Predicting complications in elderly patients undergoing oral cancer resection with free flap reconstruction in China: a retrospective cohort study using the modified Frailty Index and Prognostic Nutritional Index

Abstract

Objectives

This study aimed to evaluate the predictive abilities of the 5-item modified Frailty Index (5-mFI), Prognostic Nutrition Index (PNI), and their combination in older adult patients undergoing oral cancer resection and free flap reconstruction.

Design

Retrospective cohort study.

Setting

Secondary care involving multiple centres treating older adult patients for oral cancer.

Participants

This study included a total of 1197 patients aged ≥60 years who underwent oral cancer resection with free flap reconstruction between January 2014 and December 2022. The study included patients aged ≥60 years with malignant tumours who underwent selective radical surgery, such as mandibulectomy, maxillectomy, glossectomy or laryngectomy, followed by free flap reconstruction under general anaesthesia. Exclusion criteria included the presence of any inflammatory disease affecting blood test results, incomplete clinical records or missing data for any of the five items in the 5-mFI. Patients were categorised into four groups based on PNI and 5-mFI values: (1) ‘Control’ (neither frail nor malnourished), (2) ‘Frailty’ (frail only), (3) ‘Malnutrition’ (malnourished only) and (4) ‘Frailty+Malnutrition’ (both frail and malnourished).

Primary and secondary outcome measures

The primary outcome was the rate of complications within 30 days after surgery. Secondary outcomes included unplanned reoperation rates, length of postoperative hospital stay and the predictive performance of PNI, 5-mFI and their combination.

Results

The overall complication rate within 30 days post-surgery was 34.6%. The Frailty+Malnutrition group exhibited the highest risk of complications, longer postoperative hospital stays and increased rates of unplanned reoperation compared with the Control, Frailty and Malnutrition groups. The combined PNI and 5-mFI model significantly improved the predictive value for postoperative complications compared with either PNI or 5-mFI alone.

Conclusions

Older adult patients undergoing oral cancer resection with free flap reconstruction face considerable risk from frailty and malnutrition. Although both 5-mFI and PNI independently demonstrated good predictive abilities for postoperative complications, the combined model provided the best prediction. These findings could help optimise preoperative management in this high-risk population.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study retrospectively analysed a large cohort of 1197 older adult patients, enhancing statistical power and generalisability.

• Incorporation of the 5-item modified Frailty Index and Prognostic Nutrition Index provided a rigorous and comprehensive risk assessment model.

• All data were collected from a single centre, which may limit the generalisability of the findings.

• The retrospective design of the study introduces potential selection bias and reliance on existing records.

• Postoperative complications were only assessed during hospitalisation, potentially underestimating the true incidence.

INTRODUCTION

Globally, neoplasms of the oral cavity and oropharynx rank as the seventh most prevalent type of cancer and the ninth leading cause of cancer-related mortality.¹ Approximately 69% of oral and oropharyngeal cancers are reported in patients aged ≥60 years in China.² Currently, oral cancer is primarily treated through surgery, which includes tumour removal, selective neck dissection and functional reconstruction.³ The application of free flap surgery aims to restore and repair tissue defects, with the primary purpose being to enhance the patient’s quality of life.⁴ However, this particular surgical procedure is complex, time-consuming and carries a substantial risk of postoperative adverse events.⁵ Although geriatric patients are more prone to complications and prolonged hospitalisation due to decreased physiological reserve capacity, frailty and malnutrition,6,8 they should not be excluded from reconstructive oral cancer surgery. Therefore, preoperative assessment and optimisation should be carefully undertaken.

