# Predictors of Intensive Care Unit Outcomes Among Oncology Patients: A Retrospective Cohort Study at a Tertiary Referral Center in Saudi Arabia

## TL;DR

This study identifies factors that predict ICU mortality in cancer patients in Saudi Arabia, emphasizing the importance of early risk identification for better clinical decisions.

## Contribution

The study provides novel regional data on ICU outcomes for oncology patients in southwestern Saudi Arabia.

## Key findings

- ICU mortality was 38.1%, primarily due to cancer progression.
- Older age, advanced-stage disease, respiratory infections, and higher SOFA scores were independent predictors of mortality.
- Advanced-stage patients had significantly lower 14-day survival compared to non-advanced stage patients.

## Abstract

Background

Caring for critically ill cancer patients is challenging due to the complexity of malignancy combined with acute organ dysfunction. Despite an increasing cancer burden in Saudi Arabia, data on intensive care unit (ICU) outcomes, particularly in the Najran region, remain limited. This study aims to identify clinical characteristics, ICU treatments, and mortality predictors in this population.

Patients and methods

This retrospective cohort study included 105 adult patients with active solid or hematologic malignancies admitted to the ICU at King Khalid Hospital, Najran, between January 2014 and February 2023. Data, including demographics, cancer details, ICU admission causes, treatments, and outcomes, were collected. Multivariate logistic regression was used to assess mortality predictors, and Kaplan-Meier analysis compared survival by disease stage.

Results

The cohort had a mean age of 56.8 ± 16.5 years, with 57 females (54.3%). Advanced-stage disease was present in 84 patients (80.0%), with gastrointestinal cancers being the most common primary site (35, 33.3%). Cancer-related complications accounted for 77 ICU admissions (73.3%). The median Sequential Organ Failure Assessment (SOFA) score at admission was 1.0 (interquartile range (IQR) 0.0-3.0). ICU mortality was 40 patients (38.1%), primarily due to cancer progression. Independent mortality predictors included older age (adjusted odds ratio (aOR) 1.04 per year; 95% CI 1.01-1.06), advanced-stage disease (aOR 3.37; 95% CI 1.56-7.26), respiratory infections (aOR 2.19; 95% CI 1.04-4.59), and higher SOFA scores (aOR 1.70 per point; 95% CI 1.28-2.27). The predictive model demonstrated good discrimination (optimism-corrected area under the curve (AUC) = 0.82). Kaplan-Meier analysis showed significantly lower 14-day survival in advanced-stage patients (41%) compared to non-advanced stage patients (78%; log-rank p < 0.001).

Conclusions

This study highlights high ICU mortality among cancer patients at a tertiary center in southwestern Saudi Arabia, particularly those with older age, advanced-stage disease, respiratory infections, and higher SOFA scores. Early identification of these high-risk predictors is essential to guide clinical decision-making, enhance infection control, and facilitate timely palliative care integration. These findings provide valuable regional data to inform critical care strategies for oncology patients in similar settings.