Factors influencing intensive care unit length of stay after craniotomy for supratentorial tumor resection: an observational study
Article information
Abstract
Background
Few studies have explored the intensive care unit (ICU) requirements of patients after craniotomy. Identifying the factors that contribute to extended ICU stay can enhance ICU efficiency and reduce costs. This study investigated the predictors of prolonged ICU stay after craniotomy for supratentorial brain tumors to establish conditions enabling ICU bypass.
Methods
We retrospectively reviewed 205 patients admitted to the ICU after craniotomy for supratentorial brain tumors. Patients were categorized based on ICU length of stay (LOS): ≤24 hours (group 1) and >24 hours (group 2). Logistic regression analysis identified factors linked to extended ICU stay.
Results
Among the patients, 172 (84%) were discharged from the ICU within 24 hours, whereas 33 (16%) had longer stays. Independent predictors of prolonged ICU LOS included midline shift (odds ratio [OR], 2.494; 95% CI, 1.289–4.824; P=0.007), operative blood transfusion (OR, 1.295; 95% CI, 1.007–1.665; P=0.044), longer operative duration (OR, 1.027; 95% CI, 1.014–1.039; P<0.001), and new postoperative neurological deficits (OR, 1.865; 95% CI, 1.426–2.439; P<0.001).
Conclusion
These findings may optimize ICU resource allocation by identifying low-risk patients suitable for early transfer to the neuroscience floor. A selective ICU admission strategy can improve resource use, minimize costs, and enhance ICU availability.
INTRODUCTION
The postoperative management of neurosurgical patients often involves admission to the intensive care unit (ICU) following craniotomy to detect complications [1,2], particularly for neurological ones [3]. However, ICU resources are limited and expensive, often necessitating the re-evaluation of routine ICU admissions. Understanding the factors leading to an extended ICU length of stay (LOS) after brain tumor surgery can aid in resource optimization and cost reduction. Existing research on factors influencing ICU LOS after craniotomy remains limited. This study aimed to identify the preoperative and perioperative factors contributing to ICU stay >24 hours after craniotomy for supratentorial tumors.
METHODS
Design and setting
This study involved a retrospective review of the medical records of adult patients who underwent elective craniotomies for supratentorial tumors between September 2018 and September 2021 in a 10-bed surgical ICU at a University Hospital. In institutional practice, all such patients are routinely admitted to the ICU postoperatively irrespective of their neurological conditions or procedural complexity. Patients with preoperative consciousness disorders were excluded as they were anticipated to have an extended ICU stay. Patients were categorized into two groups based on their ICU LOS: group 1 included patients discharged within 24 hours of surgery, and group 2 included patients who remained in the ICU for >24 hours. Extended ICU stays are typically necessitated by unforeseen postoperative events, either surgery-related or independent, or by the need for specialized ICU resources.
Data collection
Pertinent variables were extracted and compared between the groups. Preoperative data included demographic data, American Society of Anesthesiologists physical status classification, comorbidities (hypertension, diabetes, coronary artery disease, arrhythmia, and ischemic stroke), smoking, alcohol use, steroid use, Glasgow Coma Scale, impaired sensorium, motor deficit, seizures, tumor characteristics (location: midline vs. lateralized; type), raised intracranial pressure (ICP), and a midline shift of >3 mm on computed tomography (CT). ICP was assessed using preoperative imaging studies (e.g., CT or magnetic resonance imaging [MRI]). Radiological signs including midline shift, sulci effacement, and ventricular compression were used to infer increased ICP. However, invasive monitoring was not performed. Echo Doppler findings, such as an increased pulsatility index, were also used to suggest increased ICP in some cases.
Intraoperative data included the method of anesthesia (balanced or total intravenous anesthesia), positioning, extent of tumor resection (gross total resection or subtotal resection), duration of surgery, total volume of crystalloid intake, volume of intraoperative blood transfusion, and use of vasopressors for the treatment of hypotension. Neurological deficits were assessed during the postoperative period and after ICU admission. Neurological status was evaluated immediately upon ICU admission by the neurosurgical team and subsequently at scheduled intervals within the first 24 hours. The assessments included motor function, sensory testing, speech and language evaluation, consciousness level, and cranial nerve examination. Any newly documented deficits compared to the preoperative baseline were recorded, and in cases of neurological deterioration, additional imaging (CT or MRI) was performed to investigate potential complications such as hematoma or ischemia.
