Abstract
Background
The decision to transfer postoperative oral cancer patients to the intensive care unit (ICU) lacks standardized criteria and remains contentious. This study aimed to determine whether ICU transfer independently prolongs hospital length of stay (LOS) and identify the primary determinants of LOS in these patients.
Methods
A retrospective single-center cohort study was conducted in 674 patients who underwent oral cancer surgery. Patients were categorized into the ICU and general ward groups. Data on demographics (age, sex, BMI, smoking, drinking, and betel chewing), comorbidities (hypertension, diabetes and other chronic diseases), intraoperative factors (tracheotomy, duration of operation, lymph node dissection, flap transplantation, and blood transfusion), and clinical scoring systems (SOFA, APACHE-II, CAPRINI and NRS 2002), were all analysed. Linear regression analyses were performed to identify independent determinants of LOS.
Results
Patients in the ICU group were older and had higher rates of hypertension and diabetes (p < 0.05). Between-group comparisons exhibited a higher incidence of intraoperative blood transfusion in the ICU group (p = 0.001), whereas flap transplantation rate was lower (p < 0.001). Comparisons of postoperative clinical scores showed that the ICU group had significantly higher scores of SOFA (p < 0.001), APACHE-II (p < 0.05), and CAPRINI (p < 0.05). Additionally, patients with a SOFA score > 2 or an APACHE-II score ≥ 15 accounted for a higher proportion in this group. Subsequent analysis revealed that there was a longer unadjusted LOS in the ICU group (p < 0.001). However, ICU transfer was not an independent predictor of LOS after adjustment for confounders (p > 0.05). Independent determinants of prolonged LOS included tracheotomy (p < 0.001), intraoperative blood transfusion (p < 0.001), and bilateral lymph node dissection (p < 0.001). Age-stratified analyses indicated that ICU admission independently prolonged LOS in older adults (≥ 65 years, p < 0.05), but not in younger adults (< 65 years, p > 0.05).
Conclusion
ICU transfer is not an independent determinant of the overall LOS of postoperative oral cancer patients, instead, surgical complexity and lymph node dissection play a more critical role. In older adults, ICU transfer represents a key modifiable factor associated with prolonged LOS. These findings support a risk-stratified and age-based triage protocol to optimize ICU utilization and improve healthcare efficacy.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12903-026-07997-1.
Keywords: Hospital Length of Stay, ICU Transfer, Oral Cancer
Background
Oral cancer is a tumour with high morbidity and mortality worldwide, especially in developing countries. According to the recent released report, an estimated 370,000 new cases and 190,000 deaths from lip and oral cancers (LOC) occurred globally in 2019 [1]. Among all cancer types, lip and oral cancers ranked 16th in terms of incidence worldwide [2]. The incidence of oral cancer is significantly higher in Asia, and this phenomenon is closely related to the specific diet and lifestyle, including areca nut chewing, smoking, excessive alcohol consumption, and other high-risk factors, which constitute the core aetiologies of oral cancer [1, 2].
Currently, surgery remains the primary treatment for locally advanced or aggressive oral cancer. Extensive resection and complex reconstructive procedures during the surgery usually increase its risk [3, 4]. Consequently, postoperative monitoring and patient management after surgery is critical, particularly in the first few days following surgery [5, 6]. Some patients may be directly transferred to the intensive care unit (ICU) for close monitoring during this vulnerable period in order to prevent and manage the potential postoperative complications without delay [5]. However, there are currently no standardized clinical criteria to determine whether all postoperative patients require ICU admission. At present, decision-making of ICU admission is largely based on individual clinical experience of surgeons and anesthesiologists [7].
In recent years, debates have been continued on the pros and cons of ICU admission. Abundant evidences have indicated that routine ICU monitoring is not mandatory for all postoperative head and neck cancer patients. Clinical outcomes of the patients who are carefully selected and managed in general wards are even comparable to those receiving the ICU care [8, 9]. Additionally, patients admitted to the ICU tend to have a longer length of stay (LOS), which might adversely affect their prognosis. Therefore, it remains unclear whether ICU transfer prolongs LOS or simply serves as a perioperative risk factor. in this study, we aimed to evaluate whether postoperative ICU admission independently influences LOS in patients undergoing oral cancer surgery. Relevant clinical data and postoperative indicators were compared and analysed between the ICU and general ward care. Stratified analyses were performed afterwards according to different clinical scoring systems and indexes.
Materials and methods
Inclusion and exclusion criteria
Inclusion criteria: patients with a definite diagnosis of oral cancer, and with surgical treatment in the Oral and Maxillofacial Department of Sun Yat-sen Memorial Hospital (SYSMH), Sun Yat-sen University, Guangzhou, China. Patients were admitted to the ICU immediately after the surgery if they had severe organ function impairment, risk factors, or comorbidity. In some cases, admission to the ICU was due to the decision of the operating surgeon. All participants were 18 or more years of age and had complete medical records. Exclusion criteria: those with repeated admission records or with non-malignant tumours were excluded from the research (Fig. 1).
Fig. 1.
Flow chart of the study design. (A) Work flow of data analysis, which included patient data collection, data grouping, statistical analysis, filtering, and identification of the influencing factors. (B) The specific clinical indexes analyzed in the study. SOFA: Sequential Organ Failure Assessment; APACHE-II: Acute Physiology and Chronic Health Evaluation-II; NRS 2002: Nutritional Risk Screening 2002; CAPRINI: Caprini Venous Thromboembolism Risk Assessment Model
General data and group divisions
674 patients undergoing surgical treatment for oral cancer were eventually recruited in the stomatology department of SYSMH from 1 January 2022 to 31 December 2022. This study was approved by the ethics committee of SYSMH (Ethics Code: SYSKY-2024-783-01). The detailed study design and data processing are shown in the flow chart in Fig. 2.
Fig. 2.
Detailed data processing of the study, including the selection and group divisions for the patients enrolled in the analysis. ICU: intensive care unit
The enrolled patients (n = 674) were divided into the ICU group (n = 134, 19.88%) and the general ward group (n = 540, 80.12%). Further division of the subgroups was performed according to Sequential Organ Failure Assessment (SOFA), Acute Physiology and Chronic Health Evaluation (APACHE)-II scores [10, 11]. The detailed information was as follows: (1) ICU group: SOFA ≤ 2 (n = 33, 24.63%) and SOFA > 2 (n = 101, 75.37%); APACHE-II < 15 (n = 72, 53.73%) and APACHE-II ≥ 15 (n = 62, 46.27%); (2) General ward group: SOFA ≤ 2 (n = 437, 80.93%) and SOFA > 2 (n = 103, 19.07%); APACHE-II < 15 (n = 534, 98.89%) and APACHE-II ≥ 15 (n = 6, 1.11%).
For the age stratification, the whole set of 674 patients was separately divided into two groups: (1) Older adults: age ≥ 65 years (n = 199, 29.53%); (2) Younger adults: age < 65 years (n = 475, 70.47%).
