30-Second Takeaway
- Dexmedetomidine reduces postoperative delirium in elderly spinal surgery patients (**OR 0.35**) and is the only intervention with moderate-certainty evidence.
Latest - Week ending May 2, 2026
Grand Rounds: Practical Evidence Briefs for Geriatricians — Care coordination, telehealth, readmission risk, caregiver support, and delirium prevention
Proactive care coordination before hospitalization did not reduce ED visits or hospitalizations
Randomized trial in 400 adults ≥65 with CVD risk and highly fragmented outpatient care compared proactive outreach versus usual post‑hospital coordination. No difference in ED visits or hospitalizations (0.25 vs 0.21 events per 100 person‑days; P=0.29). Uptake was low: 26.5% of eligible intervention patients accepted proactive coordination versus 100% in usual care. Conclusion: offering coordination earlier did not improve utilization and many patients declined help, suggesting targeted allocation may be preferable.
Older veterans face structural, usability, and age‑bias barriers to video telehealth
Formative ethnography observed 20 veterans ≥65 during pharmacist video visits, documenting previsit and in‑visit barriers. Common problems included broadband/device access and audio‑video setup difficulties requiring technical support. Investigators identified 'digital ageism'—internalized age expectations and surprised pride after success—distinct from access issues. Implication: design telepharmacy trials to address access, usability, and age‑related expectations to improve recruitment and engagement.
Swiss EHR model modestly discriminates but poorly calibrates for 30‑day readmission
Retrospective cohort of 9,429 discharges age ≥65 developed a logistic model predicting unplanned 30‑day readmission (19.6% event rate). Predictors included polypharmacy (OR 1.83), home care (OR 1.61), male sex (OR 1.42), comorbidity count (OR 1.12/diagnosis), shorter stay, and younger age. AUC was 0.716 but calibration was poor and sensitivity at a 10% threshold was 7.5%, limiting clinical utility. Recommendation: do not deploy this model for automated interventions until externally validated and recalibrated.
References
Numbered in order of appearance. Click any reference to view details.
Additional Reads
Optional additional studies from this edition.