Beyond the Hazard Ratio: Development of a Synthesis Readiness Framework for Effect-Scale Heterogeneity in Continuous Biomarker Evidence

Abstract Background Quantitative synthesis of studies of continuous biomarkers assumes that reported effect estimates target sufficiently comparable estimands. However, differences in exposure scaling, cohort-specific variability, biomarker derivation, and reporting of distributional parameters may create effect-scale heterogeneity. Such heterogeneity can produce pooled estimates that are statistically precise but difficult to interpret on a common biological scale. We developed a proof-of-concept Synthesis Readiness Framework and applied it to associations between GrimAge-related measures and all-cause mortality. Methods We conducted a systematic review of prospective cohort studies, registered prospectively in PROSPERO (CRD420261449280) and reported in accordance with PRISMA 2020. We conceptualised heterogeneity in three layers: biological/demographic, analytical/methodological, and metric/effect-scale heterogeneity. We operationalised the Synthesis Readiness Index (SRI), a four-domain tool assessing estimand alignment, metric comparability, distributional availability, and calibration feasibility. Each domain was assigned a prespecified score from 0 to 2, yielding a total score ranging from 0 to 8. We also conducted a deterministic numerical illustration to demonstrate how an identical log-linear association can yield different hazard ratios when reported using cohort-specific standard deviation units. Results Five prospective cohorts comprising 7,438 participants were included. The SRI distinguished studies with high, conditional, and low readiness for effect-scale harmonisation. In several cohorts, incomplete reporting of standard deviations or equivalent distributional information precluded direct transformation of standardised estimates to a common absolute exposure scale without additional assumptions or external data. In the numerical illustration, a reference hazard ratio of 1.20 per 1-year increase corresponded to hazard ratios of 2.49 and 1.73 per 1-standard-deviation increase when the corresponding standard deviations were 5 and 3 years, respectively. Conclusions Consistent effect direction does not establish quantitative comparability. The Synthesis Readiness Framework provides a structured pre-pooling assessment of estimand alignment, exposure-scale comparability, distributional reporting, and calibration feasibility. The SRI is intended as a transparent methodological decision aid to guide the choice between quantitative harmonisation, structured narrative synthesis, and individual participant data meta-analysis; it is not a validated rule for determining meta-analysis eligibility.