Clinical data collection and evaluation
The clinical data of the patients were collected through the electronic medical records system of the hospital. The baseline information included age, gender, and body mass index (BMI) [12]. Personal histories were also recorded, which contained smoking, drinking, and betel nut chewing habits. In addition, information on medical histories was collected, including hypertension, diabetes, prior radiotherapy or chemotherapy, and other diseases, including cardiovascular disease (CVD) and cerebral infarction (CI).
In terms of clinical indexes, we mainly focused on preoperative treatments, operation duration, tracheotomy, lymph node dissection, flap transplantation, fibula/femur flap transplantation, intraoperative blood transfusion, as well as other relevant factors that might influence prognosis. Relevant clinical scoring systems, including SOFA, APACHE-II, nutrition risk screening (NRS) 2002 and Caprini Venous Thromboembolism Risk Assessment Model (CAPRINI) scores were applied in the evaluation of organ dysfunction and postoperative complications [10, 11, 13, 14]. The SOFA was comprised by respiratory function, coagulation function, liver function, cardiovascular system, central nervous system and kidney function. SOFA ≤ 2 usually indicates mild organ function damage [10], while SOFA > 2 indicated a moderate to severe damage. APACHE-II score was specifically designed to assess the severity of illness in ICU patients and predict their prognosis [11]. Patients with an APACHE-II score ≥ 15 could be defined as critically ill patients. NRS 2002 score was used to assess functional status impairment and the overall condition of the patient [13]. The cumulative score ≥ 3 often indicated a risk of malnutrition. The CAPRINI score was an important assessment scale for the risk of venous thrombosis in clinical practice [14]. Higher total score represented higher risk of venous thrombosis. Comparisons of the LOS were conducted between groups and subgroups based on relevant stratifications.
Statistical analysis
The R (R Software, Tsinghua University, China) language and relevant packages were used for statistical analysis. Detailed packages included readxl (data import), dplyr (data processing), stats (t-test, correlation, rank-sum, Mann-Whitney U), broom (regression output), writexl (export), and tibble (variable management). A p-value of less than 0.05 was considered to be statistically significant.
Continuous variables were initially assessed for normality using the Shapiro–Wilk test and Q-Q plots prior to further data analysis. The variable distribution was presented as the normal (or approximately normal) or skewed type according to the normality test results. Data with normal (or approximately normal) distribution included LOS, SOFA score, BMI, APACHE-II score, NRS 2002 score, and CAPRINI score. Other variables, including age and duration of operation, showed a skewed distribution. Continuous variables with normal distributions were described as mean ± standard deviation (SD), while those with skewed distribution were described as median and interquartile range (IQR). Categorical variables were described as frequencies and percentages. In addition, they were dummy-coded for inclusion in regression models. For univariate analysis, between-group differences were assessed using two-tailed independent-samples t-tests in continuous variables with normal distribution, while the Mann-Whitney U test was applied to variables with skewed distribution (i.e., age and duration of operation). Associations between categorical variables and relevant outcome were evaluated using the χ² test or Fisher’s exact test, and results were reported as corresponding test statistics (t, U, or χ²) and p-values (Supplementary Table 1). Variables showing a significant association (p < 0.05) in the univariate analysis were included in the multivariate analysis (Supplementary Table 2) [15, 16].
Multiple linear regression was employed to identify key factors correlated with LOS (Supplementary Table 2) [15, 16]. The exported results were mainly presented as regression coefficients, standard errors (SE), t-values, and p-values. To explore potential effect modifications, patients in the ICU group were further stratified by age for subgroup analysis. Additionally, the clinical implications of the relevant scoring systems were further investigated.
In the univariate analysis, we used a p-value threshold of p < 0.1 to select candidate variables for inclusion in the multivariate model. This approach is widely adopted in clinical and epidemiological research because the goal of univariate screening is not to test hypotheses but to identify potential predictors that could be relevant in a multivariable context. Using a more liberal threshold (e.g., p < 0.10 instead of p < 0.05) reduces the risk of excluding variables that may not reach conventional significance in univariate tests but could be important predictors when adjusted for confounders in the multivariate model. This approach has been recommended in regression modeling literature, including the purposeful selection strategy described by Hosmer et al. [17] and further supported by subsequent methodological studies [18].
Results
Demographic baseline characteristics of patients in ICU and general ward
Among all the participants, 134 patients were transferred to the ICU after surgery, and 540 patients were transferred to the general ward. The demographic characteristics of the patients were shown in Table 1.
Table 1.
Baseline characteristics of postoperative patients in the ICU and general ward groups
| Variable | Total n = 674 |
ICU n = 134 |
General ward n = 540 |
Z / χ² | p |
|---|---|---|---|---|---|
|
Age, Y (median (IQR)/ mean ± SD) |
56.0 (46.00, 66.00); 55.92 ± 14 |
65.0 (55.00, 72.00); 62.81 ± 15.98 |
54.0 (45.00, 64.00); 54.2 ± 12.91 |
−6.42 | < 0.001 |
| Male, n (%) | 465 (68.99) | 88 (65.67) | 377 (69.81) | 0.86 | 0.353 |
| BMI (kg/m²), n (%) | 2.77 | 0.251 | |||
| < 18.5 | 81 (12.02) | 21 (15.67) | 60 (11.11) | ||
| 18.5–24.0 | 363 (53.86) | 73 (54.48) | 290 (53.70) | ||
| > 24.0 | 230 (34.12) | 40 (29.85) | 190 (35.19) | ||
| LOS ( median (IQR)/ mean ± SD ) | 13 (10, 16); 13.64 ± 5.67 |
13 (11, 18); 15.44 ± 6.91 |
13 (10, 15.75); 13.19 ± 5.24 | 3.53 | < 0.001 |
| Smoking, n (%) | 268(39.76) | 54 (40.30) | 214 (39.63) | 0.02 | 0.887 |
| Drinking, n (%) | 162 (24.04) | 20 (14.93) | 142 (26.30) | 7.60 | 0.006 |
| Betel chewing, n (%) | 58 (8.61) | 7 (5.22) | 51 (9.44) | 2.432 | 0.119 |
| Hypertension, n (%) | 158 (23.44) | 45 (33.58) | 113 (20.93) | 9.58 | 0.002 |
| Diabetes, n (%) | 84 (12.46) | 25 (18.66) | 59 (10.93) | 5.88 | 0.015 |
| Other chronic diseases, n (%) | 404 (59.94) | 61 (45.52) | 343 (63.52) | 14.48 | < 0.001 |
| Pretreatments, n (%) | 260 (38.58) | 57 (42.54) | 203 (37.59) | 1.11 | 0.293 |
| Pre-operation, n (%) | 89 (13.20) | 18 (13.43) | 71 (13.15) | 0.01 | 0.931 |
| Chemoradiotherapy, n (%) | 200 (29.67) | 46 (34.33) | 154 (28.52) | 1.74 | 0.188 |
Continuous variables were compared using the independent-samples t-test or the Mann-Whitney U test, depending on data distribution; categorical variables were analysed using the χ² test
BMI Body mass index, BMI = mass [kg]/ (height [m])², kg/m kilogram/meter, other chronic diseases: such as cardiovascular disease (CVD) and cerebral infarction (CI), LOS Length of stay, SD Standard deviation
The overall age of the 674 patients ranged from 18 to 99 years old, and the median age was 56. There was a significant difference in the median age between the ICU and the general ward (65 vs. 54, p < 0.001, Table 1). The number of patients with a personal history of alcohol consumption was lower in the ICU group in contrast with the general ward group (14.93% vs. 26.3%, p = 0.006, Table 1). ICU group had a significantly longer LOS (15.44 ± 6.91 vs. 13.19 ± 5.24 days, p < 0.001, Table 1). Statistical analysis also showed obvious differences in previous disease history in the two groups. There were more patients with hypertension and diabetes in the ICU group than the general ward group (hypertension: 33.58% vs. 20.93%, p = 0.002; diabetes: 18.66% vs. 10.93%, p = 0.015, Table 1). However, the percentage of patients with other diseases, including cardiovascular disease (CVD) and cerebral infarction (CI) (CVD and CI) were much lower in the ICU group compared with the general ward group (45.52% vs. 63.52%, p < 0.001, Table 1). There were no significant differences between the two groups in other indexes such as sex, BMI, smoking, betel chewing and preoperative treatments, including surgery and chemoradiotherapy (p > 0.05, Table 1).
