Study characteristics |
Methods |
Study design: CBA Study grouping: N/A How were missing data handled? NR. Randomisation ratio: N/A Recruitment method: state governments identify families eligible for PDS support and issue ration cards to such HHs. Sample size justification and outcome used: NR Sampling method: the National Sample Survey Organization (NSSO) conducts a survey of 100,000–125,000 HHs representative of the state and national level. Study aim or objective: to examine the impact of subsidies on pulses in select Indian states on HH consumption and protein intake. Study period: 2007 (for Himachal Pradesh, Punjab and Tamil Nadu) or 2008 (Andhra Pradesh) to 2010 Unit of allocation or exposure: states |
Participants |
Baseline characteristics Intervention or exposure
Age: HH head, mean: 44.30 (SE 0.146)
Place of residence: proportion of HHs with LPG, mean: 0.286 (SE 0.004); proportion of HHs with electricity, mean: 0.889 (SE 0.003)
Sex: proportion of HHs with male head, mean: 0.867 (SE 0.003)
Ethnicity and language: proportion of HHs, mean: Hindu 0.848 (SE 0.003); Islam 0.048 (SE 0.001); Christian 0.030 (SE 0.001)
Occupation: proportion of HHs, mean: self‐employed in non‐agriculture 0.108 (SE 0.002); employed in agricultural labour 0.239 (SE 0.004); employed in casual labour in rural India 0.085 (SE 0.002); self‐employed in agriculture 0.159 (SE 0.003); self‐employed in urban areas 0.110 (SE 0.002); employed in labour in urban areas 0.040 (SE 0.001)
Education: education level of HH head, mean: 1.780 (SE 0.017)
SES: proportion of HHs which, mean: purchased subsidised rice from PDS: 0.521 (SE 0.004); purchased subsidised wheat from PDS: 0.047 (SE 0.001); purchased subsidised sugar from PDS: 0.414 (SE 0.004); purchased subsidised kerosene from PDS: 0.509 (SE 0.004). Land ownership quintiles, mean: 2.542 (SE 0.014)
Social capital: proportion of HHs, mean: scheduled tribes 0.041 (SE 0.002); scheduled castes 0.222 (SE 0.004); other backward classes 0.517 (SE 0.004)
Nutritional status: quantity of pulses, mean, kg/HH/day: 3.210 (SE 0.020); protein from pulses, mean: g/HH/day: 23.24 (SE 0.148); protein from all food, mean: g/HH/day: 191.6 (SE 1.037). Total number of meals, mean: consumed in the HH per day: 11.24 (SE 0.053); consumed on payment per HH: 2.604 (SE 0.174); consumed at home per HH per month: 321.5 (SE 1.623)
Morbidities: NR
Concomitant or previous care: NR
Control
Age: HH head, mean: 44.70 (SE 0.070)
Place of residence: proportion of HHs with LPG, mean: 0.203 (SE 0.002); proportion of HHs with electricity, mean: 0.596 (SE 0.002)
Sex: proportion of HHs with male head, mean: 0.897 (SE 0.001)
Ethnicity and language: proportion of HHs, mean: Hindu 0.830 (SE 0.001); Islam 0.127 (SE 0.001); Christian 0.022 (SE 0.000)
Occupation: proportion of HHs, mean: self‐employed in non‐agriculture 0.114 (SE 0.001); employed in agricultural labour 0.183 (SE 0.002); employed in casual labour in rural India 0.075 (SE 0.001); self‐employed in agriculture 0.281 (SE 0.002); self‐employed in urban areas 0.102 (SE 0.001); employed in labour in urban areas 0.029 (SE 0.000)
Education: level of HH head, mean: 1.883 (SE 0.009)
SES: proportion of HHs which, mean: purchased subsidised rice from PDS: 0.139 (SE 0.001); purchased subsidised wheat from PDS: 0.107 (SE 0.001); purchased subsidised sugar from PDS: 0.083 (SE 0.001); purchased subsidised kerosene from PDS: 0.644 (SE 0.002). Land ownership quintiles, mean: 2.595 (SE 0.006)
Social capital: proportion of HHs, mean: scheduled tribes 0.098 (SE 0.001); scheduled castes 0.191 (SE 0.002); other backward classes 0.374 (SE 0.002)
Nutritional status: quantity of pulses, mean, kg/HH/day: 3.550 (SE 0.016); protein from pulses, mean, g/HH/day: 25.98 (SE 0.123); protein from all food, mean, g/HH/day: 270.5 (SE 0.933). Total number of meals, mean: consumed in the HH per day: 11.61 (SE 0.031); consumed on payment per HH: 1.354 (SE 0.042); consumed at home per HH per month: 336.5 (SE 0.921)
Morbidities: NR
Concomitant or previous care: NR
Overall: NR Inclusion criteria: HHs qualifying for PDS subsidies in selected Indian states. Details of eligibility criteria NR. Exclusion criteria: NR Pretreatment: none Attrition per relevant group: total of 4872/23,558 (20.7%) fewer HHs in intervention and 18,917/101,086 (18.7%) fewer HHs in control states were included in the follow‐up survey. Description of subgroups measured and reported: poorest HHs, vegetarian HHs and HHs with different quantities of pulse consumption. Total number completed and analysed per relevant group: 18,686 HHs in intervention and 82,169 HHs in control states. Total number enrolled per relevant group: 23,558 HHs in intervention states and 101,086 HHs in control states. Total number randomised per relevant group: N/A |
Interventions |
Intervention characteristics Intervention or exposure
Food access intervention category: address food prices
Intervention type: food subsidies
Description: subsidising of a variety of pulses (arhar daal in Andhra Pradesh; moong, urad daal and chana daal in Himachal Pradesh; various pulses in Punjab; arhar daal and urad daal in Tamil Nadu) as part of the PDS, in addition to the usual subsidising of rice, wheat, sugar and kerosene oil.
