Search

This paper uses programme theory to assess, in the context of an effectiveness evaluation, the feasibility and acceptability of distributing micronutrient Sprinkles through a food-assisted maternal and child health and nutrition programme in rural Haiti. We laid out the steps related to programme delivery and household utilization of Sprinkles and used qualitative and quantitative methods to gather data on these steps. Methods included structured observations, checks of beneficiary ration cards, exit interviews, focus group discussions (FGD), individual interviews and survey data from the effectiveness evaluation. Results are as follows: (1) information on use of Sprinkles was provided before mothers first received them, as planned; (2) Sprinkles were re-packaged and distributed as planned and in the appropriate amount; (3) almost all mothers (96%) received two monthly rations of Sprinkles and received timely information on their use; (4) mothers understood instructions about use of Sprinkles and acceptance was high, and no selling of the product was reported or observed; and (5) mothers reported using Sprinkles as instructed, every day (63% in survey; 86% at exit interviews), and for the child only (99%). FGD with staff highlighted the acceptance of the intervention, with a reported ‘modest’ increase in workload. Within this well-established programme, it proved feasible to distribute Sprinkles and to ensure appropriate use by beneficiary mothers. Existing programme venues were suitable for distributing Sprinkles and educating mothers about their use. Use of programme theory helped to assess feasibility and acceptability of the Sprinkles intervention and provided useful information for programme replication or scale-up. [ABSTRACT FROM AUTHOR]

We evaluated the effectiveness of a 2-mo treatment of Sprinkles containing 12.5 mg iron, 5 mg zinc, 400 microg vitamin A, 160 microg folic acid, and 30 mg vitamin C in reducing anemia among children 9- to 24 mo old in Haiti. Ten food distribution points (FDP) where children received take-home rations of fortified wheat-soy blend (WSB) were randomly allocated into 2 groups: 1) Sprinkles-WSB (S-WSB) (6 FDP; n = 254), receiving 30 sachets of Sprinkles monthly for 2 mo; and 2) WSB only (WSB) (4 FDP; n = 161), not receiving Sprinkles. At baseline, anemia prevalence [hemoglobin (Hb) < 100 g/L], adjusted for age and sex, was 54 and 39% in S-WSB and WSB groups, respectively. After the 2-mo intervention (1st follow-up), anemia, adjusted for baseline prevalence, age, and sex dropped to 24% in S-WSB (P < 0.001) and increased to 43% in WSB (P = 0.07). At 7 mo postintervention, anemia in S-WSB declined to 14%; 92% of children who were nonanemic at 1st follow-up remained so without further Sprinkles consumption. From baseline to 1st follow-up, mean Hb increased by 5.5 g/L and dropped by 1.0 g/L in the S-WSB and WSB groups, respectively (P < 0.001). From baseline to 2nd follow-up, mean Hb increased by 10.9 g/L in S-WSB (P < 0.001). Changes in mean Hb were greater for younger children (<21 mo at onset of intervention) (P < 0.05) and for children who were anemic at baseline (P < 0.001). In populations with a high prevalence of anemia, such as rural Haiti, 2 mo of Sprinkles are effective in reducing anemia among 9- to 24-mo-old children. [ABSTRACT FROM AUTHOR]

We compared child feeding practices (CFP) in rural Haitian communities exposed to two models for delivering food aid & behavior change communication (BCC) within a maternal & child health & nutrition program: 1) a preventive model (P), targeting all children 6-24 mo; or 2) a recuperative model (R), targeting malnourished children under 5 y. Twenty paired clusters of communities were randomized to the models. BCC, designed using formative research, was delivered to small peer groups & was identical in P & R from pregnancy to early lactation (to 6 mo). Thereafter, the focus in P was on age-specific messages for optimal CFP & in R on needs of the malnourished child. Baseline & endline surveys (3 yr apart; N∼1500) assessed maternal recall of CFP. No baseline differences were seen. At endline, P area children had greater dietary diversity & were more likely to have received an extra meal after diarrhea (p<0.05); most other differences favored P but were small. Larger differences were found btn baseline & endline for both models, & btn participants & non-participants. Exclusive breastfeeding increased from 42% at baseline to 92%; non-participants were 4 times more likely to report introducing liquids or semi-solids at < 6 mo (p<0.05). Results suggest that a well-designed BCC program improved many important CFP even as Haiti's food security deteriorate. [ABSTRACT FROM AUTHOR]