Analysis of intraoperative indicators of the enrolled patients in the ICU group and the general ward group
Intraoperative situations of the patients with oral cancer were detected in the ICU and the general ward groups (Table 2). Analysis indicated that there were significant differences in flap transplantation and blood transfusion in the two groups (85.82% vs. 95.00%, p < 0.001; 32.8% vs. 20.00%, p = 0.001, Table 2). Subgroup analysis for flap transplantation revealed that differences existed in fibula flap transplantation (26.12% vs. 36.67%, p = 0.022, Table 2). No significant differences were found in other indexes, including femur flap transplantation, other flap transplantation types, operation duration, tracheotomy, and lymph node dissection (p > 0.05, Table 2).
Table 2.
Analysis of intraoperative indicators for the enrolled patients in the ICU and general ward groups
| Variable | Total n = 674 |
ICU n = 134 |
General ward n = 540 |
Z / χ² | p |
|---|---|---|---|---|---|
| Tracheotomy, n (%) | 615 (91.25) | 118 (88.06) | 497 (92.04) | 2.13 | 0.145 |
| Duration of operation (minutes, median (IQR)/ mean ± SD) | 340.00 (270.00, 475.00); 367.52 ± 156.38 | 347.50 (270.00, 477.50); 372.28 ± 157.83 | 333.00 (270.00, 475.00); 366.34 ± 156.15 | −0.32 | 0.747 |
| Unilateral lymph node dissection, n (%) | 455 (67.51) | 84 (62.69) | 371 (68.70) | 1.77 | 0.183 |
| Bilateral lymph node dissection, n (%) | 156 (23.15) | 35 (26.12) | 121 (22.41) | 0.83 | 0.362 |
| Flap transplantation, n (%) | 628 (93.18) | 115 (85.82) | 513 (95.00) | 14.22 | < 0.001 |
|
Fibula flap transplantation, n (%) |
233 (34.57) | 35 (26.12) | 198 (36.67) | 5.28 | 0.022 |
| Femur flap transplantation, n (%) | 337 (50.00) | 71 (52.99) | 266 (49.26) | 0.60 | 0.44 |
| Other flap transplantation, n (%) | 58 (8.61) | 9 (6.72) | 49 (9.07) | 0.76 | 0.384 |
| Blood transfusion, n (%) | 152 (22.55) | 44(32.8) | 108 (20.00) | 10.13 | 0.001 |
Continuous variables were compared using the Mann-Whitney U test, depending on data distribution; categorical variables were analyzed using the χ² test
Comparison of postoperative clinical indicators between ICU and general ward
Postoperative clinical indicators were further probed in the ICU and the general ward groups. Series of clinical scoring systems were adopted for the assessments of organ dysfunction and postoperative complications, which mainly contained SOFA, APACHE-II and CAPRINI scores (Table 3). Comparisons of relevant indexes between the two groups revealed that there were remarkable differences in these three types of scores, and the ICU group presented much higher scores of SOFA (3.34 vs. 2.03, p < 0.001, Fig. 3A), APACHE-II (14.89 vs. 8.6, p = 0.002, Fig. 3B), and CAPRINI score (10.92 ± 2.87 vs. 10.33 ± 2.37, p = 0.03, Fig. 3C). Subgroup analysis based on SOFA and APACHE-II scores demonstrated that patients with a SOFA score > 2 or APACHE-II score ≥ 15 constituted much larger proportions in the ICU in contrast with the general ward (75.37% vs. 19.07%, p < 0.001, Tables 3 and 46.27% vs. 1.11%, p < 0.001, Table 3).
Table 3.
Comparison of postoperative indicators for the patients in the ICU and general ward groups
| Variable | Total n = 674 |
ICU n = 134 |
General ward n = 540 |
Z / χ² | p |
|---|---|---|---|---|---|
| SOFA score, n (%) | 161.23 | < 0.001 | |||
| ≤ 2 | 470 (69.73) | 33(24.63) | 437 (80.93) | ||
| > 2 | 204 (30.27) | 101(75.37) | 103 (19.07) | ||
| APACHE II score, n (%) | 241.34 | < 0.001 | |||
| < 15 | 606 (89.91) | 72 (53.73) | 534 (98.89) | ||
| ≥ 15 | 68 (10.09) | 62 (46.27) | 6 (1.11) | ||
| CAPRINI score (median (IQR)/ mean ± SD) | 11 (9, 12); 10.45 ± 2.49 | 12 (9, 13); 10.92 ± 2.87 | 11 (9, 12); 10.33 ± 2.37 | 2.45 | 0.03 |
Continuous variables were compared using the independent-samples t-test, depending on data distribution; categorical variables were analyzed using the χ² test
ICU Intensive care unit ICU, SOFA Sequential Organ Failure Assessment, APACHE-II Acute Physiology and Chronic Health Evaluation- II, CAPRINI Caprini Venous Thromboembolism Risk Assessment Model
Fig. 3.
Comparisons of baseline characteristics between the ICU group and the general ward group. (A) Comparison of the SOFA core between the two groups (p < 0.001). (B) Comparison of the APACHE II score between the two groups (p < 0.001). (C) Comparison of the CAPRINI score between the two groups (p = 0.03). (D) Comparison of LOS between the two groups (p < 0.001). ICU: intensive care unit; SOFA: Sequential Organ Failure Assessment; APACHE-II: Acute Physiology and Chronic Health Evaluation-II; CAPRINI: Caprini Venous Thromboembolism Risk Assessment Model; LOS: length of stay. (***, p < 0.001;**, p < 0.01; *, p < 0.05; ns: p > 0.05)
Comparisons of LOS in postoperative patients in ICU and general ward
LOS was checked from electronic medical records of the oral cancer patients. It was found that LOS in the ICU group was longer than the general ward group (15.44 ± 6.9 vs. 13.19 ± 5.24, p < 0.001, Fig. 3D). Subgroup analysis based on APACHE-II score did not display a significant difference in LOS between groups when APACHE-II < 15 (Fig. 4A), whereas LOS appeared to be longer in the ICU when APACHE-II ≥ 15 (14.00 vs. 12.00, p = 0.001, Fig. 4B). SOFA score stratification did not exhibit significant differences in LOS (Fig. 4C and D).
