Predicting the clinical trajectory in critically ill patients with sepsis: a cohort study

Abstract

Background: To develop a mathematical model to estimate daily evolution of disease severity using routinely available parameters in patients admitted to the intensive care unit (ICU).

Methods: Over a 3-year period, we prospectively enrolled consecutive adults with sepsis and categorized patients as (1) being at risk for developing (more severe) organ dysfunction, (2) having (potentially still reversible) limited organ failure, or (3) having multiple-organ failure. Daily probabilities for transitions between these disease states, and to death or discharge, during the first 2 weeks in ICU were calculated using a multi-state model that was updated every 2 days using both baseline and time-varying information. The model was validated in independent patients.

Results: We studied 1371 sepsis admissions in 1251 patients. Upon presentation, 53 (4%) were classed at risk, 1151 (84%) had limited organ failure, and 167 (12%) had multiple-organ failure. Among patients with limited organ failure, 197 (17%) evolved to multiple-organ failure or died and 809 (70%) improved or were discharged alive within 14 days. Among patients with multiple-organ failure, 67 (40%) died and 91 (54%) improved or were discharged. Treatment response could be predicted with reasonable accuracy (c-statistic ranging from 0.55 to 0.81 for individual disease states, and 0.67 overall). Model performance in the validation cohort was similar.

Conclusions: This prediction model that estimates daily evolution of disease severity during sepsis may eventually support clinicians in making better informed treatment decisions and could be used to evaluate prognostic biomarkers or perform in silico modeling of novel sepsis therapies during trial design.

Clinical trial registration: ClinicalTrials.gov NCT01905033.

Keywords: Epidemiology; Intensive care unit; Markov model; Organ failure; Outcome; Sepsis.

Conflict of interest statement

Not applicable.

Figures

Fig. 1

Proposed Markov model showing all possible transitions. The arrows represent forward or backward progression between transitional (disease severity) states, as well as to the final absorbing states death or discharge. The probabilities of advancing to a more advanced stage or regressing to a less severe stage or to an absorbing state are calculated by the multi-state Markov model with piecewise constant intensities. Forty-three out of a total of 3855 transitions (1%) were from “at risk” directly to “failure” or death or from “failure” directly to “at risk” or discharge and were not estimated due to the insufficient number of events

Fig. 2

Flowchart of patient inclusion with patient disposition at admission

Fig. 3

Modeled incidences of organ failure, death, and discharge in three illustrative patients. Patient 1 is a 72-year-old immunocompromised male admitted for a community-acquired pneumonia with mild hypoxemia (60% oxygen mask), a lactate level of 0.5 mg/L and a C-reactive protein level of 153 mg/L upon presentation. He has an absolute risk for discharge alive of 58% and death of 22% at day 14. Patient 2 represents another (but similar) patient with a community-acquired pneumonia in acute respiratory distress (requiring prompt intubation), hypotension (requiring norepinephrine), mottled skin, oliguria, lactate 4.2 mg/L, and a C-reactive protein of 268 mg/L. He has a risk for discharge alive of 36% and death of 40% at day 14. Patient 3 is a 53-year-old previously healthy female patient with a urinary tract infection, lactate of 0.4 mg/L, and a C-reactive protein of 50 mg/L. She has a probability of discharge alive of 79% and a probability of death of 5% at day 14

Fig. 4

Outcome of patients who improve or worsen over time. Patient 4 is a 59-year-old male patient admitted for a severe peritonitis requiring noradrenaline at a rate of 0.05 μg/kg/min, a lactate level of 5.6 mmol/L, and a C-reactive protein level of 256 mg/L. At day 3, the noradrenaline can be stopped, his lactate levels are 0.5 mmol/L, and his C-reactive protein levels decrease to 170 mg/L (indicated by “improvement”), and at day 7, C-reactive protein levels dropped to 50 mg/L. However, if the same patient would develop refractory shock and atrial fibrillation at day 3, his outcome is as shown by “worsening”; at day 7, he develops an ICU-acquired pneumonia but noradrenalin is stopped, showing the net positive effect of worsening (pneumonia) and improvement (stopping of noradrenalin)