Duration of intervention period: 2007 (Himachal Pradesh, Punjab and Tamil Nadu) or 2008 (Andhra Pradesh) to present
Frequency: monthly subsidy of pulses: 1 kg of 1 pulse in Andhra Pradesh; 1 kg of 3 pulses in Himachal Pradesh; 0.5 kg of various pulses in Punjab and 1 kg of 2 pulses in Tamil Nadu.
Number of study contacts: baseline (2004 and 2005) and follow‐up (2009 and 2010).
Providers: state governments (pulses) and central Indian government (rice, wheat, sugar and kerosene).
Delivery: government‐issued ration cards are given to poor HHs enabling them to purchase from fair‐price shops 1 kg arhar daal for INR 50 in Andhra Pradesh; 1 kg moong for INR 49.99, 1 kg urad daal for INR 34.99 and chana daal for INR 25 in Himachal Pradesh; 0.5–2.5 kg per family of various pulses for INR 20/kg in Punjab; and 1 kg arhar daal for INR 30 as well as 1 kg urad daal for INR 30 in Tamil Nadu.
Co‐interventions: rural poor HHs may have had access to other welfare programmes, e.g. Mahatma Gandhi National Rural Employment Guarantee Scheme, implemented between 2004/2005 and 2007/2008 or 2009/2010.
Resource requirements: NR
Economic indicators: NR
Control
Food access intervention category: address food prices
Intervention type: food subsidies
Description: usual subsidy of rice, wheat, sugar and kerosene oil as part of the PDS.
Duration of intervention period: 1947 to present
Frequency: NR
Number of study contacts: baseline (2004 and 2005) and follow‐up (2009 and 2010)
Providers: Central Indian government
Delivery: Central government responsible for procurement, storage, transportation and the bulk allocation of food grains to the state governments. State governments responsible for identification of eligible families, issue of ration cards, distribution of subsidised goods and supervision of the programme. Fair‐price shops spread throughout the country.
Co‐interventions: rural poor HHs may have had access to other welfare programmes, e.g. Mahatma Gandhi National Rural Employment Guarantee Scheme, implemented between 2004/2005 and 2007/2008 or 2009/2010
Resource requirements: NR
Economic indicators: NR
|
Outcomes |
HH expenditure on food (excluding pulses) per month Dietary diversity: amount of different types of food consumed by HH per month Dietary intake: amount of protein from pulses/all foods consumed per HH per day |
Identification |
Sponsorship source: Indian Council of Agricultural Research (ICAR); Swiss Agency for Development and Cooperation. This work was also undertaken as part of the Consultative Group on International Agricultural Research (CGIAR) Research Program on Agriculture for Nutrition and Health (A4NH). Country: India Setting: rural and urban HHs in selected states (Andra Pradesh; Maharashtra) Authors' names: Suman Chakrabarti; D Roy Email: d.roy@cgiar.org Declarations of interest: NR Study or programme name and acronym: Public Distribution System (PDS) Type of record: journal article |
Notes |
|
Risk of bias |
Bias |
Authors' judgement |
Support for judgement |
Random sequence generation (Selection bias) |
High risk |
No randomisation. CBA. |
Allocation concealment (Selection bias) |
High risk |
No randomisation. CBA. |
Baseline characteristics similar (Selection bias) |
Low risk |
No formal tests of hypotheses were done, but between‐group baseline equivalence appeared to have been attained for baseline characteristics. |
Baseline outcome measurements similar (Selection bias) |
Low risk |
No formal tests of hypotheses were done, but between‐group baseline equivalence appeared to have been attained for quantity of pulses, protein from pulses and protein from all foods at baseline. |
Blinding of participants and personnel (Performance bias) |
Low risk |
Although the participants were not blinded, the intervention (introduction of pulses) was implemented as part of a central government programme. The programme was implemented in a similar way in both intervention and control villages. |
Blinding of outcome assessment (Detection bias) |
High risk |
Not described by study authors. |
Protection against contamination (Performance bias) |
Low risk |
As intervention and control treatments were assigned by state government, it was unlikely that meaningful contamination could have occurred. 2 intervention states in the north (Himachal Pradesh and Punjab) and 2 in the south (Tamil Nadu and Andhra Pradesh) border each other, while Maharashtra (a control state) does not border any of the intervention states; making contamination highly unlikely. |
Incomplete outcome data (Attrition bias) |
Unclear risk |
Total attrition was fairly high (13.0%), but comparable between intervention (20.7%) and control (18.7%) groups. |
Selective outcome reporting (Reporting bias) |
Unclear risk |
No protocol was available and the paper did not follow the usual journal format with a designated Methods section. The aim of the study was to assess the effect of pulse subsidy on pulse consumption and protein intake, both of which are reported on. |
Other bias |
Low risk |
No other biases identified. |