Medical authority values the consistency of health economic evaluations

Health economic evaluations bring additional information to the authorities’ decision-making concerning medicinal products. How does a good health economic evaluation look like from the perspective of an expert in the Pharmaceutical Pricing Board and which issues should be developed?

Expert in Health Economics, Kalle Aaltonen, has worked five years in the Pharmaceutical Pricing Board Hila. During the time, he has developed a good understanding of what sort of a health economic evaluation optimally supports the work of authorities and decision-makers and which issues usually raise questions.

Hila decides on the medicinal products for outpatient healthcare included in the reimbursement system as well as their wholesale prices and reimbursement groups.

What does a good health economic evaluation look like?

According to Aaltonen, from the perspective of an expert at Hila, the key in health economic evaluations is open and detailed reporting illustrating the utilised reference data, assumptions made, structure of the model and justifications for the choices.

“The calculation model with which the results have been modelled should be freely adjustable and repeatable. This is usually the case.”

A common requirement for the evaluations is that their source data includes the most recent available research data openly and non-selectively. Consistency is also an essential requirement of the evaluations. According to Aaltonen, this is an issue that they are often forced to give feedback in Hila’s assessments.

The internal validity of a review requires that the used source data are internally consistent. This means, for example, that the health impacts received with treatment and the used dose are based on the same source.

“The results of the evaluation may not be fully generalisable to Finland, if a clinical trial used as source data deviates from the Finnish treatment praxis. Then again, individual source data, such as follow-up treatment received by a patient, cannot be changed without breaking the internal validity of the evaluation unless the wider impact of the change is taken into consideration”, Aaltonen states.

Another very typical feedback especially in the cost-effectiveness models of cancer medicine relates to the assumption on how the benefit of treatment is modelled after treatment discontinuation to the lifetime benefit.

“If the market authorisation holder models that a cancer medicine should have a lower mortality than a comparison product, it should be modelled whether the benefit continues after treatment discontinuation. In these stratified lifetime models, there should always be more than only the most optimistic scenario under review.”

A third key issue in a good evaluation is the consistency in the proposed coverage of reimbursement and the patient population of a health economic evaluation and in the key assumptions, such as reaching treatment response as a condition for continuing treatment.

Need for domestic data

Local Finnish real-world evidence and observation-based research data can improve the generalisability of a health economic review, according to Aaltonen. However, it requires that the data can be connected to the source data of a clinical trial without endangering the internal validity.

“Typical source data based on Finnish research evidence, or expert assessments, include the utilisation of healthcare resources and their unit costs.”

Regression models can be utilised in the improvement of generalisability. For example, a clinical trial on the relative effect of treatment can be extrapolated to the Finnish patient population.

“In some situations, this can be useful, but the challenge is that a company is not required to generalise. This may lead to generalisation being conducted when the results are more favourable and not being conducted if they are not favourable. This is why we always want to see also the non-adjusted result of the original clinical trial.”

It is difficult to receive national data on treatment practices

In the decision-making of the Pharmaceutical Pricing Board, data on current treatment praxis, size of patient population and costs of treatment play a central role. Receiving national data is, however, difficult, and Aaltonen hopes this would change.

“In terms of outpatient medicine and their users, we receive some nationally covered data from Kela’s subsidy register. The situation is more difficult for hospital-only therapies. I have big interest in and high expectations for the possibility of studies based on nationally covered patient document data, inter alia, with the help of new quality registers”, Aaltonen says.

According to Aaltonen, the significance of RWE in the assessment of medicinal treatment in the future is impacted, among other issues, by how patient documents can be used, how authorisation processes work and how pharmacotherapies can be validated in everyday life. RWE does not, however, replace randomised clinical trials but validates and completes data received from them.