Mouth and tongue cancers are the most frequently diagnosed cancers and the leading causes of death among all oral cancers in China.² Tumours located in the oral cavity and oropharynx cause difficulties with swallowing and pain, which often result in inadequate oral intake, unintentional weight loss and loss of muscle mass. As a result, these patients frequently experience malnutrition.⁹ The Prognostic Nutritional Index (PNI), first proposed by Buzby et al,¹⁰ was originally used to predict the risk of patients undergoing gastrointestinal surgery. It is defined by serum albumin concentration and peripheral blood lymphocyte count and has been used to evaluate the nutritional status of patients with cancer.11,13 However, PNI solely assesses indicators of nutritional and immunological well-being, disregarding a patient’s comorbidities and functional capacity. The concept of frailty has been introduced in recent years to address the discrepancy between a patient’s chronological and physiological age.¹⁴ A diminishing physiological reserve characterises frailty and can be assessed using different models.^{15 16} Frailty assessments generally fall into two broad categories: cumulative deficit models and phenotypic models. The Frailty Index, including the modified Frailty Index (mFI), represents a cumulative deficit model, where frailty is quantified based on the accumulation of various health deficits.¹⁶ The 5-item mFI (5-mFI),¹⁷ a streamlined version developed from the original 70-item scale,^{16 18} maintains its predictive capacity while offering simplicity and ease of use.¹⁹ This index has been extensively validated in older adult patients undergoing various surgeries, including minimally invasive partial nephrectomy,²⁰ abdominal surgery,²¹ and colorectal cancer surgery,²² proving its reliability in predicting patient outcomes.

There are very few studies on the use of 5-mFI in older adult patients undergoing oral and oropharyngeal surgery with free flap repair. Abt’s study focused on the predictive effect of mFI on Clavien-Dindo grade IV intensive care complications and grade V complications (mortality) after major surgery for head and neck cancer. However, it lacked a comprehensive description of overall postoperative complications, without considering nutritional factors.²³ In addition, because of the possible coexistence of malnutrition and frailty in patients with oral cancer, it is necessary to screen for both conditions.⁹ Based on this, we hypothesised that the combination of PNI and 5-mFI would better reflect nutritional and frailty statuses and improve predictive significance. This combination, which incorporates both frailty and nutritional status indicators, is expected to address the limitations of Abt et al’s work, which primarily focused on critical complications without accounting for nutritional factors, thereby expanding on Abt et al’s findings. Therefore, this large sample study aimed to determine whether PNI and 5-mFI could reliably predict the occurrence of postoperative complications in older adult patients with oral cancer undergoing free flap reconstruction. In addition, we developed a new screening model that combines PNI with 5-mFI to improve predictive performance.

Materials and methods

Patient and public involvement statement

Patients were not involved in the design, conduct, reporting or dissemination of the research.

Study design and patients

This retrospective cohort study enrolled older adult patients admitted to Shanghai Ninth People’s Hospital, affiliated with Shanghai Jiao Tong University School of Medicine, between 1 January 2014 and 31 December 2022. The study adhered to the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis principle,²⁴ for developing prediction models. All patients were diagnosed with oral or oropharyngeal cancer.

The inclusion criteria were: (1) age ≥60 years, (2) malignant tumours and (3) having undergone selective radical surgery, such as mandibulectomy, maxillectomy, glossectomy and laryngectomy, followed by free flap reconstruction under general anaesthesia. These procedures were selected for their relevance to the surgical management of oral and oropharyngeal cancers. The exclusion criteria were: (1) any inflammatory disease that could affect the results of blood testing, (2) incomplete clinical records and (3) missing at least one item from the five items in the 5-mFI. Detailed enrolment information is shown in online supplemental figure 1.

Data collection

The data were obtained from the Union Digital Medical Record Browser (V.2012.4; Unionnet, Shanghai, China) at our hospital. Demographic data were retrospectively collected from medical records, including age, body mass index (BMI), 5-mFI, PNI, duration of surgery, American Society of Anesthesiologists Physical Status (ASA) score, history of radiotherapy and chemotherapy, smoking and alcohol consumption habits and previous tumours during hospitalisation.

