The narrow scope of existing information systems

Despite the significant use of the private sector by the population to seek services, majority of private health facilities do not provide data routinely to the MoH.6 As a result, health sector monitoring is not comprehensive to realistically assess progress in the attainment of SDG3 targets and indicators. For example, the use of the private for-profit sector is as high as 39.7% in Uganda7 and 47% of the poorest quintile in Kenya,8 yet the private sector in these countries does not provide data to ministries of health. Several scholars cite similar challenges of multiple sources of data outside the public health system with limited information sharing.9 Acknowledging this fact, De Costa and Diwan emphasised the need for the public sector to play an oversight role over the entire sector and build strong public–private partnerships.10 This oversight role should also be extended to data sharing and monitoring sector performance. However, ensuring that the private sector abides with the reporting obligations may have cost implications. The provision of reporting tools and training in data management are not cost neutral, and these costs may have to be borne by the government. The narrow scope has also been noted in reference to specific data sources where survey questionnaires are influenced by funding agencies as opposed to realistic information needs. The skewing in the data collected is another concern. In monitoring health sector performance, the balance of indicators across the domains of inputs, outputs, outcomes and impact is important.11 Data must be collected on the inputs into service delivery, for example, medicines and number of health workers; the processes undertaken to translate inputs into outputs, for example, training of health workers to deliver services better; the outputs, for example, number of patient attended to; the outcomes realised, for example, coverage of services; and the impact attained like reduction in mortality. However, evidence has shown focus on a limited set of indicators and targets1 3 as well as inadequate attention paid to certain health system building blocks.10 Strengthening information systems to assess performance of SDG3 targets should not be pursued at the expense of comprehensive information systems strengthening to assess health sector performance. Boerma et al cautioned against this possibility given their findings in the MDG era where readily available data sets centred around MDG-related indicators.3

Weaknesses in existing information systems and lack of capacity for data analysis

Available data are not disaggregated enough to monitor equity in line with policy-relevant equity stratifiers. The WHO proposes equity stratifiers along several dimensions, including sex, age, residence, social economic category and minority status.1 However, in pursuant of equity, stratifies of interest should be country specific, implying that there may be additional parameters of policy relevance. For example, in some countries, access to health services by religious groups is of policy interest.12 Although population-based surveys are a good sources of some of the equity stratifiers like socioeconomic categories, rural/urban residence and education status, the frequency of these is irregular. Periodicity of population-based surveys is noted to vary between 4 and 15 years. This will not suffice for the SDG era, given the onus of regular monitoring to keep track of progress of indicators. Boerma et al cited similar challenges and noted the erratic and inaccessible nature of the surveys conducted by international agencies.3 Berman et al, in their work on MDG count down in 75 priority countries, highlighted limitations related to population-based surveys with reference to periodicity and different recall periods, making causal inferences, cross-country comparisons and linkages between policy adoption and intervention coverage trends difficult.13 Furthermore, in most cases, sample sizes from population-based surveys are not adequate to generate reliable subnational and special population group estimates.14 However, large country-wide surveys are costly, and low-income and middle-income countries do not possess adequate resources to conduct more frequent surveys. Additional challenges relate to the inadequate capacity of country statistical units in reference to skills, staffing and required resources, which must be addressed. On the other hand, we need to critically examine the extent to which these equity stratifiers can be accommodated within existing information systems. The envisaged time frame to realise changes given investments made may guide the process of mainstreaming equity stratifiers in the existing information systems. For example, even with the best of investments, equity in access by social economic category cannot be realised in a period of 1 year. This then is best mainstreamed in surveys that take place every 3 years. On the other hand, equity in access by gender can be monitored annually. Furthermore, some of the equity analysis could be in the research domain emphasising the need to strengthen national health research systems, which are currently performing at varying levels in the WHO African Region.15

