Purpose: The effectiveness of acute geriatric units in improving outcomes of hospitalized seniors in the real world is unclear. We sought to answer this question by focusing on pneumonia. We hypothesized that acute geriatric units reduce short-term mortality for seniors hospitalised for pneumonia compared with those receiving usual internal medicine care.
Method: We merged medical records and administrative data of adults aged 65 years or older admitted to 3 acute care hospitals over one year. The outcome variable was 30-day mortality. The treatment variable was admission to acute geriatric units, with usual internal medical care as the reference. Other explanatory variables included demography, admission information, severity of acute illness (CURB score), co-morbidity, and functional status. We obtained propensity scores for admission to acute geriatric units and stratified seniors into quintiles according to scores sorted in ascending order. We performed logistic regression and propensity score matching (PSM) to estimate treatment effects for all seniors and for those within quintiles. Finally, we explored distribution of explanatory variables across quintiles.
Result: The 30-day mortality for 3034 seniors included for analyses was 25.8%. There was a significant reduction in 30-day mortality for senior admitted to acute geriatric units (adjusted OR 0.77; 95%CI 0.60 to 0.99) using logistic regression analyses. The following table shows 30-day mortality and selected patient characteristics across quintiles.
| Quintile 1 (n=607) | Quintile 2 (n=607) | Quintile 3 (n=606) | Quintile 4 (n=607) | Quintile 5 (n=607) | |
| Treatment effects: | |||||
| Acute geriatric units: OR (95% CI) | 0.82 (0.35-1.91) | 1.16 (0.53-2.55) | 0.97 (0.60-1.59) | 0.66 (0.40-1.11) | 0.57 (0.36-0.90) |
| Patient characteristics: | |||||
| Age>80 years | 0.0% | 3.3% | 68.0% | 100.0% | 100.0% |
| Hospitalization in prior 90 days | 28.2% | 24.7% | 23.6% | 36.1% | 36.6% |
| CURB score>2 | 12.5% | 15.3% | 18.2% | 18.5% | 21.1% |
Conclusion: Acute geriatric units reduced short-term mortality among seniors hospitalized for pneumonia, when compared with usual internal medicine care. Seniors who were more likely to receive care at these units had greater mortality reduction. These findings have implications on targeting policies for these and similar units including acute care for elders (ACE) units.