As previously mentioned, the 5-mFI was calculated based on the number of concurrent medical conditions: hypertension (managed with a standardised pharmaceutical regimen), diabetes, chronic obstructive pulmonary disease (COPD), history of congestive heart failure (CHF) within 30 days prior to surgery and non-independent functional status.²⁵ The American Cancer Society’s National Surgical Quality Improvement Programme defines functional dependency as the need for assistance from others in performing daily activities, such as dressing, bathing, using the toilet, eating and walking, during the 30 days prior to surgery. The assessment of non-independent functional status was derived from the Barthel Index Rating Scale, as documented in the nursing evaluation sheet.²¹ For each item, patients were assigned scores of either 0 (indicating non-compliance with frailty criteria) or 1 (indicating compliance with frailty criteria). The cumulative scores of the five tasks were then calculated. According to previous literature, patients with a score of 0 or 1 were categorised as non-frail; in contrast, those with a score ≥2 were classified as frail.^{26 27}

PNI was calculated as previously reported: PNI = (10 × albumin (g/dL)) + (0.05 × absolute preoperative lymphocyte count (cells per mm³)).²⁸

We established four patient groups, categorised as follows: ‘Control’ with high PNI and 5-mFI <2, ‘Malnutrition’ with low PNI and 5-mFI <2, ‘Frailty’ with high PNI and 5-mFI ≥2 and ‘Malnutrition+Frailty’ with low PNI and 5-mFI ≥2.

Outcome and complication data

Data were collected on various postoperative outcomes that occurred within 30 days following the principal surgical procedure. These postoperative complications included: pulmonary complications (pneumonia, pulmonary embolism, hypoxaemia, acute respiratory failure),²⁹ flap complications (infection, bleeding, flap crisis) acute renal injury (an absolute increase in plasma creatinine of 26.5 µmol/L within 48 hours or a 1.5-fold increase in creatinine within 7 days),³⁰ hepatic insufficiency (an increase in serum total bilirubin after hepatectomy ≥7 mg/dL),³¹ cardiac insufficiency (CHF, permanent atrial fibrillation), postoperative delirium (POD), deep vein thrombosis, cerebrovascular incidents (cerebral embolism, cerebral haemorrhage) and acute pancreatitis. The resource utilisation parameters included the duration of intensive care unit (ICU) stay, the need for respiratory support beyond 48 hours, length of postoperative hospitalisation, unplanned surgical reintervention and mortality rate.

Statistical analysis

Statistical analyses were performed using SPSS (V.26; IBM, Armonk, New York, USA). One-way analysis was employed initially to measure and compare the occurrence of each complication across the groups. Multivariate regression analysis was then conducted to assess differences between the various cohorts. Age, sex, BMI, ASA score and history of smoking and alcohol consumption were controlled for in our model. Comparisons were made between groups for complications, length of postoperative hospital stays, ICU stay duration, unplanned reoperations and mortality rates. The Control group served as the reference level for comparison. Subgroup analysis was conducted according to the cancer subsite and the type of free flap. Comparisons were made using the Malnutrition group or the Frailty group as the reference level against the Malnutrition+Frail group. A comprehensive risk assessment model was created by integrating 5-mFI and PNI. The predicted abilities of PNI and 5-mFI were evaluated using the receiver operating characteristic (ROC) curve. We constructed a new model by incorporating all variables used to evaluate PNI and 5-mFI into a logistic regression model. These variables included preoperative albumin, absolute preoperative lymphocyte count, hypertension (managed with a standardised pharmaceutical regimen), diabetes, COPD, history of CHF within 30 days prior to surgery and non-independent functional status). The datasets were randomly divided into training set (70%), and test set (30%). Five-fold cross-validation was performed on the training set to construct the model. The test set was used to evaluate the predictive ability of this new model. The DeLong Method was used to compare the significance of differences between the ROC curves. A two-tailed p value of<0.05 indicated statistical significance.

Results

A total of 1197 patients were included in this study. Among these, 54.0% of patients were neither frail nor had low PNI; 24.6% and 10.6% were identified purely as frail or with low PNI, respectively; and 10.8% were both frail and had low PNI.

