Hitting the Vasopressor Ceiling: Finding Norepinephrine Associated Mortality in the Critically Ill

Explainable time-series deep learning models for the prediction of mortality, prolonged length of stay and 30-day readmission in intensive care patients

Background

In-hospital mortality, prolonged length of stay (LOS), and 30-day readmission are common outcomes in the intensive care unit (ICU). Traditional scoring systems and machine learning models for predicting these outcomes usually ignore the characteristics of ICU data, which are time-series forms. We aimed to use time-series deep learning models with the selective combination of three widely used scoring systems to predict these outcomes.

Materials and methods

A retrospective cohort study was conducted on 40,083 patients in ICU from the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database. Three deep learning models, namely, recurrent neural network (RNN), gated recurrent unit (GRU), and long short-term memory (LSTM) with attention mechanisms, were trained for the prediction of in-hospital mortality, prolonged LOS, and 30-day readmission with variables collected during the initial 24 h after ICU admission or the last 24 h before discharge. The inclusion of variables was based on three widely used scoring systems, namely, APACHE II, SOFA, and SAPS II, and the predictors consisted of time-series vital signs, laboratory tests, medication, and procedures. The patients were randomly divided into a training set (80%) and a test set (20%), which were used for model development and model evaluation, respectively. The area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and Brier scores were used to evaluate model performance. Variable significance was identified through attention mechanisms.

Results

A total of 33 variables for 40,083 patients were enrolled for mortality and prolonged LOS prediction and 36,180 for readmission prediction. The rates of occurrence of the three outcomes were 9.74%, 27.54%, and 11.79%, respectively. In each of the three outcomes, the performance of RNN, GRU, and LSTM did not differ greatly. Mortality prediction models, prolonged LOS prediction models, and readmission prediction models achieved AUCs of 0.870 ± 0.001, 0.765 ± 0.003, and 0.635 ± 0.018, respectively. The top significant variables co-selected by the three deep learning models were Glasgow Coma Scale (GCS), age, blood urea nitrogen, and norepinephrine for mortality; GCS, invasive ventilation, and blood urea nitrogen for prolonged LOS; and blood urea nitrogen, GCS, and ethnicity for readmission.

Conclusion

The prognostic prediction models established in our study achieved good performance in predicting common outcomes of patients in ICU, especially in mortality prediction. In addition, GCS and blood urea nitrogen were identified as the most important factors strongly associated with adverse ICU events.

The Association of Norepinephrine Utilization With Mortality Risk in Trauma Patients

INTRODUCTION

While the pillars of trauma resuscitation are surgical hemostasis and blood product administration, norepinephrine (NE) can be used as an adjunct. The goal of this study was to evaluate the relationship between the maximum dose of NE, timing of NE administration, and mortality in trauma patients.

METHODS

Patients admitted between January 2013 and January 2021 treated with NE were reviewed. Univariate and multivariate logistic regression were used to assess whether maximum NE dose was independently associated with mortality. Optimal dosage rates for NE were determined via Youden Index. Subgroup analyses comparing those who received NE within versus after the first 24 h of admission were conducted.

RESULTS

Three hundred fifty-first trauma patients were included, with 217 (62%) surviving. Patients who died received an average maximum dose of 16.7 mcg/min compared to 9.1 mcg/min in survivors (P = 0.0003). Mortality rate increased with dosage (P < 0.0001), with doses greater than 20 mcg/min having 79% mortality. Those who received NE within the first 24 h had an inflection point in mortality at 16 mcg/min (Youden = 0.45) (OR 1.06; 95% CI 1.03-1.10). For patients who received NE after the first 24 h, an inflection point in mortality was at 10 mcg/min (Youden = 0.34) (OR 1.09; 95% CI 1.04-1.14).

CONCLUSIONS

Higher maximum doses of NE were associated with increased mortality. Patients initiated on NE more than 24 h into their admission displayed an inflection point at a lower dose than those initiated later. This suggests that trauma patients initiated on NE after 24 h from injury may have a dire prognosis.

Correlation and Prognostic Assessment of Low T3 Syndrome and Norepinephrine Dosage for Patients with Sepsis: A Retrospective Single-Center (Cohort) Study

Purpose

To investigate the correlation and prognostic significance of low triiodothyronine (T3) syndrome and norepinephrine dosage in patients with sepsis and septic shock.

Methods

This single-center, retrospective, cohort study enrolled 169 patients with sepsis and septic shock that were admitted to the intensive care unit of First Hospital of Nanchang, Nanchang, China from June 2017 to July 2019. All included patients were followed up for 28 days or died, whichever was earlier. Patients with free T3 (FT3) of <3.1 pmol/L were considered with low T3 syndrome. The correlation and prognostic significance of the FT3 and maximum dosage of norepinephrine (MDN) within 72 h, as well as other clinical indicators, were analyzed by using correlation analysis, principal component analysis, receiver operating characteristic curve, Youden index, and logistic regression.

Results

A total of 138 patients were allocated to the low T3 group. FT3 inversely correlated with the Sequential Organ Failure Assessment (SOFA) score within 24 h, fluid resuscitation volume within 24 h, and lactic acid levels, and positively correlated with the mean arterial pressure. The critical values of age, SOFA, and MDN for predicting the 28-day mortality were 79.5 years, 8.5 points, and 0.61 µg/kg/min, respectively. The mortality of the low T3 and normal T3 groups was similar. Considering the MDN of 0.61 µg/kg/min as the cutoff value, the mortality between the two groups was significantly different.

Conclusion

Among patients with sepsis and septic shock, FT3 was inversely correlated with the disease severity. An MDN ≥ 0.61 µg/kg/min within 72 h may be an important prognostic indicator.