There were 18,690,469 eligible hospitalizations involving an ICU stay during the study period. The overall number of Medicare ICU admissions declined each year, from 1,901,630 in 1997 to 1,637,581 in 2006 (). Yet during that time, the absolute number of long-term acute care transfers, as well as long-term acute care transfers as a proportion of all ICU discharges, steadily increased. Critical care hospitalizations ending in a long-term acute care transfer increased from 13,732 (0.7%) in 1997 to 40,353 (2.5%) in 2006 (p<0.001). During that time, the percentage of critical care hospitalizations ending in transfer to skilled nursing or rehabilitation facility also increased, while the percentage of critical care hospitalizations ending in discharge to home fell. Mortality rates were relatively constant. Examining patients receiving invasive mechanical ventilation separately (n=2,380,881; 12.7% of total), ICU hospitalizations ending in a long-term acute care transfer increased from 7,126 of 217,514 critical care hospitalizations (3.3%) in 1997 to 19,781 of 227,152 critical care hospitalizations (8.7%) in 2006 (p<0.001). During that time period, the percentage of critical care hospitalizations involving mechanical ventilation ending in discharge to home fell, while other discharge dispositions (skilled nursing or rehabilitation facility, death) remained relatively constant.
Temporal trends in discharge location after critical illness, both for all patients and for patients receiving invasive mechanical ventilation.
The increase in the number of transfers was reflected in an increase in the number of long-term acute care hospitals, long-term acute care beds and long-term acute care-associated costs following critical illness (). The number of long-term acute care hospitals increased at a mean rate of 8.8% per year, from 192 in 1997 to 408 in 2006 (p-value for linear trend <0.001). The number of long-term acute care hospital beds increased from 16,523 in 1997 to 27,623 in 2006 (mean rate of increase: 5.9% per year, p<0.001) while the total associated costs increased from $484 million in 1997 to $1.325 billion in 2006 (mean rate of increase: 12.1%, p<0.001).
Temporal trends in long-term acute care hospitals (panel A), long-term acute care beds (panel B) and long-term acute care -associated costs after critical illness (panel C).a
We observed similar increases in the age-adjusted incidence of long-term acute care hospital transfer at the population level (). Transfers after critical illness increased from 38.1 per 100,000 capita in 1997 to 99.7 per 100,000 capital in 2006 (, p<0.001). The age-adjusted incidence of transfers was higher for male individuals () and black individuals () in all time periods (p <0.001). Race differences were particularly large, with black individuals consistently experiencing over twice the transfer rate as white individuals. For example, in 2006 the age-adjusted transfer rate was 182.0 per 100,000 for black individuals compared to 89.6 per 100,000 for white individuals. The rate of increase was not different by gender (p-value for interaction term: 0.60) or race (p-value for interaction term: 0.10).
Kaplan-Meier curves for one-year survival after transfer to a long-term acute care hospital, both for all patients (panel A) and for patients by ventilation status at originating hospital and long-term acute care hospital (panel B).
Patient characteristics over time are shown in . Age, gender, and racial distribution and length of stay prior to transfer varied little over the study period. However, the incidence of co-morbid conditions, the incidence of sepsis at the originating hospital, and the percentage of patients receiving mechanical ventilation at the long-term acute care hospital all increased. Additionally, the final discharge destination after transfer changed over time, with more patients discharged to a skilled nursing or rehabilitation facility (19.9% in 1997–2000, 34.9% in 2004–2006, p<0.001), and fewer patients discharged to home (32.3% in 1997–2000, 27.4% in 2004–2006, p<0.001). Nearly 20% of patients were transferred back to an acute care facility, a figure that decreased over time (19.6% in 1997–2000, 14.2% in 2004–2006, p<0.001).
One year mortality was poor, ranging from 48.2% to 52.2% over the study period. Compared to other patients, lower survival was observed for patients receiving mechanical ventilation at the long-term acute care hospital (Figure 4, data for 2006 only, n=38,423). One year mortality was 46.2% for patients never receiving mechanical ventilation (n=18,660), 48% for patients receiving mechanical ventilation in the short stay hospital only (n=8,068), and 69.1% for patients receiving mechanical ventilation in both the short stay hospital and the long-term acute care hospital (n=11,695) (p<0.001 using log-rank test).