Notably, significant differences were observed in age, BMI, previous tumours, smoking and alcohol consumption history and ASA scores across these groups (p<0.05, online supplemental table 1). The occurrence rate of any complication within 30 days after surgery was 34.6%. The most common complications were pulmonary complications (16.5%), POD (8.1%), flap complications (6.9%), cardiac insufficiency (5.6%) and hepatic insufficiency (5.3%) (online supplemental table 2). The rate of any complication in the Malnutrition+Frailty group (70.5%) was significantly higher compared with that in the Control (18.1%), Malnutrition (45.1%) and Frailty (57.5%) groups (p<0.001, online supplemental table 2). The Malnutrition+Frailty group had the highest occurrence rate of each complication among the four groups; in contrast, the Control group had the lowest. The multivariate model showed that the Malnutrition and Frailty groups had a higher risk of pulmonary and flap complications, as well as cardiac insufficiency, compared with the Control group (p<0.05, table 1).

Table 1. Multivariate analysis of Malnutrition, Frailty, Malnutrition+Frailty patients compared with Control.

Complication	Malnutrition		Frailty		Malnutrition+Frailty
	OR, 95% CI	P value	OR, 95% CI	P value	OR, 95% CI	P value
Any complication	3.82 (2.81 to 5.2)	<0.001	6.57 (4.34 to 10.03)	<0.001	12.08 (7.7 to 19.3)	<0.001
Flap complication	3.32 (1.89 to 5.93)	<0.001	2.85 (1.24 to 6.16)	0.01	6.61 (3.29 to 13.26)	<0.001
Pulmonary complication	3.4 (2.25 to 5.19)	<0.001	6.01 (3.65 to 9.9)	<0.001	9.46 (5.77 to 15.66)	<0.001
Acute renal injury	2.23 (0.61 to 8.12)	0.209	3.2 (0.63 to 13.65)	0.124	5.36 (1.33 to 21.5)	0.016
Hepatic insufficiency	2.55 (1.4 to 4.67)	0.002	1.83 (0.7 to 4.27)	0.183	2.8 (1.18 to 6.26)	0.015
Cardiac insufficiency	2.35 (1.17 to 4.72)	0.015	6.49 (3.22 to 13.23)	<0.001	3.97 (1.75 to 8.82)	<0.001
Postoperative delirium	1.5 (0.89 to 2.49)	0.123	1.44 (0.69 to 2.8)	0.305	1.99 (1.03 to 3.75)	0.036
Deep venous thrombosis	0.87 (0.24 to 2.65)	0.821	1.49 (0.32 to 5.11)	0.561	3.58 (1.16 to 10.53)	0.022
Unplanned reoperation	1.2 (0.53 to 2.59)	0.651	0.95 (0.22 to 2.91)	0.933	4.24 (1.82 to 9.65)	<0.001
Death	0.68 (0.03 to 5.45)	0.744	4.14 (0.51 to 27.1)	0.136	1.87 (0.08 to 17.9)	0.619
Postoperative days	1.65 (1.04 to 2.69)	0.039	1.17 (0.65 to 2.26)	0.614	3.95 (1.67 to 11.68)	0.005
Ventilator support	1.08 (0.54 to 2.05)	0.82	1.82 (0.85 to 3.69)	0.108	1.71 (0.77 to 3.59)	0.17

Compared with the Control group, the Malnutrition+Frailty group had a higher rate of pulmonary and flap complications, acute renal injuries, hepatic and cardiac insufficiency, POD, deep vein thromboses, unplanned reoperations and longer postoperative hospital stays (p<0.05, table 1). In the subgroup analysis of cancer subsite and free flap type, the interaction p values did not reveal a statistically significant trend (interaction p values were 0.521 and 0.338, respectively, online supplemental table 3). This indicates that there was no clear association between the occurrence of postoperative complications in frail, malnourished or frail+malnourished older patients and their surgical approach, which contrasting with findings from Abt et al’s study. Compared with the Frailty and Malnutrition groups, more complications occurred in the Malnutrition+Frailty group (OR: 2.88, 95% CI: 1.8 to 4.68, p<0.001; and OR: 1.83, 95% CI: 1.08 to 3.14, p<0.026, respectively). The Malnutrition+Frailty group also had more unplanned reoperations and longer postoperative hospital stays compared with the Frailty and Malnutrition groups (p<0.05, table 2).

Table 2. Multivariate analysis of Frail+Malnourished patients compared with referent categories.

