gms | German Medical Science

Background/research question: Older patients with multiple chronic conditions identified optimal blood pressure targets and diabetes second-line medications as top priority questions. We engaged stakeholders throughout the process of quantitative benefit-harm assessments to inform patient-centered care, in the context of guideline development. Because guideline developers enquired about the uncertainty in the benefit-harm balance, we aimed to determine the potential impact of lacking evidence and the overall quality of evidence.

Methods: We identified patient-important outcomes for both questions in a systematic manner based on input from focus groups of patients. We elicited patient preferences quantitatively using best-worst scaling in members of Kaiser Permanente Colorado who had hypertension resp. diabetes type 2, at least two other chronic conditions and were 60 or older. We performed a systematic literature review, selected the most applicable, valid and precise evidence and modeled the benefit-harm balance quantitatively. To address lacking evidence, we performed sensitivity analyses and estimated the potential impact of each patient-relevant outcome on the benefit-harm balance, and assessed the risk of bias for each input parameter.

Results: For blood pressure targets, most patient-relevant outcomes were reported, and the benefit-harm balance varied depending on individual preferences and risk factors. However, for the comparison of adding basal insulin vs sulfonylurea to metformin, half of the patient-relevant outcomes were not measured in the trials. The largest uncertainty was caused by lacking evidence on incident chronic kidney disease. The overall quality of evidence was low, i.e. additional trials are likely to change the estimated benefit-harm balance.

Conclusion: Stakeholders should be engaged to design trials to ensure all patient-relevant outcomes are measured and reported. Quantitative benefit-harm assessments can be used to inform patient-centered care and can pinpoint the most crucial lacking evidence. Guideline recommendations should clarify the uncertainty in the benefit-harm balance.

Competing interests: The authors declare no conflicts of interests.