Fig. 4.
Comparisons of LOS of the postoperative patients between the ICU group and the general ward group. (A) Comparison of LOS between the two groups in the condition of APACHE II < 15. (B) Comparison of LOS between the two groups in the condition of APACHE II ≥ 15 (p < 0.01). (C) Comparison of LOS between the two groups in the condition of SOFA ≤ 2. (D) Comparison of LOS between the two groups in the condition of SOFA > 2. ICU: intensive care unit. SOFA: Sequential Organ Failure Assessment; APACHE-II: Acute Physiology and Chronic Health Evaluation-II; LOS: length of stay. (***, p < 0.001;**, p < 0.01; *, p < 0.05; ns: p > 0.05)
Factors influencing LOS
The factors that might impact LOS were examined step by step via linear regression analysis afterwards (Table 4). There was a weak correlation between ICU transfer and LOS in a univariate regression model, but with relatively lower effect size (point biserial correlation coefficient, rPB = 0.158, p < 0.001, Table 4). Other variables with statistical significances in univariate analysis included intraoperative blood transfusion (rPB = 0.306, p < 0.001, Table 4), tracheotomy (rPB = 0.286, p < 0.001, Table 4), bilateral lymph node dissection (rPB = 0.204, p < 0.001, Table 4), history of other diseases (rPB = 0.095, p = 0.009, Table 4), age (correlation coefficient, r = 0.088, p = 0.022, Table 4), BMI ( r = − 0.136, p < 0.001, Table 4), and surgical duration (r = 0.192, p < 0.001, Table 4). However, factors such as diabetes, skin flap transplantation, and hypertension did not demonstrate statistical significances (p > 0.05, Table 4).
Table 4.
Univariate and multivariate linear regression analyses of factors associated with length of stay (LOS)
| Variable | Univariate linear regression | Multivariate regression | ||||
|---|---|---|---|---|---|---|
| Total n = 674 |
Correlation coefficient, r | p value | Regression coefficient, β | t value | p value | |
| ICU transfer, n (%) | 134 (19.88) | 0.158 | 0.001 | 0.912 | 1.398 | 0.163 |
| SOFA score (median (IQR)/ mean ± SD) |
2 (2, 3); 2.48 ± 1.23 |
0.305 | < 0.001 | 0.695 | 3.335 | < 0.001 |
| APACHE-II score (median (IQR)/ mean ± SD) |
9 (7, 12); 9.82 ± 4.46 |
0.224 | < 0.001 | −0.010 | −0.157 | 0.875 |
| CAPRINI score (median (IQR)/ mean ± SD) |
11 (9, 12); 10.45 ± 2.49 |
0.254 | < 0.001 | 0.237 | 2.618 | 0.009 |
| NRS2002 score (median (IQR)/ mean ± SD) |
3 (1, 4); 2.86 ± 1.61 |
0.057 | 0.138 | |||
| Age, Y (median (IQR)/ mean ± SD) | 56.00 (46.00, 66.00); 55.92 ± 14 | 0.088 | 0.022 | 0.003 | 0.170 | 0.865 |
| Sex, n (%) | 0.004 | 0.914 | ||||
| Female | 209 (31.01) | |||||
| Male | 465 (68.99) | |||||
| BMI (kg/m², median (IQR)/ mean ± SD) |
22.35(20.08, 4.92); 22.54 ± 3.67 |
−0.136 | < 0.001 | −0.053 | −0.962 | 0.337 |
| Hypertension, n (%) | 158 (23.44) | 0.001 | 0.983 | |||
| Diabetes, n (%) | 84 (12.46) | 0.026 | 0.495 | |||
| Other chronic diseases | 404 (59.94) | 0.095 | 0.009 | 1.283 | 3.195 | 0.001 |
| Pretreatment, n (%) | 260 (38.58) | −0.042 | 0.407 | |||
| Pre-operation, n (%) | 89 (13.20) | −0.062 | 0.122 | |||
| Chemoradiotherapy, n (%) | 200 (29.67) | −0.005 | 0.903 | |||
| Smoking, n (%) | 268 (39.76) | 0.012 | 0.752 | |||
| Drinking, n (%) | 162 (24.04) | 0.033 | 0.369 | |||
| Betel chewing, n (%) | 58 (8.61) | −0.031 | 0.397 | |||
| Tracheotomy, n (%) | 615 (91.25) | 0.285 | < 0.001 | 3.381 | 3.988 | < 0.001 |
| Duration of operation (median (IQR)/ mean ± SD) |
367.52 ± 156.38; 340.00; (270.00, 475.00) |
0.192 | < 0.001 | 0.001 | 0.327 | 0.744 |
| Unilateral lymph node dissection, n (%) | 455 (67.51) | −0.055 | 0.197 | |||
| Bilateral lymph node dissection, n (%) | 156 (23.15) | 0.204 | < 0.001 | 1.831 | 3.952 | < 0.001 |
| Flap transplantation, n (%) | 628 (93.18) | 0.099 | 0.052 | |||
| Fibula flaps, n (%) | 233 (34.57) | 0.045 | 0.207 | |||
| Femur flaps, n (%) | 337 (50.00) | 0.027 | 0.481 | |||
| Other flaps, n (%) | 58 (8.61) | 0.093 | 0.082 | |||
| Blood transfusion, n (%) | 152 (22.55) | 0.306 | < 0.001 | 2.905 | 5.935 | < 0.001 |
| Intercept | 5.281 | 2.911 | 0.004 | |||
LOS Length of stay, ICU Intensive care unit, BMI Body mass index, BMI = mass [kg]/ (height [m])², kg/m kilogram/meter, other chronic diseases: such as cardiovascular disease (CVD) and cerebral infarction (CI), SOFA Sequential Organ Failure Assessment, APACHE-II Acute Physiology and Chronic Health Evaluation-II, NRS 2002 Nutritional Risk Screening 2002, CAPRINI Caprini Venous Thromboembolism Risk Assessment Model
Variables with p < 0.1 in the univariate analysis were further employed in the multifactor linear regression model to evaluate their independent effects on LOS. No statistical significance was revealed in the regression coefficient of ICU transfer (estimate, β = 0.9118, p = 0.1626, Table 4), suggesting that the effect of ICU transfer on LOS was not statistically significant after adjustment of the other variables. The independent factors that demonstrated statistically significant influence on LOS included intraoperative blood transfusion (β = 2.905, p < 0.001, Table 4), tracheotomy (β = 3.381, p < 0.001, Table 4), bilateral lymph node dissection (β = 1.832, p < 0.001, Table 4), history of other diseases (β = 1.283, p = 0.001, Table 4), SOFA score (β = 0.695, p = 0.001, Table 4), and CAPRINI scores (β = 0.238, p = 0.009, Table 4). The abovementioned results showed that ICU transfer was not an independent influencing factor that dominated LOS, and the main determinants were intraoperative blood transfusion, tracheotomy, bilateral lymph node dissection, history of other diseases, CAPRINI and SOFA scores.
