The Impact of a High Sensitivity Troponin HEART Pathway Based Clinical Decision Protocol on Observation Visits.

Background: Use of high-sensitivity troponin (hs-cTn) might lead to an increase in hospital observation visits due to higher number of abnormal troponin levels.

Study objectives: To determine the impact of incorporating hs-cTn into a chest pain clinical decision protocol (CDP) on observation visits in a large academic health system.

Methods: This is a retrospective observational cohort study of all chest pain observation patients in four hospitals in an academic health system over 24 months. All hospitals used the Beckman Coulter Unicel Dxi instrument, and all shared the same emergency department (ED) chest pain protocol, which used the HEART pathway and serial troponins and directed ED dispositions to either an observation stay, ED discharge, or inpatient admission. Outcomes studied before and after introduction of a hs-cTn protocol included daily chest pain observation census, cost, observation hours, and inpatient admit rate. Census was reported as the daily chest pain observation census and as a proportion of all observation visits. Data was retrieved from a health system data warehouse and a cost accounting program.

Results: There were 6,712 chest pain observation visits over 24-months, with 4,087 visits before and 2,634 visits after the hs-cTn protocol implementation. Comparison groups were similar in terms of age, gender, and type of insurance. There were 10.59 (95% CI: 10.24 - 10.95) daily chest pain observation visits before and 7.66 (95% CI: 7.34 - 7.97) visits after implementation, with a 28% (95% CI: 35% - 20%) decrease in the total daily census. As a portion of all observation visits, there was a 22% drop in the proportion that were observed for chest pain. The daily number of chest pain patients requiring inpatient admission was unchanged. The daily total direct cost for chest pain observation decreased with an effective daily cost savings of $4,313 USD (95% CI: $1,534 - $6,998). The total daily number of chest pain observation bed hours also decreased by 41.5 hours (95% CI 13.4 - 96.4 hr).

Conclusion: Implementation of a hs-cTn chest pain protocol was associated with a significant decrease in the number and proportion of observation visits, a decrease in total daily cost and bed hours used, and no increase in inpatient admissions.