Associations of Government-issued Intensive Care Unit Admission Criteria with Clinical Practices, Outcomes, and Intensive Care Unit Bed Occupancy

Rationale: In Japan, the government officially issued intensive care unit (ICU) admission criteria that require ICUs to admit patients who need a certain level of monitoring and procedures to ensure their reimbursement for ICU costs from April 2014. Objectives: To assess whether the newly issued health policy on ICU admission criteria based on financial incentives for monitoring and procedures had any impact on clinical practices, outcomes, and ICU bed occupancy. Methods: Using a nationwide inpatient health claims database in Japan, we identified patients who were admitted to the ICU from April 2012 to March 2018. Outcomes were monitoring and procedures in the ICU, clinical outcomes, and ICU bed occupancy. The outcomes of monitoring and procedures and clinical outcomes were adjusted for patient characteristics. Interrupted time-series analyses were used to compare the trends in outcomes for two separate periods before and after the issue of the new health policy on ICU admission criteria in April 2014. Results: A total of 1,660,601 patients in 259 ICU-equipped hospitals were eligible. There were significant upward slope changes between the pre- and post-issue periods for all monitoring and procedures in the ICU, including invasive arterial pressure monitoring (5.62% change in trend per year; 95% confidence interval [CI], 4.75-6.49%) and central venous pressure monitoring (1.22% change in trend per year; 95% CI, 0.78-1.67%). There was no significant slope change between the pre - and post-issue periods for in-hospital mortality, but there were significant upward slope changes for complications of pneumonia (0.27% change in trend per year; 95% CI, 0.14-0.39%) and catheter-related bloodstream infection (0.02% change in trend per year; 95% CI, 0.00-0.14%). There were also significant upward slope changes in length of hospital stay, length of ICU stay, and hospitalization costs between the pre- and post-issue periods. There was no significant slope change between the pre- and post-issue periods for ICU bed occupancy. Conclusions: The health policy on ICU admission criteria based on financial incentives for actions taken by providers was associated with increased monitoring and procedures, complications, lengths of hospital and ICU stay, and hospitalization costs without decreasing ICU bed occupancy.