Whether all indicators can be realistically measured is another question. Majority of indicators can however be measured given the fact that these were derived from the 100 WHO core indicators, which were compiled in consultation with member states based on measurement feasibility. Data for one-third (approximately) of the indicators can be collected from civil registration systems, one-third from surveillance and routine systems and the rest from the population-based surveys. On the contrary, for a few new indicators, measurement feasibility needs to be deliberated. For example, how do we measure ‘well-being’? Challenges are viewed from two fronts: (1) the contextually appropriate definition of such an indicator and (2) the feasibility of existing information systems to collect relevant information. AbouZahr and Boerma, in the review of health information systems (HIS), highlighted the strong focus on quantitative data.4 How do countries incorporate ‘well-being’ in routine data collection instruments? In some contexts, incorporating all relevant indicators in national information systems is unrealistic, and in such cases, prioritisation of indicators and linkages with already existing national health indicators needs to be explored to minimise duplication. Boerma et al emphasised the need for strengthening health management information systems to incorporate the measurement of essential SDG-type indicators, specifically on quality of care, safety, client centredness and NCDs.3

Multiplicity of information systems

Although information systems exist in majority of countries, these are inadequate due to several reasons. There is a lot of data collected by the different stakeholder, which are not available to the MoH due to lack of interoperability of the systems and lack of institutionalised data sharing procedures. Gloyd et al documented inconsistencies, numerator–denominator incompatibility, missing and doubtful data regarding the data for the same indicators derived from routine information system and programme-based information systems.2 Fragmentation of HIS and lack of effective coordination and leadership are known hindrances to timely availability of data to inform planning and decision-making.16 17 Seitio-Kgokgwe et al5 also drew our attention to similar concerns where the development of the national HIS in Botswana was negatively affected by weak leadership, weak policy and regulatory frameworks.

Given the multisectoral nature of SDGs, some relevant data sources are not managed by the MoH. Information systems relevant to the monitoring of SDGs are managed by several ministries and government agencies creating fragmentation and compromising countries’ overall efforts to meet their monitoring commitments. Particularly in the WHO African region, ministries of finance, health, international cooperation, agriculture and planning can all be involved at different levels. Coordination between ministries needs to improve in order to harmonise and create a whole-of-government approach to monitoring. The weak intersectoral collaboration in health development in general impacts similar collaboration in accessing data. Little success has been realised in institutionalising multisectoral collaboration in health development despite a boost by the social determinants of health18 and the macroeconomics and health commissions19 that put forward evidence and strongly advocated for this approach. Highlighted challenges include operation issues like who chairs the multisectoral forum, who takes charge of the multisectoral agenda once agreed and competition for limited resources.20–22 Other documented challenges include the lack of feedback mechanisms, shared responsibility and incentives to reward collaboration efforts.21

Undue emphasis on ensuring measurement of SDGs

The emphasis on compiling statics was noted in the era of MDGs.1 On a positive note though, capacity was strengthened to track progress. This notwithstanding, we need to balance investments in strengthening monitoring and implementation to attain SDGs, and in this regard, emphasis would be on using existing information systems. The fear of undue emphasis on monitoring of SDGs should be discussed in light of the need for a holistic approach to strengthening information systems. Data are primarily important in informing planning and resource allocation even beyond the SDG targets and indicators. We need to look at the HIS in a more holistic manner so that we improve health planning in general.

Available opportunities

Global health initiatives (GHIs) offer opportunities that can be exploited to strengthen information systems. However, skewed investments in information needs that address GHI-funded programme have been documented.3 As such, the need for better negotiation by ministries of health to ensure a holistic approach to strengthening information systems is crucial.23 The health data collaborative (HDC) that seeks to align partners’ support towards a consensual country-owned monitoring and evaluation (M&E) framework is another opportunity. There are positive lessons of the HDC from Kenya with realised partner alignment to one M&E plan, greater engagement of national statistical office and improved data quality. Key prerequisites are stated as clear country priorities with stakeholder roles and responsibilities defined, strong country leadership and strengthening existing systems as opposed to creating new ones.