Factors influencing LOS based on age group analysis
The enrolled patients were separately stratified into the older adults and younger adults based on age (Tables 5 and 6). Linear regression analysis indicated that ICU transfer could extend LOS for the patients in the older adults (p = 0.021, Table 5). There were several other positive factors that strongly impacted LOS, including bilateral lymph node dissection (p = 0.010, Table 5), tracheotomy (p = 0.002, Table 5), and intraoperative blood transfusion (p = 0.005, Table 5). However, preoperative treatment was shown to reduce LOS (p = 0.017, Table 5). Other variables (SOFA score, APACHE-II score, BMI, and operation duration) did not demonstrate statistically significant effects in multivariate analysis.
Table 5.
Multivariate linear regression analysis of factors associated with length of stay (LOS) in patients aged ≥ 65 years
| Variable | Univariate linear regression | Multivariate regression | ||||
|---|---|---|---|---|---|---|
| Total n = 199 |
Correlation coefficient, r | p value | Regression coefficient, β | t value | p value | |
| ICU transfer, n (%) | 70 (35.2) | −3.507 | < 0.001 | 2.036 | 2.326 | 0.021 |
| SOFA score (median (IQR)/ mean ± SD) |
2 (2, 4); 2.9 ± 1.5 |
0.303 | < 0.001 | 0.289 | 1.003 | 0.317 |
| APACHE II score (median (IQR)/ mean ± SD) |
12 (10, 15); 12.8 ± 4.6 |
0.245 | < 0.001 | 0.052 | 0.535 | 0.593 |
| NRS2002 score (median (IQR)/ mean ± SD) |
4 (2, 5); 3.291 ± 1.7 |
0.099 | 0.163 | |||
| CAPRINI score (median (IQR)/ mean ± SD) |
12 (10 ,13); 11.2 ± 2.7 |
0.130 | 0.067 | −0.017 | −0.120 | 0.889 |
| Male, n (%) | 115 (57.8) | −1.161 | 0.247 | |||
| BMI (kg/m², median (IQR)/ mean ± SD) |
21.57 (19.72, 24.30); 21.91 ± 3.474 |
−0.123 | 0.084 | −0.092 | −0.962 | 0.337 |
| Hypertension, n (%) | 87 (43.7) | −0.352 | 0.725 | |||
| Diabetes, n (%) | 34 (17.1) | 1.028 | 0.307 | |||
| Other chronic diseases, n (%) | 121 (60.8) | −1.211 | 0.228 | |||
| Pretreatment, n (%) | 67 (33.7) | 2.408 | 0.017 | −1.748 | −2.418 | 0.017 |
| Pre-operation, n (%) | 29 (14.6) | 1.686 | 0.101 | |||
| Chemoradiotherapy, n (%) | 46 (23.1) | 0.860 | 0.392 | |||
| Smoking, n (%) | 64 (32.2) | 0.023 | 0.981882661 | |||
| Drinking, n (%) | 37 (18.6) | −0.47 | 0.640 | |||
| Betel chewing, n (%) | 199 (100) | |||||
| Duration of operation (median (IQR)/ mean ± SD) | 305 (240, 420); 330 ± 147 | 0.192 | < 0.001 | −0.0003 | −0.147 | 0.884 |
| Bilateral lymph node dissection, n (%) | 43 (21.6) | −3.364 | 0.002 | 2.084 | 2.616 | 0.010 |
| Unilateral lymph node dissection, n (%) | 133 (66.8) | 1.138 | 0.258 | |||
| Flap transplantation, n (%) | 182 (91.5) | −0.840 | 0.412 | |||
| Fibula flaps, n (%) | 65 (32.7) | −1.576 | 0.117 | |||
| Femur flaps, n (%) | 96 (48.2) | 0.072 | 0.943 | |||
| Other flaps, n (%) | 19 (9.5) | −0.295 | 0.771 | |||
| Tracheotomy, n (%) | 175 (87.9) | −3.656 | 0.001 | 3.812 | 3.146 | 0.002 |
| Blood transfusion, n (%) | 45 (22.6) | −2.911 | 0.005 | 2.295 | 2.833 | 0.005 |
| Intercept | 10.523 | 3.931 | < 0.001 | |||
LOS Length of stay, ICU Intensive care unit, BMI Body mass index, BMI = mass [kg]/ (height [m])², kg/m kilogram/meter, other chronic diseases: such as cardiovascular disease (CVD) and cerebral infarction (CI), SOFA Sequential Organ Failure Assessment, APACHE-II Acute Physiology and Chronic Health Evaluation-II, NRS 2002 Nutritional Risk Screening 2002, CAPRINI Caprini Venous Thromboembolism Risk Assessment Model
Table 6.
