Health data consistency and management, case study of maternal health data in Malawi

Good quality data is vital. It informs decision making in a wide range of sectors, at all levels. Accuracy, completeness, consistency, timeliness and standard-based are the properties of good quality data. Data which satisfies these properties is deemed fit and appropriate for its intended use. Data lacking these qualities poses challenges in operations, decision making and planning.The Millennium Development Goals (MDG), specifically MDG number 5- the reduction of maternal mortality by two-thirds between 1990 and 2015- now succeeded by Sustainable Development Goal (SDG) number 3 - has led to a significant need for reliable maternal health data. Accurate data is needed on the levels of (and trends in) maternal death in developing countries in order to address and improve maternal health and survival. The health care system in Malawi lacks vital registration systems which are rich and valuable sources of health data, including maternal health. Malawi had a health information system that did not produce reliable data, and therefore could not be used for decision making in terms of planning and management with respect to maternal health. In 1999, after reviewing the Health Management Information System (HMIS), Malawi developed a new system based on the strengths of the old system while addressing its weaknesses.This study aimed to use maternal health data to investigate management processes and procedures within the HMIS from data collection by different entities, transitioning through the hierarchy from the community to the point of use at district level. The study also assessed the consistency of the data itself as it transitioned through this hierarchy. Transitions of data were explored, and difficulties in maternal health data collection were assessed among stakeholders including community members, government health staff and non-governmental organisations (NGOs). Monthly and annual data collected and compiled by various personnel over one year, was also tested for consistency using chi-square test.The study was carried out in three phases. The first was done in the Southern Region of Malawi (Phase 1). The second one was done in three health facilities which had interventions (Phase 2) in terms of training, infrastructure and provision of resources in maternal and neonatal health. The third one was conducted in six health facilities (three with (as in Phase 2) and three without intervention (Phase 3);Ten HMIS officers and 10 data users (programme coordinators at the District Health Office (DHO)) were selected for Phase One since they were involved in data management and decision making processes respectively given the data. For Phase Two participants included 14 Secret Women, 16 Health Surveillance Assistants (HSAs), Three Village Health Committees (VHCs) and two Health Personnel. Phase Three participants included 17 Secret Women, 42 Chiefs, 40 Health Surveillance Assistants (HSAs), six Health Personnel and also one Safe Motherhood Coordinator, one Community Based Maternal and Neonatal Health Coordinator, one NGO and one HMIS Officer chosen from the District Health Office, all of whom are involved in the data management process until the data is sent to the users.Cross-tabulations, frequency tables and graphs were used to assess data management processes, procedures and problems among the personnel. Testing data uniformity was achieved using the Chi-square test of homogeneity to compare monthly data and annual data aggregates for the groups of personnel to check for data quality.The results showed that data management was compromised by problems faced by data collection personnel such as lack of transportation affecting timeliness of data submission, lack of basic needs (e.g. proper housing and low salaries for HSAs and Health Personnel) which affected their motivation to work, and lack of reporting forms and writing materials which led to data gaps and missing information. Discrepancies arose in compilation and transfer of information since some information was forgotten or not recorded during the process. Furthermore, lack of supervision coupled with lack transportation and stationary led to inconsistent, incomplete, inaccurate and unreliable data.The quantitative analysis showed that there were significant differences, thus no consistency, in the monthly and annual data for the selected variables i.e. new pregnancies, births, live births among the groups of personnel. Monthly data for maternal and neonatal deaths also showed differences among the personnel, with annual aggregates also showing differences.;Important resources such as stationary and reporting forms should be provided in good time (and in adequate numbers) to ensure that there are no data gaps. In addition, the study strongly recommends the use of eHealth/mHealth in rural communities to reduce errors and data gaps during entry, so as to increase accuracy, reliability, consistency,completeness and timeliness. It also recommends training for new officers, and refresher courses for those already in the system to instil procedures and for the purpose of review (supervision) of work that has been completed.Key words: Accuracy, Consistency, Maternal