This Mentored Clinical Scientist Development Award proposal describes the 5-year training program for Dr. Elliott R. Haut. The proposal builds upon the candidate's strengths and prior research skills, while utilizing the unique opportunities of The Johns Hopkins University. The mentorship of Dr. Peter Pronovost and formal advanced degree coursework in the Graduate Training Program in Clinical Investigation (GTPCI), will give Dr. Haut the tools and training to advance to the level of a fully independent health services researcher. Pay for performance and public reporting of quality are growing exponentially in medicine. Implicit in these systems is the assumed presence of a valid way to measure clinical quality. Many national agencies (AHRQ, NQF) have suggested deep vein thrombosis (DVT) incidence as a metric of health care quality. DVT is a significant cause of morbidity and mortality in trauma patients, yet may be completely asymptomatic until the clot embolizes and causes sudden death from massive pulmonary embolism. Duplex ultrasound serves an important role as a noninvasive diagnostic tool for DVT detection. However, screening asymptomatic trauma patients for DVT is controversial, these practices likely vary widely among hospitals, and no national agency has recommended a standardized screening approach. In the absence of standardized surveillance, DVT rates may be influenced more by how often caregivers look for these events rather than the quality of care provided. This is a classic example of surveillance bias in which event rates identified and reported may merely reflect levels of vigilance. Higher DVT rates may not be a marker of poor quality of care, but rather an outcome of more aggressive screening practices. Providers who look harder by ordering more screening duplex ultrasounds, may report higher rates of DVT, and appear to provide poorer quality of care than those providers who do not screen. Therefore, DVT rates reported may be unreliable measures of quality of care and should not be used to compare hospitals or for pay-for-performance initiatives. SA#1 will examine the impact of hospital-level ultrasound rates on DVT rates reported after major trauma. SA#2 will determine whether hospital duplex and DVT rates correlate with the presence of a hospital duplex ultrasound screening protocol for high-risk trauma patients. SA#3 will determine if patients are more likely to have DVT diagnosed and reported based on hospital characteristics, controlling for patient-level DVT risk factors using multi-level multiple logistic regression.
Some hospitals use duplex ultrasound to screen asymptomatic trauma patients for deep vein thrombosis (DVT) while other hospitals do not. Hospitals that screen trauma patients may identify and report higher DVT rates merely because they search for DVTs more aggressively.Therefore, reported DVT rates are unreliable quality of care measures and should not be used to compare hospitals or for pay-for-performance initiatives.
Showing the most recent 10 out of 25 publications