We compared the impact of 2 models of targeting food aid & BCC on knowledge, trial & adoption of child feeding practices (CFP) in rural Haiti. Models targeted to mothers of all children 6-24 mo (preventive) or mothers of malnourished children <60 mo (recuperative) were randomized within 10 pairs of matched communities. BCC, designed based on formative research, was delivered via small peer group clubs. Baseline & endline surveys 3 y apart provided data on health & nutrition knowledge (N∼1500); data on awareness, trial & adoption of 7 key CFP were collected at endline only. No differences were seen in baseline knowledge. Endline knowledge was greater in the preventive group (p<0.05) & among exposed mothers (p<0.05). Awareness of 6 key CFP was higher (p<0.05) among preventive participants than recuperative participants (74-98% vs. 68-98%). In both groups, trial was lower than awareness (24-84%) & adoption lower than trial (15-57%). Conclusions: (1) A preventive approach leads to increased CFP knowledge, trial & adoption; (2) A BCC program designed based on formative research & implemented via small peer group clubs is effective at improving CFP knowledge, trial & adoption; (3) Constraints other than knowledge affect ability to try & adopt new behaviors. [ABSTRACT FROM AUTHOR]

We compared the impact on household food insecurity (HHFI) of targeting a food assistance and behavior change communication (BCC) intervention for up to 18 mo to all households with children 6-24 mo (preventive model) vs. up to 9 mo to those with a malnourished under-5 child (recuperative model) in rural Haiti. The models were randomly assigned within 10 pairs of matched communities. HHFI was assessed using an experience-based food insecurity scale (E-FI) at baseline & endline surveys (3 y apart; N∼1500); and mo of food insufficiency (MFI) in the past year at endline only. There were no differences between program models in baseline HHFI, which was severe. Endline HHFI remained high in both groups, but E-FI and M-FI scores were lower in the preventive group (p<0.05). Current participants in both programs had lower HHFI than non-participants (p<0.01). Current participation rates were double in preventive areas (37% vs. 14%, p<0.05) due to targeting (age vs. malnutrition), duration of intervention (18 vs. 9 too) and greater overall participation in preventive communities. Results suggest that targeting food aid using a preventive approach leads to greater reductions in HHFI than a recuperative approach, mainly via differences in participation patterns. [ABSTRACT FROM AUTHOR]

We compared differences in childhood undernutrition in rural Haitian communities exposed to a food aid/behavior change communication (BCC) intervention linked with a maternal and child health/nutrition (MCHN) program and delivered through either: 1) a preventive model, targeting all children 6-24 mo; or 2) a recuperative model, targeting malnourished children under 5 y. Clusters of communities (n=20) were paired on access to services and other factors and randomly assigned to each model. Two cross-sectional surveys were conducted (baseline and 3 years later) to test differences in undernutrition in 12-42 mo old children (N∼1500 at each survey). There were no differences at baseline. At follow-up, stunting, underweight and wasting were 5, 7, and 4 percentage points lower respectively in preventive (23.2%, 22.9% and 5.2%) compared to recuperative areas (28.6% 30.1%, 9.2%) (p<0.05); mean z-scores for all 3 indicators were significantly higher in preventive communities. Differences in favor of preventive areas were statistically significant across the age range (except for 18-24 and 36-42 mo). We conclude that to reduce childhood undernutrition a preventive approach to targeting food aid and BCC in the context of an MCHN program is more effective than the traditional recuperative approach. [ABSTRACT FROM AUTHOR]

The article presents a cluster randomized trial to determine if earlier age-based preventive targeting of childhood nutrition programs could reduce childhood malnutrition in developing countries. Most such programs have used a benchmark of five years or younger, and the trial attempts to determine if a lower benchmark would be more effective, and if preventive targeting of nursing mothers is more effective than remedial nutrition. Two childhood nutrition programs in Haiti were examined in the trial. The results showed that preventive targeting was more effective in reducing cases of child malnutrition.