Complication	Frailty		Malnutrition
	OR, 95% CI	P value	OR, 95% CI	P value
Any complication	1.83 (1.08 to 3.14)	0.026	2.88 (1.8 to 4.68)	<0.001
Flap complication	2.37 (1.06 to 5.63)	0.04	1.79 (0.92 to 3.45)	0.084
Pulmonary complication	1.58 (0.93 to 2.7)	0.095	2.81 (1.72 to 4.62)	<0.001
Acute renal injury	1.69 (0.38 to 8.81)	0.496	2.27 (0.56 to 9.06)	0.239
Hepatic insufficiency	1.18 (0.42 to 3.45)	0.757	0.99 (0.42 to 2.21)	0.98
Cardia insufficiency	0.63 (0.29 to 1.34)	0.233	1.68 (0.71 to 3.9)	0.23
Postoperative delirium	1.45 (0.66 to 3.29)	0.356	1.18 (0.57 to 2.36)	0.645
Deep vein thrombosis	2.15 (0.53 to 10.85)	0.302	2.75 (0.7 to 11.9)	0.153
Cerebrovascular incidents	0.96 (0.17 to 5.43)	0.96	2.07 (0.32 to 12.88)	0.422
Unplanned reoperation	4.26 (1.28 to 19.34)	0.03	2.68 (1.03 to 7.05)	0.043
Death	0.27 (0.01 to 3.24)	0.319	1.69 (0.05 to 50.35)	0.737
Postoperative days	3.77 (1.34 to 12.42)	0.017	3.22 (1.19 to 10.44)	0.032
Ventilator support	0.98 (0.41 to 2.36)	0.971	1.6 (0.65 to 3.88)	0.297

The ROC curve of the PNI was similar to that of the 5-mFI in terms of predicting postoperative complications (PNI area under the curve (AUC): 0.66; 5-mFI AUC: 0.65; p=0.586). However, the predictive model integrating 5-mFI and PNI demonstrated superior performance (AUC: 0.80), with a sensitivity of 75% and a specificity of 85%, compared with PNI (p<0.001) or 5-mFI alone (p<0.001) (figure 1).

Figure 1. ROC curves for PNI, Frailty and PNI+Frailty. AUC, area under the curve; PNI, Prognostic Nutrition Index; ROC, receiver operating characteristic.

Discussion

There is a limited understanding of the interaction between malnutrition and frailty and their impact on older adult patients undergoing oral cancer resection with free flap reconstruction. Therefore, in the present investigation, we examined the association between PNI and 5-mFI and clinical outcomes. The results indicated a noteworthy association between low PNI and frailty with an unfavourable prognosis. Furthermore, our findings suggest that our innovative predictive model, which amalgamated PNI and 5-mFI, demonstrated enhanced predictive efficacy compared to the application of PNI or 5-mFI in isolation.

In our study, 24.6% of patients were identified as having low PNI, 10.6% exhibited frailty and 10.8% were classified as having both frailty and malnutrition. The data showed that only one-third of patients with malnutrition were frail. In addition, the occurrence of frailty was higher in patients with malnutrition compared with the occurrence of malnutrition in patients with frailty. This suggests that these two conditions were coexistent but distinct.⁹ In our study, the rate of frailty based on 5-mFI (≥2) in older adult patients was higher than that of patients undergoing total hip arthroplasty (15%),³² meningioma surgery (16%)²⁷ and gynaecological surgery (6.5%);³³ however, it was lower than that of patients undergoing colorectal cancer surgery (45%).³⁴ ROC curve analysis identified 44.85 as the most suitable cut-off value for PNI. Our rates of low PNI (35.4%) were similar to those previously reported in patients with oral cancer.35,37 These results confirm that patients with oral cancer may have chronic nutritional problems due to odynophagia and dysphagia.