Health, Data Management, HMISGood quality data is vital. It informs decision making in a wide range of sectors, at all levels. Accuracy, completeness, consistency, timeliness and standard-based are the properties of good quality data. Data which satisfies these properties is deemed fit and appropriate for its intended use. Data lacking these qualities poses challenges in operations, decision making and planning.The Millennium Development Goals (MDG), specifically MDG number 5- the reduction of maternal mortality by two-thirds between 1990 and 2015- now succeeded by Sustainable Development Goal (SDG) number 3 - has led to a significant need for reliable maternal health data. Accurate data is needed on the levels of (and trends in) maternal death in developing countries in order to address and improve maternal health and survival. The health care system in Malawi lacks vital registration systems which are rich and valuable sources of health data, including maternal health. Malawi had a health information system that did not produce reliable data, and therefore could not be used for decision making in terms of planning and management with respect to maternal health. In 1999, after reviewing the Health Management Information System (HMIS), Malawi developed a new system based on the strengths of the old system while addressing its weaknesses.This study aimed to use maternal health data to investigate management processes and procedures within the HMIS from data collection by different entities, transitioning through the hierarchy from the community to the point of use at district level. The study also assessed the consistency of the data itself as it transitioned through this hierarchy. Transitions of data were explored, and difficulties in maternal health data collection were assessed among stakeholders including community members, government health staff and non-governmental organisations (NGOs). Monthly and annual data collected and compiled by various personnel over one year, was also tested for consistency using chi-square test.The study was carried out in three phases. The first was done in the Southern Region of Malawi (Phase 1). The second one was done in three health facilities which had interventions (Phase 2) in terms of training, infrastructure and provision of resources in maternal and neonatal health. The third one was conducted in six health facilities (three with (as in Phase 2) and three without intervention (Phase 3);Ten HMIS officers and 10 data users (programme coordinators at the District Health Office (DHO)) were selected for Phase One since they were involved in data management and decision making processes respectively given the data. For Phase Two participants included 14 Secret Women, 16 Health Surveillance Assistants (HSAs), Three Village Health Committees (VHCs) and two Health Personnel. Phase Three participants included 17 Secret Women, 42 Chiefs, 40 Health Surveillance Assistants (HSAs), six Health Personnel and also one Safe Motherhood Coordinator, one Community Based Maternal and Neonatal Health Coordinator, one NGO and one HMIS Officer chosen from the District Health Office, all of whom are involved in the data management process until the data is sent to the users.Cross-tabulations, frequency tables and graphs were used to assess data management processes, procedures and problems among the personnel. Testing data uniformity was achieved using the Chi-square test of homogeneity to compare monthly data and annual data aggregates for the groups of personnel to check for data quality.The results showed that data management was compromised by problems faced by data collection personnel such as lack of transportation affecting timeliness of data submission, lack of basic needs (e.g. proper housing and low salaries for HSAs and Health Personnel) which affected their motivation to work, and lack of reporting forms and writing materials which led to data gaps and missing information. Discrepancies arose in compilation and transfer of information since some information was forgotten or not recorded during the process. Furthermore, lack of supervision coupled with lack transportation and stationary led to inconsistent, incomplete, inaccurate and unreliable data.The quantitative analysis showed that there were significant differences, thus no consistency, in the monthly and annual data for the selected variables i.e. new pregnancies, births, live births among the groups of personnel. Monthly data for maternal and neonatal deaths also showed differences among the personnel, with annual aggregates also showing differences.;Important resources such as stationary and reporting forms should be provided in good time (and in adequate numbers) to ensure that there are no data gaps. In addition, the study strongly recommends the use of eHealth/mHealth in rural communities to reduce errors and data gaps during entry, so as to increase accuracy, reliability, consistency,completeness and timeliness. It also recommends training for new officers, and refresher courses for those already in the system to instil procedures and for the purpose of review (supervision) of work that has been completed.Key words: Accuracy, Consistency, Maternal Health, Data Management, HMI