A Self-Esteem Scale showed that the intervention group had a higher overall score (0.73) compared with the comparison group on the waiting list to receive cash transfers (0.68).

7.55% of children in the intervention group had an abnormal Strength and Difficulties Questionnaire (behavioural health screening tool) score compared with 9.18% in the comparison group.

Household member affected by illness or died

Expenditure on medicine

Affected by illness: −0.01 (0.08) cash, 0.59 (0.50) comparison using voucher (P-value 0.87)

Death: 0.03 (0.05) cash, 0.11 (0.31) comparison using voucher (p value 0.57)

Households receiving cash transfers were more likely to use the funds to pay for health expenses

Presence of acute malnutrition (WHZ <2 or MUAC <125 mm)

MUAC

WHZ

Child dietary diversity

Child meal frequency

Child illness

Maternal mental health

Factors found to be associated with risk of acute malnutrition in households receiving cash transfers included low WHZ, household poverty status, and occurrence of child illness.

Household food expenditures and other diet-related factors were not found to be associated with the risk of acute malnutrition.

Over the course of the study, 18% (n=74) children in the cash transfer programme became acutely malnourished.

Weight gain

Weight gain velocity (g/kg/d)

MUAC

WHZ

Presence of acute malnutrition (WHZ <2 or MUAC <125 mm or bilateral pitting oedema)

Meal frequency

Dietary diversity

Weight Gain: difference estimations found that cash transfers were associated with a 1.27 kg greater overall weight gain (p<0.001) compared with the control group that did not receive cash.

Weight-for-height Z scores: 1.82 greater overall gain in WHZ (p value<0.001) in the cash group.

Acute malnutrition: odds of acute malnutrition at the end of the intervention was 25 times higher among children in the comparison group compared those in the intervention group receiving cash (p value<0.001).

Meal frequency: compared to the comparison group, increased by one meal more on average among cash group (p value<0.001).

Dietary diversity: children in cash group consumed more food groups (p value<0.001).

Household Hunger Scale (HHS)

Minimum Dietary Diversity for Women (MDDW)

MUAC (MUAC <21.0 cm classified as acute malnutrition)

Meal frequency

Hunger: in the mixed transfer group fewer households reported moderate or severe hunger (35.4% compared with 44.0% and 94.9% in voucher and non-assistance groups)

Meal frequency: there was a significant increase in meal frequency of cash transfer recipients (0.3 meals/day, CI: 0.1 to 0.5, p value 0.001) compared with voucher recipients.

MUAC: an increase in mean MUAC was statistically significant in both voucher (0.9 cm, CI: 0.6 to 1.3, p=0.001) and mixed transfer recipients (1.3 cm, CI: 1.1 to 1.5, p value 0.001). Yet, mean MUAC increased by 0.4 cm (95% CI: −0.1 to 0.8, p value 0.086) more in mixed transfer compared with voucher.

MDDW: adjusted model reported no significant difference between mixed transfer and voucher (7.7%, 95% CI: −7.0 to 22.5, p value 0.3)

Household Hunger Scale (HHS)

Children’s dietary diversity

Meal frequency

Minimum acceptable diet (MAD)

Mean MUAC

Acute malnutrition (MUAC <12.5 cm)

Hunger: 56.0% of the food voucher group and 64.6% of the mixed transfer group were classified as having little to no hunger at baseline

Dietary diversity: 64.4% of children in the mixed transfer group compared with 51.0% of children in the food voucher group achieved the minimum dietary diversity

MUAC: the mean MUAC values increased significantly by 0.4 cm (CI: 0.1 to 0.8, p=0.001) in the food voucher group, but there was no statistically significant change in the mixed transfer group (−0.1, CI: −0.4 to 0.2, p=0.001).

Acute malnutrition: while it was not statistically significant, the prevalence of acute malnutrition was 2.9% (CI: 0.4 to 6.1, p=0.086) greater in the food voucher group as compared with the mixed transfer group.

Child delivery in a health facility

Receiving at least one ANC visit

Receiving one PNC visit

Receiving at least one CHW home visit

ANC visits: there was a statistically significant increase in ANC visits in the intervention villages (AMD 45.0%, CI: 0.18 to .0.72, p value 0.004).

Facility delivery: facility deliveries (AMD 3.3%, CI: −0.14 to 0.21, p=0.685) and PNC visits (AMD 31.8%, CI: −0.05 to 0.68, p=0.080) also increased, but these findings were not statistically significant.

CHW Home visits: CHW home visits did not change significantly in the intervention villages, but did decrease by 23.9% in control villages (AMD 12.2%, CI: −0.27 to 0.51, p=0.508).

