On the Nature of Population-based Information Systems and Health Services
In section 2.7 we argued that (also) being population-based in terms of data sources and use areas are key characteristics of public health information systems. This focus on the population, or target groups, or service-user groups is in contrast to the clinical information systems perspective where the focus will typically be on the individual. Population and patient-based clinical systems are different in that they are used for different purposes and are developed and designed according to different logics. In the context of public health informatics, it is important to understand the difference between these types of systems and what population-based means in terms of system design. Here we will identify and describe two important population-based features in public health information systems: first, the use of population-based data, such as from censuses and surveys, as denominators in calculating indicators; and second, information support of population-based health services, such as antenatal care and immunization.
Population-based data, such as from censuses and estimates of target populations, make up a central component of the public health information system, as these data are used as denominators in the calculation of coverage indicators. Public health information systems have, as a rule, a strong denominator focus. These denominator data will always have a certain margin of error and will rarely be accurate. The number used for expected pregnancies in African countries will typically be calculated as four per cent of the estimated population for an area, which again will be based on annual estimates of population growth since the last census, typically carried out many years earlier. Such ‘estimates of estimates’ will never be accurate. These denominator data are therefore of a different type than the numerator data, which are typically the ‘counts’ of the services being rendered at the health facility, whether the numbers are aggregated from electronic records, paper registers, or tick sheets. While the target population denominator data in most cases will be estimates and characterized with a certain uncertainty and fuzziness, the numerator data are different in that they aim at being accurate. Of course, there are also huge data quality issues with the numerator data aggregated from the health services records. But the point here is that there are different types of systems and methodologies used to arrive at the data components of an indicator. We discuss data quality issues related to population data and their relation to complexity in Chapter 7.
We also include information support to population-based health services more generally, to be part of what we term the population-based perspective of public health information systems. Immunization and mother and child health services are population based, in that the primary aim is to reach the entire target population and not only those that are coming to the health facilities. While aiming at identifying and reaching the population unable to access particular health services, we have even greater challenges regarding open-endedness and uncertainty with population data in general. Information systems to support population-based health services will therefore also have a certain open-ended scope in its reach and design. As primary healthcare expands with increasing resources and political commitment the number of such population-based health services increases.
While highlighting the population-based aspects of public health information systems and what distinguishes these systems from typical clinical approaches, it is important to emphasize that public health information systems also include medical records and other clinical approaches. Ensuring continuity of care for tuberculosis (TB) and HIV/AIDS patients, and for other patients with chronic diseases in communities, is typically central to clinical approaches, but is also part of the PHI approach, since this is a public obligation with considerable externalities. Similarly, many aspects of district hospital management form part of both clinical and public health informatics. Clinical approaches are therefore also part of the public health informatics framework, which may be understood as approaches to extend appropriate clinical approaches; from the individual treated in the hospital, to all those requiring care in the community. The other way is also true; population-based ‘serve the people’ perspectives will help focus and also improve clinical approaches. It is therefore important to see population and clinically-based approaches as a continuum of mutual synergies.
