Candidate for the Lee B. Lusted Student Prize Competition
Purpose: Patients with an active malignancy and septic shock presenting to the intensive care unit (ICU) often have a high mortality despite aggressive supportive care. As a result, some investigators have postulated that a 72-hour “trial of ICU” can accurately differentiate survivors from non-survivors. This recommendation is, however, based on clinical gestalt and has never been studied.
Methods: The primary objectives of this study were to determine 1) if a trial of ICU offers similar overall life expectancy compared to unlimited aggressive care, and 2) to define the optimal length of time for a trial of ICU. The secondary objective was to determine whether a trial of ICU is more cost effective than aggressive care.
A five-state stochastic model was built to simulate the clinical problem, with 3 possible strategies: aggressive care, ICU trial, or comfort-measures-only as a comparison arm. 939 patients with advanced cancer and septic shock were identified from the Multiparameter Intelligent Monitoring in Intensive Care (MIMIC) II at the Beth Israel Deaconess Hospital. Daily transition probabilities, as well as probabilities of improvement or worsening disease, were determined using Kaplan-Meier time-to-event analysis on this patient population, conditional on their severity of illness database.
Benefit was defined as average days lived out of 30, and costs were calculated in 2011 US dollars. The model was validated internally as well as externally against two other ICU databases, one in Boston and one in Riyadh, Saudi Arabia.
Results: Modeled survival analysis did not differ significantly from actual survival analyses of patients in the MIMIC database or the external validation databases. Trial lengths that afforded the same probability of 30-day survival as aggressive care varied by illness severity. In the healthiest half of patients, trial lengths were clinically indistinguishable from aggressive care. Even in the sickest patients, optimal trial length was 5 days, longer than the current clinical recommendations (see figure). With costs considered, however, aggressive care became the desired solution only above WTP values of $1,100,00/life saved
Conclusions: Aggressive care for cancer patients with septic shock leads to longer overall survival compared to a trial of ICU, but at significant cost. Optimal trial of ICU varies with clinical severity but, for all patients, is longer than current recommendations.
Figure: Optimal trial lengths by disease severity
|
|
See more of: The 35th Annual Meeting of the Society for Medical Decision Making