Food insecurity via (HFIAS)

Depressive symptoms via the PHQ-9

Disability status, via an adapted version of the WG-SS

Women’s experiences of violence

, Between baseline and endline, Food insecurity items decreased by 0.92 points (95% CI: −1.17 to −0.68; p<0.0001) in the unadjusted linear model

This significant decrease in food insecurity remained robust when adjusting for demographics (β=−0.90; 95% CI: −1.14 to −0.65; p<0.0001)

Or when includinghousehold fixed effects (β=−0.95; 95% CI: −1.19 to −0.71; p<0.0001)

Women agreed on average with 12.08 statements at baseline (SD: 3.32) on the 20 item HESPER scale. This signifieshigh household daily stressors and perceived serious needs. This was similar at endline (Mean: 12.11; SD: 3.87).

Unadjusted, adjusted and household fixed effects models were not statistically significant (β=0.04; p=0.83; β=0.05; p=0.81; β=0.12; p=0.52, respectively).

Women had a mean of 11.08 on the PHQ-9 scale on average at baseline and 11.93 at endline.

Between baseline and endline, depressive symptoms changed by 0.86 points (95% CI: 0.32 to 1.40; p=0.002) in the unadjusted model. This was similar in the adjusted model (β=0.92; 95% CI: 0.35 to 1.49; p=0.001) as well as in the household fixed effects model (β=0.89; 95% CI: 0.34 to 1.43; p=0.001).

At the individual and household levels, many respondents reported improvements in health.

There were decreases in depression severity in both the treatment and control groups over time. At endline, the treatment group mean decreased by 29%, from 0.85 to 0.60. Likewise, the control group mean decreased by 21%, from 0.75 to 0.59

The average treatment effect on symptoms of depression was not statistically significant and small

Mean child, household and women DDS values

Incidence acute malnutrition, defined by low MUAC or oedema

Prevalence acute malnutrition, defined by low WHZ or oedema

Mean WHZ value

Mean FCS

Mean HFIAS score

Mean Reduced Coping Strategies Index (rCSI) score

Increased Child Dietary Diversity score by 0.53 (95% CI 0.01 to 1.05)

Increased monthly household expenditure by US$29.60 (95% CI 3.51 to 55.68)

Increased household Food Consumption Score by 14.8 (95% CI 4,83 to 24.8)

Decreased Reduced Coping Strategies Index by 11.6 (95% CI 17.5 to 5.96)

Did not reduce risk of acute childhood malnutrition

Psychosocial distress

Health expenses

No significant difference in change in dietary quality observed between food voucher and mixed transfer recipients

A significant difference in change in mean meal frequency was observed (0.3 meals/day, CI: 0.1 to 0.5, p=0.001).

Mean MUAC increased significantly among both voucher (0.9 cm, CI: 0.6 to 1.3, p=0.001) and mixed transfer recipients (1.3 cm, CI: 1.1 to 1.5, p=0.001)

Fewer households in the mixed transfer group had moderate or severe hunger (35.4% compared with 44.0% and 94.9% in voucher and non-assistance groups, respectively)

After the flood, households not receiving FbF assistance felt miserable or unhappy significantly more frequently compared to the intervention group not receiving cash assistance

In the last seven days before the survey, compared to the intervention group, FbF-assisted households were significantly less likely to have felt anxious or depressed.

A third of participants reported the cash transfer improved their mental well-being.

The cash transfers also alleviated stress or anxiety related to paying rent.

The cash transfer helped to reduce small health expenditures by enabling recipients to partially cover the costs of treatment or medication. For other beneficiaries it helped to secure a loan covering healthcare expenses. Whilst the cash transfers alleviated some financial burdens of accessing healthcare, this was not a decisive factor in recipients' behaviours related to accessing health treatment.

DDS

HDDS

FCS

Caloric intake per capita (daily)

All three groups (cash, food vouchers, food transfers) experienced significant improvements in households’ caloric intake and dietary diversity, however caloric intake increased by 21% in the food group and only by 12% in the cash group (p=0.05).

The FCS, which measures households’ food consumption, increased by 11% in the cash group, 12% in the food group, and 16% in the voucher group. However, only the voucher and food groups saw statistically significant reductions in the percentage of households with poor to borderline FCS

Total calorie consumption

Diversity of diet including vegetables, fruits, and animal products

When drought affects income, households tend to buy cheaper calories (such as grains), which results in a net increase in total calories consumed, but these calories are more likely to cause chronic diseases. The CCT (PROGRESA) mitigates the negative effects of drought on calorie availability from fruits, vegetables and other sources. The CCT does not mitigate the impact of drought on calories from grains.