Other postoperative complications, including hematoma, seizures, cerebrospinal fluid leak, metabolic disturbances, hemodynamic instability, respiratory complications, cardiac arrhythmia, deep vein thrombosis, coagulopathy, postoperative nausea and vomiting, and death, were also documented. Additionally, the volume of blood products transfused within the first 24 hours postoperatively and the rate of reoperation were recorded.
Statistical analysis
Univariate associations between independent variables and ICU LOS were analyzed using the chi-square test or Fisher’s exact test for categorical variables and Student t-test (or Mann-Whitney U-test when appropriate) for continuous variables. Multicollinearity among predictors was assessed using the variance inflation factor (VIF) with a threshold of five to identify potential collinearity issues. Variables with P<0.05 in univariate analysis were included in the multivariate logistic regression model to determine independent predictors of prolonged ICU stay. All statistical analyses were performed using SPSS version 23 (IBM Corp.).
RESULTS
During the study period, 233 adults met the inclusion criteria. After excluding patients with missing chart data (n=23) and those with preoperative consciousness disorder (n=5), 205 patients were included in the final analysis, of whom 172 (84%) were admitted to the ICU for 1 day (group 1). Thirty-three patients (16%) remained in the unit for >1 day (group 2; mean, 6.9±2.5 days).
The key factors associated with prolonged ICU LOS in univariate analysis are summarized in Table 1. These factors include increased ICP, midline shift, blood transfusion during the operative period, operative time, fluid balance, use of vasopressors, and new neurological complications in the postoperative period. Multicollinearity among predictors was assessed using VIF analysis, with all VIF values <5 indicating no significant multicollinearity.
Preoperative, intraoperative, and postoperative variables collected from 205 patients subjected to ICU admission after elective craniotomy for supratentorial brain surgery, according to length of stay
Multivariate analysis (Table 2) identified four independent predictors of ICU stay >24 hours: midline shift (odds ratio [OR], 4.3; 95% CI, 2.1–7.8; P<0.001), blood transfusion during surgery (OR, 3.2; 95% CI, 1.6–6.4; P=0.002), operative time (OR, 1.8 per hour increase; 95% CI, 1.2–2.5; P=0.01), and postoperative neurologic complications (OR, 6.1; 95% CI, 3.4–10.9; P<0.001).
Multivariate logistic regression evaluating predictors of extended ICU stay following craniotomy for tumors
The reasons for keeping patients in the ICU >24 hours included osmotic therapy (n=5), correction of metabolic disorders (n=5), control of convulsions (n=1), administration of vasoactive drugs (n=4), treatment of cardiac arrhythmia (n=2), glycemic control (n=5), blood pressure control (n=8), need for reoperation (n=1), prolonged ventilation for a difficult respirator weaning (n=1), and re-intubation (n=1). Among the patients who were transferred to the neurosurgery ward on day 2, only two required readmission to the ICU.
DISCUSSION
Few studies have investigated patients’ ICU needs after craniotomy. In this study, 16% of the patients required an extended ICU stay, which is consistent with the findings of previous studies [1,4]. Sarkissian and Wallace [5] reported a higher proportion of prolonged ICU stays (31%), identifying factors such as tumor type, intubation upon ICU admission, and postoperative complications as key contributors to extended ICU LOS after elective craniotomy for supratentorial tumors. Ziai et al. [1] highlighted radiological findings, fluid requirements, significant intraoperative blood loss, and decisions to keep patients intubated postoperatively as predictors of a prolonged ICU stay. Similarly, Hanak et al. [6] found that patients with diabetes, advanced age, receiving blood transfusions, and undergoing prolonged surgical procedures were more likely to require extended intensive care. Kayaalti and Kayaalti [4] also identified factors such as age, fluid balance, intraoperative bleeding, blood transfusion, anesthesia duration, intubation upon ICU admission, mechanical ventilation, and severe postoperative complications as determinants of a longer ICU stay.