Multivariate linear regression analysis of factors associated with length of stay (LOS) in patients aged < 65 years
| Variable | Univariate linear regression | Multivariate regression | ||||
|---|---|---|---|---|---|---|
| Total n = 475 |
Correlation coefficient, r | p value | Regression coefficient, β | t value | p value | |
| ICU transfer, n (%) | 64 (13.5) | −1.628 | 0.108 | −0.258 | −0.287 | 0.774 |
| SOFA score (median (IQR)/ mean ± SD) |
2 (2, 3); 2.3 ± 1.0 |
0.315 | < 0.001 | 0.923 | 3.102 | 0.002 |
| APACHE-II score (median (IQR)/ mean ± SD) |
8 (6, 10); 8.6 ± 3.8 |
0.215 | < 0.001 | −0.03 | −0.365 | 0.715 |
| NRS2002 score (median (IQR)/ mean ± SD) |
3 (1 ,4); 2.7 ± 1.5 |
0.029 | 0.532 | |||
| CAPRINI score (median (IQR)/ mean ± SD) |
10 (9, 12); 10.1 ± 2.3 |
0.303 | < 0.001 | 0.427 | 3.492 | 0.001 |
| Male, n (%) | 350 (73.7) | 0.345 | 0.731 | |||
| BMI (kg/m², median (IQR)/ mean ± SD) |
22.60 (20.28, 25.16); 22.8 ± 3.7 |
−0.134 | 0.003 | −0.022 | −0.328 | 0.743 |
| Hypertension, n (%) | 71 (14.9) | 0.894 | 0.373 | |||
| Diabetes, n (%) | 50 (10.5) | −1.038 | 0.304 | |||
| Other chronic diseases, n (%) | 283 (59.6) | −2.292 | 0.022 | 1.062 | 2.164 | 0.031 |
| Pretreatment, n (%) | 193 (40.6) | 0.412 | 0.681 | |||
| Pre-operation, n (%) | 60 (12.6) | 0.106 | 0.916 | |||
| Chemoradiotherapy, n (%) | 154 (32.4) | −0.392 | 0.695 | |||
| Smoking, n (%) | 204 (42.9) | −0.534 | 0.593 | |||
| Drinking, n (%) | 125 (26.3) | −0.91 | 0.364 | |||
| Areca nut chewing, n (%) | 58 (12.2) | 0.555 | 0.58 | |||
| Duration of operation (median (IQR)/ mean ± SD) |
360 (280, 490); 383 ± 158 |
0.223 | < 0.001 | 0.002 | 1.243 | 0.215 |
| Bilateral lymph node dissection, n (%) | 113 (23.8) | −3.804 | < 0.001 | 1.443 | 2.543 | 0.011 |
| Unilateral lymph node dissection, n (%) | 322 (67.8) | 0.811 | 0.418 | |||
| Flap transplantation, n (%) | 446 (93.9) | −1.873 | 0.071 | −1.911 | −1.634 | 0.103 |
| Fibula flaps, n (%) | 168 (35.4) | −0.597 | 0.551 | |||
| Femur flaps, n (%) | 241 (50.7) | −0.887 | 0.376 | |||
| Other flaps, n (%) | 39 (8.2) | −1.766 | 0.085 | 2.371 | 2.647 | 0.008 |
| Tracheotomy, n (%) | 440 (92.6) | −7.944 | < 0.001 | 3.112 | 2.545 | 0.011 |
| Blood transfusion, n (%) | 107 (22.5) | −6.061 | < 0.001 | 3.289 | 5.377 | < 0.001 |
| Intercept | 3.947 | 1.892 | 0.059 | |||
LOS Length of stay, ICU Intensive care unit, BMI Body mass index, BMI = mass [kg]/ (height [m])², kg/m kilogram/meter, other chronic diseases: such as cardiovascular disease (CVD) and cerebral infarction (CI), SOFA Sequential Organ Failure Assessment, APACHE-II Acute Physiology and Chronic Health Evaluation-II, NRS 2002 Nutritional Risk Screening 2002, CAPRINI Caprini Venous Thromboembolism Risk Assessment Model
For the younger adults, ICU transfer did not exhibit a statistically significant effect on LOS in multivariate analysis (p = 0.774, Table 6). Factors that were shown to extend LOS with statistical significance included SOFA score (p = 0.002, Table 6), CAPRINI score (p = 0.001, Table 6), other chronic diseases (p = 0.031, Table 6), bilateral lymph node dissection (p = 0.011, Table 6), other skin flap transplantation (p = 0.008, Table 6), tracheotomy (p = 0.011, Table 6), and intraoperative blood transfusion (p < 0.001, Table 6). Other variables (BMI, duration of surgery, and preoperative treatment), did not demonstrate any statistically significant effects in the multivariate analysis.
In summary, relevant determinants, including tracheotomy, intraoperative blood transfusion, and bilateral lymph node dissection could all extend LOS in both age groups with statistical significance. For the elderly patients, ICU transfer and preoperative treatment were the main factors that affected LOS, while for the younger adults, the key correlated factors included SOFA score, CAPRINI score, other chronic diseases, and other skin flap transplantation.
Discussion
In this study, retrospective analysis was used to comprehensively investigate the clinical indicators and LOS in 674 postoperative oral cancer patients. To address potential confounding, multivariable models were constructed adjusting for key demographic, clinical, and surgical factors, including age, sex, comorbidities (hypertension, diabetes, cardiovascular disease, and cerebral infarction), preoperative treatment, surgical type and duration, tracheotomy, intraoperative blood transfusion, lymph node dissection, and flap transplantation. After risk adjustment, ICU transfer was not identified as an independent determinant of LOS. The main determinants of LOS across different age groups included tracheotomy, intraoperative blood transfusion, and bilateral lymph node dissection. For the older adults, ICU transfer and preoperative treatment were the main factors that affected LOS, while for the younger ones, SOFA and CAPRINI scores, other chronic diseases, and other skin flap transplantation were the main factors.
In the present work, we initially compared the baseline clinical information of the enrolled patients between the ICU and the general ward group. The median age was higher in the ICU group, and hypertension and diabetes were more prevalent. The ICU plays a critical role in the postoperative care of surgical patients, and there are numerous oral cancer patients receiving ICU monitoring post surgery [7, 19]. The ICU provides advanced monitoring and comprehensive life-support for high-risk conditions in postoperative management, such as acute cardiovascular events and perioperative hyperglycaemia [20–22]. In contradistinction to hypertension and diabetes, the proportion of patients with other diseases, such as CVD and CI, were lower in the ICU group, well-controlled chronic conditions alone may not independently necessitate postoperative ICU admission. This finding might be attributable to adequate preoperative managements and comprehensive perioperative assessment, which mitigate postoperative risk and reduce the need for intensive care [23, 24].
For patients with oral cancer in the early stage, surgical resection remains the preferred primary treatment modality [25]. The complexity of surgical procedures is associated with a higher risk of postoperative complications, including infection, bleeding, flap necrosis, and aggravation of swallowing disorders [26]. Accordingly, we further investigated the intraoperative determinants of the patients. Distinct between-group differences were observed in intraoperative flap transplantation and blood transfusion. The ICU group showed a lower rate of flap transplantation, which was contrary to some previous reports and should be further validated in larger samples [27, 28]. On the other side, intraoperative blood transfusion was more frequent among patients admitted to the ICU, suggesting that transfusion may reflect greater surgical complexity or perioperative risk, which should be carefully balanced in clinical decision-making [29].
After the investigation of relevant intraoperative indexes, we further appraised postoperative organ dysfunction and complications through several clinical scoring systems. The ICU group exhibited significantly higher SOFA and APACHE-II scores, with a greater proportion of patients exceeding clinically relevant thresholds (SOFA > 2 or APACHE-II ≥ 15). For patients with elevated scores, postoperative ICU admission may facilitate early risk stratification and comprehensive supportive care, thereby reducing complication rates and improving postoperative recovery [30, 31]. Patients with lower scores (SOFA ≤ 2 or APACHE-II < 15) did not experience prolonged LOS when managed in the general ward, indicating that standard ward monitoring is sufficient for clinically stable patients, and this may also help to avoid unnecessary ICU utilization while preserving medical resources [7, 32].
In the evaluation of the key factors affecting LOS, the SOFA and APACHE-II scores demonstrated statistically significant predictive value, which was consistent with some other literatures [33, 34]. Notably, patients with higher SOFA scores did not exhibit prolonged LOS in the ICU group, suggesting that high-quality ICU care may mitigate the adverse impact of severe postoperative organ dysfunction. In addition to SOFA and APACHE-II scores, several intraoperative indexes, including blood transfusion, tracheotomy, and bilateral lymph node dissection, were also found to be associated with LOS, supporting previous reports that surgical complexity could be closely linked with prolonged recovery time [35–38]. Age-stratified analyses further revealed that ICU admission and preoperative treatment were key determinants of LOS in older adults, highlighting the importance of preoperative optimization in this population [39]. In contrast, LOS in younger adults less than 65 years was primarily influenced by baseline severity scores and flap transplantation, indicating that underlying health status plays a dominant role in postoperative recovery among younger adults [33, 40, 41]. In summary, these findings stressed the necessity of selecting appropriate postoperative management based on age groups.