HDDS

CDDS

Height-for-age z score (HAZ)

Positive significant (large) impact on dietary diversity across full sample, strongest in poorest HH

Overall the intervention increased the CDDS by 0.61 food groups across all HH

Average estimated programme impact on HAZ across all HH was not significant

HAZ of HH in lowest tercile statistically significant and large impact of 0.31 SD

Health service utilisation

Diabetes medication adherence

Blood glucose self-monitoring

BMI

HbA1C

Blood pressure

Health expenditure

Regular diabetes care visits increased in the CHV + CCT group (15.1%, CI: 5.4,24.8%; p= 0.002)

Specialist visits increased among CHV +CCT group (16.8%, CI: 6.6 to 27.0%; p= 0.001)

Specialist visits decreased in the CHV only participants (− 27.8%, CI: − 41.5,% to 14.0%; p < 0.001) (group difference in change p < 0.001)

Pharmacist consultation decreased significantly in CHV only (− 24.1%, CI: − 37.9% to 10.4%; p = 0.001) and CHV +CCT (− 12.7%, CI: − 22.2% to 3.2%; p= 0.009)

Decreased hospital visits among CHV only (−11.5%, CI: − 22.9% to 0.1%; p= 0.049)

Increase in adherence in the CHV + CCT group (6.8%, CI: 2.2 to 11.5%; p= 0.004)

Decrease in self-monitoring CHV only participants (− 16.3%, CI: − 25.2% to 7.4%; p≤ 0.001)

Decrease in BMI in the CHV + CCT group (− 1.0 kg/m2, CI: − 1.7 to –0.3; p= 0.005)

Decrease in HbA1C in CHV only 0.7% (CI: − 1.1% to 0.4%; p < 0.001), CHV +CCT − 0.5% (CI: − 0.7% to 0.3%; p < 0.001) and MPC group −0.2% (CI: −0.5 to 0.0%; p= 0.028)

Increase in CHV+CCT group of normal blood pressure 11.3% (CI: 3.2 to 19.4%; p= 0.007)

Health service utilisation (care-seeking for children and chronic or acute illness among adults)

Access to medication

(Health expenditure)

For childhood illnesses: increase was 19.3% (CI: 7.3,31.20%; p = 0.002) greater among MPC recipients compared to controls

For adult acute illnesses: increased among MPC recipients but decreased in controls (adjusted diffference-in-difference (DiD) 11.3%; p = 0.057)

Childhood Hospitalisations: significantly smaller among MPC recipients than among controls (adjusted difference − 6.1%, CI: − 11.7% to 0.4%; p = 0.037; effect size: −133.5%)

No significant changes observed within or between groups.

DDS

Calories per adult equivalent

HFIAS

Subjective well-being

Refugees who received the transfer were found to have more diverse diets (20% higher DDS), higher caloric intake (p=0.12), and be less food secure (92% vs 79%) than those arriving just before. There was suggestive evidence that refugees living in Kalobeyei felt happier and more independent from aid than their counterparts in Kakuma. These results are robust to various tests and specification changes. kitchen-garden agriculture improves refugee diets

Acute illness

Diarrhoea

Respiratory infection

Required primary healthcare

Used primary healthcare

Lower likelihood of children 0–5 years reporting acute illnesses with MPCs

Lower incidence of diarrhoea and respiratory infections in children 0–5 years with MPCs

Lower likelihood of needing PHC with MPCs

More likely to use PHC when needed with MPCs

Short-run improvement not sustainable when MPC benefits are discontinued, except for respiratory infections which don’t change

Second cycle of cash transfer results in initial improvements of acute illnesses; needing PHC and using PHC maintain in the longer term

HDSS

FCS

Value of food consumed

Caloric intake

Cash beneficiaries had a more diverse diet, fed infants and young children a wider variety of foods and consumed higher quality food.

Cash beneficiaries also consumed approximately 150 less calories per day than food recipients. Self-reported measures of food insecurity incidents and non-food expenditures, including qat use, did not differ by transfer type.

standard intervention = ~£144 over 4-month period in 4 monthly disbursements

modified intervention = ~ £144 over 6-month period in 6 monthly disbursements

Acute malnutrition

MUAC

WHZ

Dietary diversity

There was no observable difference in the nutritional impact among children in the modified and standard cash transfer interventions. The odds of children having GAM and the adjusted mean WHZ were the same in each intervention arm and the general population.

In children under 5, the GAM was 13.5% (95% CI: 10.8 to 16.8) at baseline and 14.7% (95% CI: 12.9 to 16.9, p=0.161) at endline. There was no significant difference in either the standard intervention (p=0.426) or the modified intervention (p=0.231).

Healthcare utilisation

For beneficiary households a positive impact of cash transfers on the use of healthcare services (66.6% higher in treatment group)