In agreement with Ziai et al. [1], our findings showed that midline shift was significantly correlated with prolonged ICU LOS after craniotomy. A midline shift often indicates tumor progression, contributing to surgical challenges and a heightened risk of complications, necessitating longer ICU care. Beyond the midline shift, no other preoperative factors were identified as predictors of extended ICU stay. This lack of significant preoperative determinants likely reflects the elective nature of most craniotomies for tumor resection in patients with minimal coexisting medical conditions.
This study also found that longer operative times were linked to an extended ICU stay, which is consistent with the findings of prior studies [4,6]. Operative time can be influenced by various factors; for instance, patients undergoing more intricate neurosurgical procedures typically have lengthier surgeries. Additionally, a surgeon’s experience can play a significant role in determining the time spent in the operating room. Although midline tumor location itself was not identified as a direct predictor of prolonged ICU LOS in the multivariate analysis, it was frequently associated with factors such as midline shifts and extended operative durations. These related variables were independent significant predictors of longer ICU stay, indicating an indirect effect of midline tumors.
Similarly, benign tumors often require extensive surgical resection to achieve gross total removal. This requirement can result in longer operative times, which were independently correlated with extended ICU stay in this study. Intraoperative blood transfusion has been recognized as a factor influencing ICU LOS. Comparable findings were reported in two other studies [4,6]. Among patients undergoing open neurosurgical procedures for intracranial aneurysms, perioperative blood transfusion was identified as an independent risk factor for significant morbid complications [7]. While our study corroborates the findings of other studies, it uniquely highlights the role of new postoperative neurological deficits as major determinants. The presence of such deficits strongly correlates with a prolonged ICU stay.
In a practice model of care for patients undergoing elective craniotomy for supratentorial tumors, the absence of significant new postoperative neurological impairments in the post-anesthesia care unit served as a criterion for direct transfer to the neurosurgical floor [8]. During the first postoperative day, intracranial complications, particularly hematoma formation, are of the greatest concern. Published experience indicates that clinical deterioration from hematomas often occurs within the initial 6 hours following surgery [9-11].
The patient profiles identified in this study can help optimize the utilization of ICU resources. Based on these findings, a revision of the institutional protocol for the postoperative care of patients undergoing elective craniotomy was proposed. This includes close ICU observation for 4–6 hours before transferring patients to the neuroscience floor, except in cases where the surgeon or anesthesiologist identifies concerns such as radiologic findings, surgery duration, significant blood loss, transfusion requirements, or notable postoperative neurological deficits. Given the scarcity of ICU resources in many healthcare settings, particularly in low- and middle-income countries, these insights could support more efficient ICU admission and resource allocation. A prospective study is planned to evaluate the feasibility, cost, and outcomes of this approach.
This study had several limitations. First, it reflects the experience of a single institution, which may not be generalizable to other ICU settings, despite its consistency with previously published data. Second, as a retrospective study, it is subject to the inherent limitations of such designs, including the potential underreporting of clinical events due to incomplete documentation. Data on tumor size and malignancy were not collected, which may have influenced the interpretation of the results. Finally, the exclusive focus on patients undergoing craniotomy for supratentorial tumors may limit the applicability of these findings to other neurosurgical populations.
This study identified the key factors influencing prolonged ICU stay following craniotomy for supratentorial tumors. Significant predictors of extended ICU stay included midline shift, intraoperative blood transfusion, longer operative times, and new postoperative neurological complications. These findings are consistent with those of previous studies and emphasize the importance of postoperative neurological status and surgical complexity in determining ICU care needs. Optimizing ICU resource utilization by refining postoperative care protocols, particularly in settings with limited resources, can enhance patient outcomes and improve ICU efficiency.
Notes
Ethics statement
The study was approved by the Institutional Review Board of the Military Teaching Hospital Mohammed V (No. M0102019), and the requirement for written informed consent was waived owing to its retrospective design.
Conflict of interest
No potential conflict of interest relevant to this article.
Author contributions
Conceptualization: MB, HB. Methodology: all authors. Formal analysis: all authors. Data curation: all authors. Visualization: HB, KM, MA, MS. Project administration: HB, MB, KM. Writing - original draft: AE. Writing - review & editing: all authors. All authors read and agreed to the published version of the manuscript.