Strategies for balancing internal and external validity in evaluations of interventions

Background Evaluations of interventions should be carefully designed and conducted to maintain a balance between internal and external validity, with the dual goal of minimising the influence of potential confounders and improving the generalisability or applicability of any findings to practice. Aim To review strategies to promote balance between internal and external validity in an evaluation of a cognitive-behavioural intervention for chronic insomnia. Discussion A pragmatic approach is needed to balance internal and external validity, and generate evidence relevant to practice. The authors present strategies to promote such a balance, including using strict eligibility criteria, subgroup analysis, random assignment of patients based on preferences, a no-treatment control condition, and standardised and consistent implementation of the intervention. Conclusion A balance between internal and external validity is essential to promote enrolment in the study and confidence in attributing its outcomes to an intervention, as well as to provide answers to clinically relevant questions such as who benefits most from which intervention. Implications for practice The authors recommend researchers conduct a pilot study in advance of an evaluation, to help decide which strategies to use and how to balance internal and external validity.