Maine Medical Center, Medical Education, Maine Medical Center Research Institute, Center for Outcomes Research & Evaluation
Palliative Care, Humans, Hospitalization, Pulmonary Disease, Chronic Obstructive
Abstract Background. COPD is the 4th leading cause of death and is characterized by significant functional decline, particularly near the end of life. COPD patients have lower rates of advance care planning than patients with other respiratory illnesses. Currently, there is no effective prognostic tool for COPD patients in the inpatient setting. The surprise question (SQ) (“Would you be surprised if this patient died in the next time frame”) has shown to be effective at predicting mortality in some patient populations but has not been studied in COPD. Objective. We sought to assess the performance of the SQ in estimating mortality and advance care planning among patients with acute exacerbation of COPD.
Methods. We performed a retrospective observational analysis of patients admitted to Maine Medical Center for acute exacerbation of COPD between July 2015 and September 2018. Emergency department (ED) and inpatient clinicians answer the SQ when placing admitting orders. ED clinicians answered, “Would you be surprised if this patient died in the next 30 days?”, and inpatient clinicians answered, “Would you be surprised if this patient died in the next year?” Clinicians provided a binary response (“No, I would not be surprised” [SQ+] or “Yes, I would be surprised” [SQ-]). We used sensitivity, specificity, positive predictive value, negative predictive value, area under the receiver operator curve, and multivariate logistic regression to assess the performance of the SQ in estimating 30-day and one-year mortality. Our secondary outcome was to assess if the surprise question prompted goals of care conversations, palliative care consultation and/or completion of advance care planning documents.
Results. A total of 381 patients had responses to the 30-day SQ, and 365 patients had responses to the one-year SQ. Five percent (n = 19) of 30-day SQ responses were positive, and 30% (n = 108) of one-year SQ responses were positive. The 30-day SQ had the following test characteristics in estimating 30-day mortality: sensitivity 12%, specificity 95%, positive predictive value (PPV) 11%, and negative predictive value (NPV) 96%. The one-year SQ had the following test characteristics in estimating one-year mortality: sensitivity 47%, specificity 75%, PPV 35%, and NPV 83%. Patients with a positive one-year SQ were 2.75 times more likely to die within one year than those with a negative SQ (95% CI 1.59, 4.77, p < 0.000; area under receiver operator curve 0.69) and 3.01 times more likely to have received any advance care planning during admission (95% CI 1.88, 4.82, p < 0.000; area under receiver operator curve 0.75). Patients with a positive 30-day SQ were 3.83 times more likely to receive advance care planning during admission than those with a negative SQ (95% CI 1.25, 11.77, p = 0.019; area under receiver operator curve 0.79).
Conclusion. The one-year surprise question can be an effective component of prognostication and care planning for COPD patients in the inpatient setting.
Tripp, Dana; Janis, Jaclyn; and Hutchinson, Rebecca, "How well does the Surprise Question Predict 1-year Mortality for Patients Admitted with COPD" (2020). Costas T. Lambrew Research Retreat 2020. 98.
2020 Costas T. Lambrew Research Retreat