Which good practice exists for policymaking under uncertainty? An interdisciplinary review with relevance for health technology assessment (HTA).

Appropriate policymaking under uncertainty can lead to better decisions around how constrained resources are used. Although approaches from the technical sciences can provide robust, generalisable uncertainty estimations, they cannot account for the varying social, economic, and cultural contexts in which policies are made and implemented. We evaluated current health technology assessment (HTA) practice, finding limited consideration of social science perspectives, which may be inadequate for policies that influence behaviour, cross jurisdictions, or deal with multiple vested interests. To identify approaches that could address this gap, we conducted an interdisciplinary review of frameworks for policymaking under uncertainty, using an iterative approach for cross-disciplinary learning. We applied a comprehensive pearl growing search strategy and conducted inductive analysis combining grounded theory coding with pattern matching. We identified fourteen approaches to facilitate social science conceptualisations of uncertainty that are currently missing from HTA practice, alongside features of the decision problem and context that influence approach selection. Within the context of HTA, our analysis suggests that these approaches would be most applicable for interventions that change how care is delivered, highly contentious policy questions, and settings with poor implementation of HTA recommendations. We illustrate the applicability of our findings for three HTA challenges where conventional approaches are proving insufficient: reducing unnecessary caesarean section, reimbursement of locally manufactured products, and integrating patient voice in policymaking. We conclude that the policy questions reaching HTA agencies increasingly require a social science perspective, but that successfully doing so will require capacity building, philosophical alignment, and institutional adaptation.