Data pertaining to health care in India, evidence shows, is significantly compromised in terms of its quality, its periodicity and coverage. In addition, “there is a visible discrepancy between the type of information available and what is required by health planners, medical scientists and researchers,” says a recent paper by the Health Team of the National Institute of Public Finance and Policy (NIPFP), New Delhi.
It has been well recognised and acknowledged by the government officials that data collection system in India needs to be completely revamped as different data sources lead to different conclusions. So if you want to know the proportion of births that were delivered by caesarian section in a private health facility of Andhra Pradesh, National Family Health Survey (NFHS)-4 will say it is 57 per cent but the Health Information System of the National Rural Health Mission 2015-16 pegs it at around 42 per cent.
Data gaps
In an issue brief published by the Observer Research Foundation (ORF), four major data gaps were identified. First, there is a lack of data at the sub-State or the district level, making it difficult to plan for targeted interventions. Second, data is collected at irregular intervals, which doesn’t allow for mid-course policy correction. For example, the fourth round of NFHS was conducted after a gap of nine years. Third, data remains incomplete in many surveys and tools, especially in administrative data at hospitals and nursing centres in smaller towns and districts. Fourth, which is a problem across the board, is data quality. The issue brief mentions that the lack of an independent quality control body limits the quality of available data, especially given that the information passes through various layers before reaching the stage of evaluation and analysis.
In order to ensure that we get credible and reliable data, systemic concerns must be taken into consideration. Figuring out why the data was collected often hints towards how it was collected. For instance, there are reports that have found that data for various indicators is being over reported because ground staff who put the data together, are aware that the set of indicators being reported is vital for the success of the programme, and consequently serves as a benchmark of their own performance, leading to conflict-of-interest. Further, the NIPFP paper says that the lack of training on probing skills of the enumerators of large-scale surveys has been in question. “The way questions are put forth changes the spending reported in the Consumer Expenditure Survey and National Sample Survey Office (NSSO) Morbidity and health care surveys, leading to different results.”
Lack of health-care data from the private sector is another major caveat. The NIPFP paper finds that the National Health Profile published by the Central Bureau of Health Intelligence, New Delhi, remains “severely compromised” in terms of health expenditures incurred at private facilities — even though 70 per cent of the health expenditure is reported at those institutions.
Evidence-based, informed health policies require data. But there is more to it: good quality data is quite crucial for efficient allocation of our limited resources. Consider this: the head of AIIMS, told Alex Pentland, co-creator of the MIT Media Lab, that 90 per cent of the drugs were wasted in India because of the inability to track disease prevalence, and as a result, same package had to be handed out to all districts. This is a low-hanging fruit where properly maintained data can help reduce unjustifiable wastage.
However, good quality data itself has no value if people don’t make use of it. Even when quality has improved for certain data sets such as the Health Management Information System (HMIS), stakeholders haven’t started using it much. Offtake is low by the policy community at large, and much lower by administrators, which in turn drives the lack of enthusiasm to further improve the data systems.
The way forward
To strengthen data quality, the NIPFP paper makes four suggestions. First, the definitions for various indicators need to be standardised across surveys, even if conducted by different organisations. Second, the best way to data is to ensure routine capturing of disaggregated data, without duplication of forms and formats to reduce the efforts of data producers. Third, stakeholders and decision- makers will have to place a positive value on data in order to use it in decision-making. This can be generated by building a positive experience using information to support a decision and through proper training. Fourth, the capacity building of staff needs to be a focus area, providing them with necessary skills.
Generating and maintaining high standards of data is not an end in itself, but a means to an end, that of making quality health-care accessible and available to all.
It has been well recognised and acknowledged by the government officials that data collection system in India needs to be completely revamped as different data sources lead to different conclusions. So if you want to know the proportion of births that were delivered by caesarian section in a private health facility of Andhra Pradesh, National Family Health Survey (NFHS)-4 will say it is 57 per cent but the Health Information System of the National Rural Health Mission 2015-16 pegs it at around 42 per cent.
Data gaps
In an issue brief published by the Observer Research Foundation (ORF), four major data gaps were identified. First, there is a lack of data at the sub-State or the district level, making it difficult to plan for targeted interventions. Second, data is collected at irregular intervals, which doesn’t allow for mid-course policy correction. For example, the fourth round of NFHS was conducted after a gap of nine years. Third, data remains incomplete in many surveys and tools, especially in administrative data at hospitals and nursing centres in smaller towns and districts. Fourth, which is a problem across the board, is data quality. The issue brief mentions that the lack of an independent quality control body limits the quality of available data, especially given that the information passes through various layers before reaching the stage of evaluation and analysis.
In order to ensure that we get credible and reliable data, systemic concerns must be taken into consideration. Figuring out why the data was collected often hints towards how it was collected. For instance, there are reports that have found that data for various indicators is being over reported because ground staff who put the data together, are aware that the set of indicators being reported is vital for the success of the programme, and consequently serves as a benchmark of their own performance, leading to conflict-of-interest. Further, the NIPFP paper says that the lack of training on probing skills of the enumerators of large-scale surveys has been in question. “The way questions are put forth changes the spending reported in the Consumer Expenditure Survey and National Sample Survey Office (NSSO) Morbidity and health care surveys, leading to different results.”
Lack of health-care data from the private sector is another major caveat. The NIPFP paper finds that the National Health Profile published by the Central Bureau of Health Intelligence, New Delhi, remains “severely compromised” in terms of health expenditures incurred at private facilities — even though 70 per cent of the health expenditure is reported at those institutions.
Evidence-based, informed health policies require data. But there is more to it: good quality data is quite crucial for efficient allocation of our limited resources. Consider this: the head of AIIMS, told Alex Pentland, co-creator of the MIT Media Lab, that 90 per cent of the drugs were wasted in India because of the inability to track disease prevalence, and as a result, same package had to be handed out to all districts. This is a low-hanging fruit where properly maintained data can help reduce unjustifiable wastage.
However, good quality data itself has no value if people don’t make use of it. Even when quality has improved for certain data sets such as the Health Management Information System (HMIS), stakeholders haven’t started using it much. Offtake is low by the policy community at large, and much lower by administrators, which in turn drives the lack of enthusiasm to further improve the data systems.
The way forward
To strengthen data quality, the NIPFP paper makes four suggestions. First, the definitions for various indicators need to be standardised across surveys, even if conducted by different organisations. Second, the best way to data is to ensure routine capturing of disaggregated data, without duplication of forms and formats to reduce the efforts of data producers. Third, stakeholders and decision- makers will have to place a positive value on data in order to use it in decision-making. This can be generated by building a positive experience using information to support a decision and through proper training. Fourth, the capacity building of staff needs to be a focus area, providing them with necessary skills.
Generating and maintaining high standards of data is not an end in itself, but a means to an end, that of making quality health-care accessible and available to all.
Comments
Post a Comment