Potential for cost-shifting among patients in critical care settings under the regulated and competitive hospital reimbursement systems

This paper investigates the potential for cost-shifting under the competitive hospital reimbursement system in New York State among patients undergoing mechanical ventilation in critical care settings and discharged under DRG 483, mechanical ventilation with tracheostomy, and DRG 475, mechanical ventilation with a respiratory diagnosis. The study determines whether the continuing reduction in hospital length of stay combined with the increased propensity for discharges to long term care settings among these patients is due to self-selection or the reduction in care in the inpatient setting. Discharge abstracts from the Statewide Planning and Research Cooperative System are used to identify clinical and hospital-driven determinants of resource utilization and discharge location under the regulated system and after the institution of the first an second generations of the competitive hospital reimbursement system during the period 1995-2003. The competitive reimbursement methodology was instituted in 1997 under the Health Care Reform Act of 1996, and was renewed in 2000. Identity-linked geometric regression models are used to test hypotheses concerning the impacts of reimbursement structures on the likelihood of hospital survival, length of stay and likelihood of discharges to long term care settings among survivors. Cost-shifting is identified as a combined decrease length of stay and increased likelihood to other care settings, among the survivors. Any potential for cost-shifting is investigated for the hospital system as a whole. As well, characteristics of hospitals most likely to be induced to shift costs to post-acute settings are identified. Preliminary findings show that on average across the hospital system, the institution of competitive reimbursements resulted in a combined decrease length of stay and increase in the likelihood of discharge to long term care settings, controlling for clinical risk factors. However, the length of stay has remained longer on average for hospitals that are large, have a larger proportion of managed care patients or are located in the New York City area. These findings inform public and private policy concerning the impact of reimbursement changes on overall resource utilization for the health care system's sickest and most expensive patients.