There is little research about the factors predicting mortality and prolonged stay in ICU in critically ill patients with solid tumors, and it seems that no study in Saudi Arabia identified the factors influencing these outcomes in this subgroup. Knowledge on such factors is important in making appropriate patients’ selection and improving the quality of ICU in these patients. Thus, the main objectives of this study were to determine the characteristics and the outcomes of patients with solid tumors admitted to ICU in a tertiary hospital in Saudi Arabia, and to explore factors affecting mortality and prolonged stay in ICU.
Our ICU mortality in patients with solid malignancies was 32.4% which is higher than our overall ICU mortality 19.2%. Previous research in patients with solid tumors reported large variation (15% − 59%) in the ICU mortality rates with an average rate of 31.2% (7, 14–16). A possible explanation of this large variation is the characteristics of patient’s selection in those studies, the underlying malignancy (cancer type, course of malignancy, staging and treatment), causes of ICU admission and discharge, and therapeutic intervention decisions (1, 4–7, 17–19) as well as early DNR decisions that prevent ICU admissions of cases who will not benefit from ICU care. Our ICU mortality rate is slightly higher than the average rate reported in the previous research in patients with solid malignancy, and this might be explained by the composition of underlying malignancy in our patient’s population as the rate of metastatic cancer (43.4%) and the majority had progressive disease (72.5%).
Studies of patients with a solid malignancy reported that predicting ICU mortality with adequate details is needed to be of prognostic value to physicians and for proper selection of cases for ICU care (14, 16). We found many factors associated with ICU mortality (Table 2). However, only 3 independent risk factors were found to be predicting ICU mortality rate in multivariate analysis. These factors were SOFA score at ICU admission, use of mechanical ventilation and vasopressor. These predictors have been similar to those reported in cancer patients (7, 20). Although the severity of the acute illness score (APACHE II) reported to be a main predictor factor for the ICU mortality in non-cancer patients (7), both organ failure (SOFA) and severity of the acute illness (APACHE II) scores are useful in predicting ICU mortality in cancer patients generally (20–22). However, in accordance with the study done by Aygencel et al. (7), we demonstrated that only organ failure score (SOFA) was independent risk factor for the ICU mortality in patients with solid tumors. Although we found that the course and stage of malignancy related to ICU mortality, these results were not significant predictors for ICU mortality. The impact of the stage of malignancy, response to chemotherapy and other characteristics of cancer on short-term outcome remain controversial. Some studies have reported these characteristics affect ICU mortality (6, 17, 23, 24) while others reported little or no impact on such an outcome (7).
To our knowledge, this is the first research investigating prolonged ICU stay in patients with solid malignancies. Of the total patients, 13.7% had an ICU stay of 21 days or more days. Prolonged ICU stay is a known to be associate an increased risk of severe complications that could lead to mortality. In this study, during ICU admission 66% of the patients with prolonged ICU stay developed ICU acquired infections. In our univariate analysis, we found that several factors were related to prolonged ICU stay However, this study only identified 4 predictors to be significantly related to prolonged ICU stay. The independent predictors are presence of ICU acquired infections, SOFA score within 24 hrs of ICU admission, use of mechanical ventilation and bilirubin levels. These results confirm what been reported by Soares M in a published review about the under-estimation of outcome in cancer patients using the critical care scoring systems alone and highlighted the importance of specific clinical prognostic factors such as mechanical ventilator and bilirubin for more accurate predication in cancer cases (22) with prolonged stay in ICU. APACHE II was found to be related to prolonged ICU stay, the score was not a significant predictor for prolonged ICU stay in the multivariate analysis. Prolonged stay in ICU was found to be significantly related to higher ICU mortality, however, it was not a significant predictor for ICU mortality. The influence of the prolonged stay in ICU that have on short- and long-term outcomes remains controversial. Several studies (25) have reported higher ICU mortality in patients with a prolonged stay in ICU while others have not (26, 27).
This study targeted patients with solid tumors and identified several predictors of mortality and prolonged stay in ICU. Knowledge on such predictors could offer valuable information for clinicians to avoid futile care and better management of critical care resources. ICU survival rate can be significantly increased in patients with solid tumors with careful patient selection during ICU admission (7). Patients who are at their initial phase of their malignant disease should routinely be admitted to the ICU, some selection criteria including the characteristics of the underlying malignancy are not currently reliable to make appropriate triage decisions (1, 3–7, 17–19). We found that the SOFA score a main determinant and useful in predicting ICU mortality. Understanding the factors affecting the prolonged ICU stay may help in improving the quality of care in ICU such as the infection prevention and mechanical ventilator management. Prolonged stay in ICU will definitely impact the critical care bed availability to avoid delayed ICU admission which been well documented as a significant factor that related to worse patient outcomes. Factors such as organs’ failure can be managed easier through earlier admission and evaluation by the ICU team and this might lead to shorter stay in the ICU. This is a prospective registry study with a relatively large number of patients. However, there are a few limitations including being a single-center research. A large multi-center study involving number of ICUs with larger sample size may bear out the findings. Finally, this study collected data only on short term outcomes. Collecting data on long-term outcomes after discharge from ICU could have increased the impact of the current research.