Research result: Evaluating the effectiveness of healthcare requires sufficiently sensitive indicators

In assessing the effectiveness of healthcare, indicators are not just numbers: they guide decisions, funding, and treatment practices. Therefore, the question is not only whether we measure, but whether we measure the right things. The indicators used must be able to genuinely describe the changes that are meaningful to patients and society.

In our recent study, "The ability of EQ-5D -5L and WHODAS 2.0 to detect relevant change in primary health care outpatients with diagnosed depressionOpens in a new window," we examined how well the EQ-5D-5L, which has been selected as the national impact indicator, is able to detect changes in the well-being of primary health care patients suffering from depression.

EQ-5D-5L is a five-question measure that assesses health-related quality of life. It is used in Finland and widely around the world to measure how different forms of treatment affect perceived quality of life.

We compared the results of the EQ-5D-5L with the results of the PHQ-9, a measure specialized in monitoring the response to depression treatment, in a sample of nearly 5,000 patients. The PHQ-9 is a widely used measure of the severity of depression symptoms and is recommended for monitoring the results of depression treatment.

The results of our study showed that the generic quality of life measure EQ-5D-5L was not sensitive enough to identify changes that were meaningful to patients with depression: one-third of patients were not correctly classified. The PHQ-9, which specializes in monitoring depression, showed a clearly better ability to distinguish whether the patient's condition had improved, worsened, or remained unchanged.

Depression is one of the most common and significant diseases treated in primary health care. It is a common, everyday ailment that has a major impact on both individual well-being and the functioning and costs of society as a whole. The patients in the study also had a lower quality of life on average at baseline and 25 days of sick leave in the six months prior to the study.

Our recent study shows that the sensitivity of the EQ-5D-5L, which has been selected as the national impact indicator, to detect changes that are meaningful to patients is inadequate, at least in this patient group.

Measuring effectiveness must be appropriate and responsible

It is worth pausing to consider what happens if we use indicators that do not recognize significant changes in patient well-being. At worst, an insensitive indicator can produce random results – it does not distinguish whether the treatment has been effective or not. If healthcare funding is guided by such indicators, we may end up in a situation where the choice of treatment and the allocation of resources are based more on chance than on actual effectiveness. This can lead to the wrong outcome for both the patient and society and undermine confidence in the fairness of the system.

Effectiveness must be measured, and resources must be allocated on the basis of measurement. Patients must be informed of the reasons for the treatment recommended to them. However, measurement must be based on appropriate, researched indicators that are suitable for the target group. One indicator may not be suitable for all situations or patient groups.

Responsible impact measurement requires systematic evaluation of the validity of indicators and, if necessary, their updating and piloting in different customer groups and services. This allows for reliable evaluation of treatment outcomes and the correct allocation of healthcare resources.