Push-Alert Notification of Troponin Results to Physician Smartphones Reduces the Time to Discharge Emergency Department Patients: A Randomized Controlled Trial

Ann Emerg Med. 2017 Sep;70(3):348-356. doi: 10.1016/j.annemergmed.2017.03.021. Epub 2017 May 9.


Study objective: For emergency department (ED) patients with chest pain, discharge decisions often hinge on troponin results. Push-alert notifications deliver results immediately to physician smartphones. Our objective is to determine whether troponin push alerts improve the time to discharge decisions for ED patients with chest pain.

Methods: In an academic ED, we assessed the effect of a quality improvement initiative using troponin push alerts to physician smartphones, with a cluster-randomized evaluation. Participating physicians were randomized to receive troponin push alerts (intervention) or not receive them (control). We retrospectively identified patients treated by participating physicians during the study period who were discharged from the ED with chest pain. The primary outcome was the time from final troponin result to discharge decision. Secondary outcomes included length of stay. A linear mixed model was used to adjust for physician clustering.

Results: During the study, 1,554 patients were discharged from the ED with chest pain. There were 551 patients in the control group and 554 in the intervention group who met inclusion criteria. The overall median interval from final troponin result to discharge decision was 79.7 minutes (interquartile range [IQR] 33.6 to 167.8 minutes); it was 94.3 minutes (IQR 36.2 to 177.8 minutes) in the control group and 68.5 minutes (IQR 30.5 to 157.2 minutes) in the intervention group. This 25.8-minute difference in medians (95% confidence interval 24.6 to 28.0 minutes) was statistically significant. Total ED length of stay was 345 minutes (IQR 261 to 419 minutes) in the control group and 328 minutes (IQR 250 to 408 minutes) in the intervention group.

Conclusion: Physicians who received troponin push alerts discharged their patients with chest pain 26 minutes faster than those without troponin notifications. Total ED length of stay did not significantly improve for these patients.

Publication types

  • Randomized Controlled Trial

MeSH terms

  • Adult
  • Aged
  • Biomarkers / blood
  • Canada
  • Chest Pain / blood*
  • Chest Pain / physiopathology
  • Emergency Service, Hospital* / organization & administration
  • Female
  • Humans
  • Length of Stay*
  • Male
  • Middle Aged
  • Patient Discharge*
  • Physicians
  • Point-of-Care Systems*
  • Quality Improvement / organization & administration
  • Retrospective Studies
  • Sensitivity and Specificity
  • Severity of Illness Index
  • Smartphone / statistics & numerical data*
  • Time Factors
  • Troponin I / blood*


  • Biomarkers
  • Troponin I