Clinical–psychosocial complexity drives hybrid care allocation in chronic disease cohorts

Hybrid care models are reshaping chronic disease management, yet the clinical logic behind patient allocation remains largely unexamined. Most evaluations assume comparability across modalities, overlooking the clinician-driven selection processes that introduce systematic indication bias. This is especially consequential in low- and middle-income settings, where structural inequalities shape who gets what care and when. We conducted a retrospective cohort study (2019–2025) in a tertiary chronic disease programme in Medellín, Colombia, enrolling 638 adults with hypertension, type 2 diabetes, and/or dyslipidemia (319 hybrid, 319 in-person). Care modality was assigned by treating physicians during routine practice. Multivariable logistic regression identified determinants of allocation and their associations with 3-year outcomes including therapeutic goal attainment and emergency department utilization. Hybrid patients carried substantially greater baseline complexity: poorer metabolic control, higher cardiovascular risk, and lower treatment adherence. High clinical-psychosocial complexity was present in 24.8% of hybrid patients versus only 0.9% of in-person patients. Poor adherence (OR 6.35) and elevated HbA1c (OR 2.70) were the strongest predictors of hybrid allocation. Yet despite these worse starting conditions, hybrid patients achieved markedly higher therapeutic goal attainment (OR 7.34) and lower emergency utilization (OR 0.53) over three years. These findings reframe hybrid care not as a convenient alternative but as a deliberate, complexity-driven clinical strategy. For researchers and policymakers alike, ignoring allocation mechanisms risks producing biased effectiveness estimates and missing critical opportunities to advance health equity in resource-limited settings.

Key messages

• Clinician-driven allocation directs hybrid care to highest-risk patients, revealing strong indication bias in real-world telehealth evaluations.

• Ignoring baseline complexity can misattribute benefits to care modality, risking biased evidence and inequitable scale-up of digital health systems.