Estimates and confidence intervals for mean postoperative health-care-associated infections per week among operating rooms and other anesthetizing locations
Authors' conclusion
Poisson regression with robust variance estimation resulted in more precise estimates than did the method of batch means, which will lower health care costs.
Study context
- Population
- Surgical cases in operating rooms and other anesthetizing locations at a large teaching hospital in Iowa, US, 2024 fiscal year75 rooms
- Participants
- 42,978 cases
- Exposure
- Poisson regression vs Method of batch means
- Outcome
- CI width for new postoperative health-care-associated infections per week per room
Methodological detailssingle-centre
- Single-centre
Main finding
As reported in the abstract
Poisson regression CIs narrower than batch means, 15 rooms with mean ≥ 0.50 infections per week
≈35%
CI width reduction, all rooms pooled
≈38%
Equivalent sample size increase from 1.00 year to (15 rooms; pooled)
2.35 and 2.60 years
Equivalent cost reduction (15 rooms; pooled)
88% or 92%
Observational data: associations, not effects.
Further findings
Cases with postoperative infection · 90 days
- Cases with patients who developed postoperative infections
- 1,095/42,978 (2.6%)
Infections per room day (27,450 room days)
- No infection
- 96.2%
- One infection
- 3.7%
- Two infections
- 0.2%
- Three infections
- 0.01%
Serial correlation and dependence between cases
- Preceding infected case did not significantly increase probability of next case infected, all 75 Šidák-corrected
- P ≥ 0.32
- No significant serial correlation between days by room or between rooms by day
Fit to Poisson distribution
- Rooms with counts consistent with Poisson distributions
- 73/75
- Variance-to-mean ratios of the other two rooms
- 1.32 and 1.78
Original publication
Composed by AI from the published abstract and checked against it. How Briefs are made →