Follow-up helps only if completed

Predischarge scheduling of early primary care follow-up did not significantly reduce 90-day readmission or death after hypoxemic pneumonia hospitalization.
*Retrospective target trial emulation cohort; Level 2b (OCEBM).

Citation

Davis AJ, Taylor SP, Hough CL, et al. Scheduling early primary care follow-up after pneumonia: a retrospective target trial emulation in five hospitals. Journal of Hospital Medicine. 2026;21:971-981. doi:10.1002/jhm.70281

Background

Hospitalized community-acquired pneumonia is common and often followed by readmission, disability, or death. Guidelines do not clearly recommend discharge follow-up strategies for these patients.

Patients

Adults hospitalized in 5 hospitals from 2019 to 2024 with community-acquired pneumonia requiring at least 2 liters of oxygen and discharged home. Exclusions included hospice discharge, discharge somewhere other than home, residence outside the service area, no in-system primary care in the prior year, and repeat qualifying admissions.

Intervention

Primary care follow-up scheduled before discharge to occur within 14 days.

Control

No primary care follow-up scheduled before discharge within 14 days.

Outcome

Primary outcome: composite of 90-day death or hospital readmission. Secondary outcomes included death, readmission, and emergency or urgent care visits.

Follow-up Period

90 days after discharge.

Results

Analysis Outcome Effect
Primary matched analysis 90-day death or readmission NS
Post-hoc completed-follow-up analysis 90-day death or readmission ATT −7.0% (95% CI −13.9 to −0.1); NNT 15
Post-hoc completed-follow-up analysis 90-day readmission ATT −7.2% (95% CI −14.1 to −0.1); NNT 14
Post-hoc completed-follow-up analysis Emergency visit without readmission ATT 5.0% (95% CI 0.1 to 9.9); NNH 20

ATT: average treatment effect among treated patients; NNT: number needed to treat; NNH: number needed to harm; NS: not statistically significant.

The primary analysis estimated the effect of scheduling, not attendance. The post-hoc completed-follow-up analysis was per-protocol-like and more vulnerable to bias.

Limitations

This was not randomized, so unmeasured differences between groups may remain. Readmissions and visits outside the health system may have been missed. Results apply mainly to patients already connected to in-system primary care. The completed-follow-up finding may be biased because healthier or more supported patients may have been more likely to attend.

Funding

National Heart, Lung, and Blood Institute; no company funding reported.

Clinical Application

Consider arranging early follow-up, but evidence is insufficient to mandate it; benefit may depend on attendance in higher-risk, primary-care-linked patients.