Introduction: As healthcare administrators focus on patient safety and cost-effectiveness, methodical assessment of quality outcome measures is critical. In 2008, the Centers for Medicare and Medicaid Services (CMS) published a series of “Never Events” which included 11 hospital acquired conditions (HACs), for which related costs of treatment are not reimbursed. Cerebrovascular procedures (CVPs) are complex and often performed on patients with significant medical comorbidities.
Methods: This study examines the impact of patient age and medical comorbidities on the occurrence of CMS-defined HACs, as well as their effect on the length of stay (LOS) and hospitalization charges in patients undergoing common CVPs.
Results: HACs occurred at a frequency of 0.49% (1.33% in the intracranial procedures and 0.33% in the carotid procedures). Falls/Trauma (n=4,610, 72.3% HACs, 357 HACs per 100,000 CVPs) and Catheter Associated Urinary Tract Infections (n=714, 11.2% HACs, 55 HACs per 100,000 CVPs) were the most common events. Age and presence of comorbidities were strong independent predictors of HACs (p<0.01). HAC occurrence negatively impacts both LOS and hospital costs. Patients with at least one HAC were ten times more likely to have prolonged LOS (>90th percentile) (p < 0.01), and seven times more likely to have high inpatient costs(>90th percentile) (p < 0.01) when adjusting for patient and hospital factors.
Conclusions: Improved quality protocols focused on individual patient characteristics might help to decrease the frequency of HACs in this high-risk population. This data suggests that risk-adjustment according to underlying patient factors may be warranted when considering reimbursement for costs related to HACs in the setting of CVPs.
Patient Care: Providing a patient subgroup that may benefit from risk adjustment due to high association with Hospital acquired conditions
Learning Objectives: Understanding cost and length of stay impact of hospital acquired conditions
Consideration of patient age and comorbidities on potential HAC occurrence