Limitation
This study was based on a single-centre design, which might be affected by the factors at the local medical management level and regional level, thus restricting the wide applicability of the study results. The lack of long-term follow-up data after discharge directly caused incomplete assessment of the influence of the ICU and the general ward on prognostic factors, such as postoperative survival status, quality of life, and functional recovery of the body. Additionally, there was insufficient exploration of various complications, including their incidence rates, optimal intervention times, and treatment effects. Moreover, the scoring systems used in the evaluation process were unable to comprehensively capture all potential factors influencing postoperative recovery, leaving certain aspects of patient outcomes unaddressed.
Conclusion
This study validated the prominent role of the ICU in the treatment of high-risk oral cancer patients after surgery. It was revealed that ICU admission was not an independent determinant of LOS in postoperative oral cancer patients. Major influencing factors included intraoperative blood transfusion, tracheotomy, bilateral lymph node dissection, and the SOFA score. For the older adults, ICU transfer and preoperative treatment were the main factors of LOS, while for the younger ones, SOFA and CAPRINI scores, chronic diseases, and other skin flap transplantation were the main factors.
Key Messages for Clinical Practice: ICU admission should not be a routine in postoperative oral cancer patients, while instead it could be guided by objective risk assessment strategies. Several clinical score systems would be applicable in the evaluating procedure, such as SOFA and APACHE-II. As ICU transfer is a key determinant of prolonged hospitalization in the older adults whose age have already reached 65 years old or more, particular caution should be warranted when considering ICU admission for this population group. For the patients with low illness severity (e.g., SOFA ≤ 2), postoperative care in a general ward is a safe and resource-efficient alternative. Clinicians should anticipate a longer LOS in patients requiring tracheotomy, intraoperative blood transfusion, or bilateral neck dissection, regardless of their age or ICU admission status.
Supplementary Information
Acknowledgements
Not applicable.
Authors’ contributions
Junpeng Tang, Yan Huang and Jinzhao Zhang: data curation, formal analysis, methodology, and original draft writing. Fanrong Lin, Siqi Liu and Weipeng Liu: investigation, methodology, and software processing. Jingying Hou and Zhengfei Yang: conceptualization, project administration, proofreading of the writing and editing.
Funding
Guangzhou Key Laboratory of Neuromodulation and Regenerative Medicine Rehabilitation (2024A03J0699).
Data availability
The datasets used during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
This study has been approved by the Ethics Committee of Sun Yat-sen University. All participants in the study have signed informed consent forms.
Consent for publication
Consent for publication have been obtained from all patients.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Junpeng Tang, Yan Huang and Jinzhao Zhang contributed equally to this work.
Contributor Information
Jingying Hou, Email: houjy6@mail.sysu.edu.cn.
Zhengfei Yang, Email: yangzhengfei@vip.163.com.
References
- 1.Zhang SZ, Xie L, Shang ZJ. Burden of Oral Cancer on the 10 Most Populous Countries from 1990 to 2019: Estimates from the Global Burden of Disease Study 2019. Int J Environ Res Public Health. 2022;19(2):875. [DOI] [PMC free article] [PubMed]
- 2.Warnakulasuriya S, Kerr AR. Oral Cancer Screening: Past, Present, and Future. J Dent Res. 2021;100(12):1313–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Yu S, Wei K, Zhou D, Lin Q, Li T. Predictive factors of postoperative complications related to free flap reconstruction in head and neck cancer patients admitted to intensive care unit: a retrospective cohort study. BMC Anesthesiol. 2024;24(1):258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chokshi NJ, Wani S, Bhatt Y, Doshi P, Shah S, Shah A, et al. Analysis of first week complications of microvascular free flap reconstructions after oral cavity cancer resection: A Single Institutional Experience of 314 Cases. J Head Neck Physicians Surg. 2019;7(1):11–5. [Google Scholar]
- 5.Rempel V, Grandoch A, Safi AF, Buller J, Riekert M, Schick V, et al. The prognostic implications of comorbidity and risk factors for (post)operative complications, days spent in the intensive care unit (ICU), and length of hospitalization in patients with oral squamous cell carcinoma: A prospective study. J Craniomaxillofac Surg. 2020;48(9):868–74. [DOI] [PubMed] [Google Scholar]
- 6.Wallner J, Schwaiger M, Edmondson SJ, Mischak I, Egger J, Feichtinger M et al. Effects of Pre-Operative Risk Factors on Intensive Care Unit Length of Stay (ICU-LOS) in Major Oral and Maxillofacial Cancer Surgery. Cancers (Basel). 2021;13(16):3937. [DOI] [PMC free article] [PubMed]
- 7.Mashrah MA, Aldhohrah T, Abdelrehem A, Sabri B, Ahmed H, Al-Rawi NH, et al. Postoperative care in ICU versus non-ICU after head and neck free-flap surgery: a systematic review and meta-analysis. BMJ Open. 2022;12(1):e053667. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Aponte-Ortiz JA, Greenberg-Worisek AJ, Marinelli JP, May M, Spears GM, Labott JR, et al. Cost and clinical outcomes of postoperative intensive care unit versus general floor management in head and neck free flap reconstructive surgery patients. Am J Otolaryngol. 2021;42(5):103029. [DOI] [PubMed] [Google Scholar]
- 9.Cervenka B, Olinde L, Gould E, Farwell DG, Moore M, Kaufman M, et al. Use of a non-ICU specialty ward for immediate post-operative management of head and neck free flaps; a randomized controlled trial. Oral Oncol. 2019;99:104464. [DOI] [PubMed] [Google Scholar]
- 10.Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22(7):707–10. [DOI] [PubMed] [Google Scholar]
- 11.Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med. 1985;13(10):818–29. [PubMed] [Google Scholar]
- 12.Obesity. preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser. 2000;894:i–xii. [PubMed] [Google Scholar]
- 13.Kondrup J, Allison SP, Elia M, Vellas B, Plauth M, Educational, et al. ESPEN guidelines for nutrition screening 2002. Clin Nutr. 2003;22(4):415–21. [DOI] [PubMed] [Google Scholar]
- 14.Caprini JA. Thrombosis risk assessment as a guide to quality patient care. Dis Mon. 2005;51(2–3):70–8. [DOI] [PubMed] [Google Scholar]
- 15.Lumley T, Diehr P, Emerson S, Chen L. The importance of the normality assumption in large public health data sets. Annu Rev Public Health. 2001;23:151–69. [DOI] [PubMed] [Google Scholar]
- 16.Schmidt AF, Finan C. Linear regression and the normality assumption. J Clin Epidemiol. 2017;98:146–51. [DOI] [PubMed] [Google Scholar]
- 17.Hosmer DW Jr, Lemeshow S, Sturdivant RX. Applied logistic regression. 3rd ed. Wiley; 2013.
