Predictive Performance of two Measures of Prognostic Mortality of Cancer Patients in Intensive Care Unit in Jordan: A Comparative Single-Centre Study

Outcome prediction for critical care patients with respiratory neoplasms using a multilayer perceptron neural network

OBJECTIVE

The variation in mortality rates of intensive care unit oncological patients may imply that clinical characteristics and prognoses are very different between specific subsets of patients with cancer. The specific characteristics of patients with cancer have not been included as risk factors in the established severity-of-illness scoring systems and comorbidity scores, showing limitations in predicting mortality risk. This study aimed to devise a predictive tool for in-hospital mortality for adult patients with a respiratory neoplasm admitted to the intensive care unit, using an artificial neural network.

METHODS

A total of 1,221 stays in the intensive care unit from the Beth Israel Deaconess Medical Center were studied. The primary endpoint was the all-cause in-hospital mortality prediction. An artificial neural network was developed and compared with six severity-of-illness scores and one comorbidity score. Model building was based on important predictors of lung cancer mortality, such as several laboratory parameters, demographic parameters, organ-supporting treatments, and other clinical information. Discrimination and calibration were assessed.

RESULTS

The AUROC for the multilayer perceptron was 0.885, while it was <0.74 for the conventional systems. The AUPRC for the multilayer perceptron was 0.731, whereas it was ≤0.482 for the conventional systems. The superiority of multilayer perceptron was statistically significant for all pairwise AUROC and AUPRC comparisons. The Brier Score was better for the multilayer perceptron (0.109) than for OASIS (0.148), SAPS III (0.163), and SAPS II (0.154).

CONCLUSION

Discrimination was excellent for multilayer perceptron, which may be a valuable tool for assessing critically ill patients with lung cancer.

Prognostic Value of Procalcitonin, C-Reactive Protein, and Lactate Levels in Emergency Evaluation of Cancer Patients with Suspected Infection

Simple Summary

Cancer patients are at increased risk of infections and related complications, including sepsis. We developed a scoring system for mortality prediction based on readily available clinical and laboratory data, including the quick sequential organ failure assessment (qSOFA) score, cancer subtype, and several laboratory markers (procalcitonin, C-reactive protein, lactate dehydrogenase, and albumin) that can be used in emergency departments for cancer patients with suspected infection. The prediction score, which stratifies patients into four different risk groups (from low risk to very high risk), achieved excellent performance in predicting 14-day mortality, with an area under the receiver operating characteristic curve value of 0.88 (95% confidence interval 0.85–0.91). The score was also effective in predicting intensive care unit admission and 30-day mortality.

Abstract

Cancer patients have increased risk of infections, and often present to emergency departments with infection-related problems where physicians must make decisions based on a snapshot of the patient’s condition. Although C-reactive protein, procalcitonin, and lactate are popular biomarkers of sepsis, their use in guiding emergency care of cancer patients with infections is unclear. Using these biomarkers, we created a prediction model for short-term mortality in cancer patients with suspected infection. We retrospectively analyzed all consecutive patients who visited the emergency department of MD Anderson Cancer Center between 1 April 2018 and 30 April 2019. A clinical decision model was developed using multiple logistic regression for various clinical and laboratory biomarkers; coefficients were used to generate a prediction score stratifying patients into four groups according to their 14-day mortality risk. The prediction score had an area under the receiver operating characteristic curve value of 0.88 (95% confidence interval 0.85–0.91) in predicting 14-day mortality. The prediction score also accurately predicted intensive care unit admission and 30-day mortality. Our simple new scoring system for mortality prediction, based on readily available clinical and laboratory data, including procalcitonin, C-reactive protein, and lactate, can be used in emergency departments for cancer patients with suspected infection.