A senior manager from South Africa’s Department of Health once said that managing a healthcare system without having enough information is “like trying to fly a plane blind”. Numbers and statistics can seem dry and boring, but having good data is critical when managing health systems. Ben Gaunt, a rural doctor, shares his perspective about trying to improve clinical care.
Why collect data?
Data collection takes time and money. So why do we do it? If systems are to function well, good data are essential. Health facilities need accurate figures to make sure that budgets are done correctly, that drug supply chains work, and that we make the most of the staff and facilities we have.
South Africa has thousands of clinics but its resources are limited. Good data reporting is important to national healthcare planning and keeping control of costs. But at the clinic level, accurate data reporting offers more immediate benefits. The purpose of tracking patients’ viral loads over time, for example, is to help us to improve clinical care both for individuals and communities.
Data collection must have a clear purpose and it must make things better for patients. We need to pay greater attention to feedback loops to help us improve the quality of our data. Evaluations of data quality can be done via informal analysis or more structured audits. However, clinic supervisors are often more focused on maintaining data registers than they are on receiving feedback about data quality. If we want meaningful data, we need to instil a respect for data accuracy and a clear understanding of why we gather data.
What data are we collecting?
Because data collection is costly, we need to ensure that we collect the right information. We can do this by looking closely at the purpose of the data. Is a particular set of data, for example, going to improve service delivery? South Africa’s District Health Information System requires staff to complete multiple data fields, but most nurses and clinicians find it overwhelming. Recording data badly about 100 parameters isn’t nearly as helpful as collecting 10 pieces of information really well. In fact, inaccurate or incomplete data can be harmful to patients and to health systems in general.
The monitoring of South Africa’s antiretroviral treatment (ART) programme focuses on access to care. This approach was critically important in the early years of the ART rollout. But now that more and more people are on treatment, are we ignoring quality-of-care indicators at our peril? How many sites are accurately reporting the number of patients who have had their annual viral load measured according to schedule, or the proportion of patients with viral load suppression?
Who collects the data?
In South Africa, the process of collecting data can be challenging: demand for healthcare is high and nurses are busy. Not all clinics have a data capturer yet. Many of those responsible for capturing data have a poor understanding of the health sector. Supervisor visits to some clinics [may] occur only monthly. In addition, many of those in charge place too much emphasis on numbers instead of what data can tell us about the quality of the clinical care that’s being given.
Poor maths education also impacts on data accuracy. Many data capturers and nurses lack analytical skills and feel uncomfortable when dealing with numbers and statistics. At one clinic, a vitamin A coverage rate of 290% was reported. In other instances, percentages of a whole have not added up to 100. These are basic and unacceptable mistakes.
How do we collect data?
Perhaps one of the biggest obstacles to generating good data in our country is how we collect these figures. Clinics still rely largely on paper-based data gathering. It is easy to see why this is still happening: paper-based systems are simple and don’t rely on expensive hardware. They also do not require additional staff training, they don’t depend on electricity, and their maintenance and running costs are low.
When we implement new kinds of information technology, we should be careful that we do not simply generate an electronic version of what we have already. Many electronic systems, just like our current paper-based system, are designed simply to collect statistics rather than analyse them. Few are able to track patients moving to different areas of the healthcare system. What we need is an awareness of the connection between the data we generate and the clinical care we give. The value of any electronic system depends entirely on whether it lets us meaningfully analyse the data we have.
The right attitude
We need to improve South Africa’s data systems. But we also need to improve the quality of the data we collect. It is important that the Department of Health recognises the gaps and flaws in our current system and does this with constructive support from members of civil society and the media. Improvements in data quality are critical and will help significantly in our fight to provide better healthcare.Ben Gaunt is a medical doctor working in the Eastern Cape who is passionate about providing high-quality care to rural communities. In 2013, he received a Rural Doctor of the Year Award from RuDASA (the Rural Doctors Association of South Africa).