paired) rural subdistricts Randomized 10 subdistricts A&T-intensive Intensive IYCF counseling by BRAC frontline workers + mass media 10 subdistricts A&T non-intensive Standard care by BRAC frontline workers + mass media only DATA COLLECTION Baseline survey (April-July 2010) & early process evaluation (late 2010) Process evaluation survey on implementation (September-October 2011) & qualitative research Process evaluation survey of implementation and utilization (subsample only, June-July 2012) & qualitative research Process evaluation survey on implementation and utilization (all areas, April-July 2013) Endline survey (April-July 2014) Bangladesh: Early Impacts on IYCF Practices (2013) 100 90 2010 A&T Intensive 2010 A&T Non-Intensive 2013 A&T Intensive 2013 A&T Non-Intensive 18.7 pp *** 24.2 pp *** 12.3 pp (n.s) 80 70 26.6 *** 20.3*** 7.6 (n.s.) Percent 60 24.6*** 50 40 30 20 10 0 Early initiation Exclusive Continued BF at Intro of CF at 6- Minimum Minimum Meal Minimum Consumption of Breastfeeding 1 year 8 mo Dietary Freq Acceptable Diet of Iron-Rich breastfeeding (<6 mo) Diversity (6-23 Foods (6-23 mo) mo) *** p<0.01; ** p<0.05; *p<0.1 † Double difference estimates with clustered standard errors comparing A&T intensive and non-intensive areas in 2010 and 2013 Neither No media non-A&T FLW Contact with A&T FLW Media alone, no FLW Media + untrained FLW Contact with A&T FLW & media 100 90 80 70 60 % 50 40 30 20 10 0 Baseline Bangladesh: IYCF indicators, by intervention exposure (based on aided recall; unadjusted preliminary estimates) 2013 EBF Min. Diet Diversity Min acceptable diet Baseline (all) Seen TVC; Seen A&T SS Seen TVC; See non-A&T SS Seen TVC; Not seen any SS Not seen TVC; Seen A&T SS Not seen TVC; Seen non-A&T SS Not seen TVC; Not seen any SS Health Extension Worker Health Volunteer ETHIOPIA: BUILDING FRONTLINE WORKER CAPACITY FOR IYCF IN ETHIOPIA’S HEALTH EXTENSION SYSTEM ETHIOPIA IMPACT EVALUATION DESIGN 89 IFHP woredas in 2 regions (Tigray & SNNPR) Random selection of 75 enumeration areas from 56 woredas for evaluation surveys* DATA COLLECTION Cross-sectional baseline survey in 2010 Process evaluation (qualitative research) on implementation in 8 woredas (2012) Process evaluation survey on implementation and utilization (2013) Cross-sectional endline survey for impact assessment in 2014 *The survey covered 75 enumeration areas in 19 woredas from Tigray and 37 woredas from SNNPR Shifts in IYCF practices between 2010-13, in Tigray & SNNPR (combined), Ethiopia Baseline, 2010 Process Evaluation, 2013 100 90 80 70 60 % 50 40 30 20 10 0 Early initiation of BF EBF Continued Intro of CF at Minimum BF at 1 year 6-8 mo DD Minimum meal frequency Minimum acceptable diet Iron rich foods Ethiopia: IYCF practices in 2013, by exposure to health extension workers and radio spot (Tigray region only) 100 90 60 % 50 40 20 10 0 Baseline 30 Contact with A&T FLW 70 Contact with A&T FLW & radio 80 2013 Early initiation of BF Exclusive BF Minimum diet diversity Iron-rich foods Baseline (all) Heard radio spot; Seen HEW Heard radio spot; Not seen HEW Not Heard radio spot; Seen HEW Not Heard radio spot; Not seen HEW Ethiopia: IYCF practices in 2013, by exposure to frontline volunteers and radio spot (Tigray region only) 100 70 % 60 50 40 20 10 0 Baseline 30 Contact with A&T FLW 80 Contact with A&T FLW & radio 90 2013 Early initiation of BF Exclusive BF Minimum diet diversity Iron-rich foods Baseline (all) Heard radio spot; Seen volunteer Heard radio spot; Not seen volunteer Not Heard radio spot; Seen volunteer Not Heard radio spot; Not seen volunteer VIETNAM: A SOCIAL FRANCHISE MODEL FOR DELIVERING IYCF COUNSELING AT GOVERNMENT HEALTH FACILITIES VIETNAM IMPACT EVALUATION DESIGN 40 Commune Health Centers (CHCs) from 4 provinces Randomization DATA COLLECTION 20 Comparison CHCs Standard Government Service+ mass media Full implementation in 11 nonevaluation provinces (660 franchises) 20 Intervention CHCs IYCF social franchise + Standard Government Service + mass media Cross-sectional baseline survey in 2010 Process evaluation on implementation (2012) Process evaluation on implementation and utilization (2013) Cross-sectional impact survey in 2014 Impact on IYCF practices in Vietnam – 2010 vs 2013 2010 A&T franchise 2010 non-franchise 2013 A&T franchise 2013 non-franchise 100 90 80 70 % 21.0 pp ** 60 50 40 Complementary feeding practices better at baseline: lower potential to benefit 30 20 10 0 EIBF EBF Continued BF Introduction Minimum Minimum of CF diet diversity meal frequency Minimum acceptable diet *** p<0.01; ** p<0.05; *p<0.1 † Double difference estimates with clustered standard errors comparing A&T intensive and non-intensive areas in 2010 and 2013 Iron rich foods Vietnam: Breastfeeding, by exposures to media spots and the social franchise 100 80 Neither Use of franchise 20 Baseline 40 Media only % Franchise & media 60 2013 0 Early initiation of BF Baseline Not seen TVC; visit MTBT Exclusive BF Seen TVC; visit MTBT Not seen TVC; Not visit MTBT Seen TVC; Not visit MTBT Conclusions on early impact Despite variability in the models, durations of implementation and exposures, we find: • In Bangladesh: large, and significant, impacts for several indicators of IYCF • In Vietnam: Large, and significant, impacts for exclusive breastfeeding • In Ethiopia: Improvements in most IYCF practices • Impact linked to potential to benefit In all three countries, contact with A&Tsupported frontline workers appears to be linked with improved practices; media interventions are playing a supportive & synergistic role 2.5 month old exclusively breastfed baby in Bangladesh, 2013 (Photo: Purnima Menon) Acknowledgments • Alive & Thrive leadership at HQ and at the country level • BRAC, Save the Children • Country research and data-collection collaborators: DATA, Bangladesh; Institute for Social and Medical Studies, Vietnam; Addis Continental Institute for Public Health, Ethiopia • Dozens of enumerators and field researchers • Mothers, fathers, grandmothers and program implementers • Bill & Melinda Gates Foundation for funding to Alive & Thrive & Ellen Piwoz for her support More information on Alive & Thrive programs, implementation lessons and evaluation designs: Food & Nutrition Bulletin Special Supplement STAY TUNED – more to come on full impact, process evaluation results, costs, policy wins, ethnographic insights, and more!