- 18.Bursac Z, Gauss CH, Williams DK, Hosmer DW. Purposeful selection of variables in logistic regression. Source Code Biol Med. 2008;3:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Jerath A, Austin PC, McCormack D, Wijeysundera DN. Impact of postoperative intensive care unit utilization on postoperative outcomes in adults undergoing major elective noncardiac surgery. J Clin Anesth. 2020;62:109707. [DOI] [PubMed] [Google Scholar]
- 20.Aseni P, Orsenigo S, Storti E, Pulici M, Arlati S. Current concepts of perioperative monitoring in high-risk surgical patients: a review. Patient Saf Surg. 2019;13:32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Pelosi P, Ball L, Schultz MJ. How to optimize critical care resources in surgical patients: intensive care without physical borders. Curr Opin Crit Care. 2018;24(6):581–7. [DOI] [PubMed] [Google Scholar]
- 22.Wong KHF, Murigu AK, Buongiovanni G, Mouton R, Hinchliffe RJ. Intensive care after vascular surgery: systematic review. Br J Surg. 2025;112(8):172. [DOI] [PMC free article] [PubMed]
- 23.Di Nardo M, Tikkanen J, Husain S, Singer LG, Cypel M, Ferguson ND, et al. Postoperative Management of Lung Transplant Recipients in the Intensive Care Unit. Anesthesiology. 2022;136(3):482–99. [DOI] [PubMed] [Google Scholar]
- 24.Yang TR, Luo K, Deng X, Xu L, Wang RR, Ji P. Effect of sarcopenia in predicting postoperative mortality in emergency laparotomy: a systematic review and meta-analysis. World J Emerg Surg. 2022;17(1):36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Colevas AD, Cmelak AJ, Pfister DG, Spencer S, Adkins D, Birkeland AC, et al. NCCN Guidelines(R) Insights: Head and Neck Cancers, Version 2.2025. J Natl Compr Canc Netw. 2025;23(2):2–11. [DOI] [PubMed] [Google Scholar]
- 26.Mathew A, Lockwood MB, Steffen AD, Jiwan Tirkey A, Pavamani S, Patil CL, et al. Distinct Dysphagia Profiles in Patients With Oral Cancer After Surgery. Oncol Nurs Forum. 2023;50(2):201–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Chen WC, Hung KS, Chen SH, Shieh SJ, Lee JW, Hsiao JR, et al. Intensive Care Unit Versus Ward Management After Anterolateral Thigh Flap Reconstruction After Oral Cancer Ablation. Ann Plast Surg. 2018;80(2S Suppl 1):S11–4. [DOI] [PubMed] [Google Scholar]
- 28.Chiche PL, Avry F, Drouet J, Weill P, Benateau H, Qassemyar Q, et al. Facial Artery Perforator Flap for Perioral and Perinasal Oncologic Defect Reconstruction: Surgical Technique and Postoperative Outcomes. Microsurgery. 2024;44(6):e31238. [DOI] [PubMed] [Google Scholar]
- 29.Shander A, Corwin HL, Meier J, Auerbach M, Bisbe E, Blitz J et al. Recommendations From the International Consensus Conference on Anemia Management in Surgical Patients (ICCAMS). Ann Surg. 2023; 277(4):581 – 90. [DOI] [PMC free article] [PubMed]
- 30.Grant CA, Dempsey GA, Lowe D, Brown JS, Vaughan ED, Rogers SN. APACHE II scoring for the prediction of immediate surgical complications in head and neck cancer patients. Plast Reconstr Surg. 2007;119(6):1751–8. [DOI] [PubMed] [Google Scholar]
- 31.de Ribeiro CB, Kowalski K. APACHE II, POSSUM, and ASA scores and the risk of perioperative complications in patients with oral or oropharyngeal cancer. Arch Otolaryngol Head Neck Surg. 2003;129(7):739–45. [DOI] [PubMed] [Google Scholar]
- 32.Yalamanchi P, Thomas WW, Workman AD, Rajasekaran K, Chalian AA, Shanti RM, et al. Value of Intensive Care Unit-Based Postoperative Management for Microvascular Free Flap Reconstruction in Head and Neck Surgery. Facial Plast Surg Aesthet Med. 2021;23(1):49–53. [DOI] [PubMed] [Google Scholar]
- 33.Houthooft R, Ruyssinck J, van der Herten J, Stijven S, Couckuyt I, Gadeyne B, et al. Predictive modelling of survival and length of stay in critically ill patients using sequential organ failure scores. Artif Intell Med. 2015;63(3):191–207. [DOI] [PubMed] [Google Scholar]
- 34.Milic M, Goranovic T, Holjevac JK. Correlation of APACHE II and SOFA scores with length of stay in various surgical intensive care units. Coll Antropol. 2009;33(3):831–5. [PubMed] [Google Scholar]
- 35.Runge A, Vales A, Pommer G, Raab H, Prossliner H, Randhawa A, et al. Perioperative Blood Transfusion in Head and Neck Cancer Revisited. Laryngoscope. 2023;133(7):1638–44. [DOI] [PubMed] [Google Scholar]
- 36.Riekert M, Rempel V, Keilwerth S, Zoller JE, Kreppel M, Schick VC. Airway-Associated Complications With and Without Primary Tracheotomy in Oral Squamous Cell Carcinoma Surgery. J Craniofac Surg. 2023;34(1):279–83. [DOI] [PubMed] [Google Scholar]
- 37.Fang KH, Kang CJ, Lee LY, Ng SH, Lin CY, Chen WC, et al. Prognostic impact of elective tracheotomy in resected oral cavity squamous cell carcinoma: A nationwide cohort study. Cancer Med. 2024;13(12):e7213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Eskander A, Dziegielewski PT, Patel MR, Jethwa AR, Pai PS, Silver NL, et al. Oral Cavity Cancer Surgical and Nodal Management: A Review From the American Head and Neck Society. JAMA Otolaryngol Head Neck Surg. 2024;150(2):172–8. [DOI] [PubMed] [Google Scholar]
- 39.Raab G, Restifo D, McBride SM, Wong RJ, Lee NY, Shahrokni A, et al. Outcomes following head and neck cancer surgery among older adults as determined by an electronic geriatric assessment. J Geriatr Oncol. 2022;13(2):228–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Dort JC, Farwell DG, Findlay M, Huber GF, Kerr P, Shea-Budgell MA, et al. Optimal Perioperative Care in Major Head and Neck Cancer Surgery With Free Flap Reconstruction: A Consensus Review and Recommendations From the Enhanced Recovery After Surgery Society. JAMA Otolaryngol Head Neck Surg. 2017;143(3):292–303. [DOI] [PubMed] [Google Scholar]
- 41.de Melo GM, Ribeiro KC, Kowalski LP, Deheinzelin D. Risk factors for postoperative complications in oral cancer and their prognostic implications. Arch Otolaryngol Head Neck Surg. 2001;127(7):828–33. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used during the current study are available from the corresponding author on reasonable request.