In this study, we discovered a population particularly vulnerable to malnutrition and frailty. Patients in this cohort tended to be older, have a history of tumours, engage in smoking or alcohol consumption and have a higher ASA score. Interestingly, our study showed that the BMI was lower in the Malnourished group and higher in the Frail group; however, it was normal in the Malnourished+Frail group. We believe this contradictory finding is due to the significant differences in fat mass and muscle mass across all BMI strata in patients with cancer.^{38 39} The effectiveness of BMI is low in older adults due to age-related changes in body composition.^{40 41} In studies of patients with cancer who tend to be older and whose weight and body composition change as a result of cancer, the low sensitivity of BMI to identify patients with obesity and its inability to distinguish between fat and muscle mass significantly challenged its usefulness.⁴² Therefore, BMI alone is insufficient for assessing nutrition and frailty status, particularly in older adults with cancer.

Regardless of whether the patients were malnourished, frail or both, they were significantly more likely to experience complications compared with Control group patients. Moreover, the Malnutrition+Frailty group had the highest rates of complications. These results were alarming but not surprising. Similar to other microvascular reconstructions, oral cancer resections with free flap repairs require longer operative times and extended bed rest postoperatively. Both the Malnutrition and Frailty groups may have had more comorbidities and poorer nutritional or function status, resulting in more complications. Furthermore, in the Malnutrition+Frailty group, there were significantly more unplanned reoperations and longer postoperative hospital stays. These findings clearly demonstrate that the combination of malnutrition and frailty has a synergistic detrimental effect on older adult patients undergoing oral cancer resection with free flap reconstruction.

Wang et al conducted a retrospective cohort study of 254 patients aged >60 years undergoing oral cancer surgery with flap reconstruction.⁴³ The postoperative complications within 30 days were similar in the frail (5-mFI=0) and non-frail (5-mFI=1) groups. Due to the small sample size, that study did not include patients with a higher Frailty Index for analysis. The statistically significant results of our study may be attributed to the criteria we chose to determine frailty (5-mFI≥2) and the larger sample size. Thus, although geriatric patients with frailty or malnutrition can tolerate oral cancer surgery with flap reconstruction, surgeons and anaesthetists should consider the patient’s functional and nutritional status when developing personalised preoperative management plans. This will help minimise the negative impact of frailty and malnutrition on postoperative recovery.

In this study, pulmonary complications (16.5%) and POD (8.1%) were the most common postoperative complications. The higher rate of postoperative respiratory complications was due to extended bed rest after surgery, tracheotomy and a reduction in the ability of patients with oral cancer resection to cough up phlegm.⁴⁴ Moreover, the incidence of pulmonary complications was highest in the Malnutrition+Frailty group. This could be attributed to lower preoperative albumin levels, more cardiopulmonary system disease and older age in these patients, all of which have been identified as risk factors for pulmonary complications after oral cancer resection with free flap reconstruction.^{29 44} The occurrence of POD reported in one meta-analysis was 4.2–36.9%, with a pooled incidence of 20%,⁴⁵ which is higher than the incidence of POD in our study. Owing to its rapid onset and fluctuating course, we cannot rule out the possibility that the occurrence of POD may sometimes have been omitted from medical records in our study.

Although both PNI and 5-mFI were suitable predictors for prognosis, they were not perfect. On the one hand, this was because the two indicators focus on different aspects, which might overlap but do not equate. On the other hand, these two indicators themselves have certain limitations. For example, it has been reported that 5-mFI is more useful in patients aged <65 years than in older adult patients when predicting the risk of postoperative outcomes.²⁷ In addition, the definition of frailty is not fully understood. It is also important to note that many aspects of frailty can potentially be modified. Current frailty assessments do not distinguish whether blood pressure or blood sugar is well-controlled during the preoperative period.³³ Some studies have suggested that hypertension, which has been managed with a standardised pharmaceutical regimen, is a strong mitigating factor influencing surgical complications.⁴⁶ The same is true for nutritional assessments—there is no unified consensus on the use of nutrition scales. One large prospective study reported that the nutritional risk index, calculated according to albumin and body weight, was superior to PNI for nutritional assessments in patients with oral cancer.³⁵ Hence, we integrated 5-mFI and PNI to establish a risk prediction model to account for these deficiencies. Unsurprisingly, the combined model significantly enhanced the ability to predict postoperative complications in geriatric patients undergoing oral cancer resection with free flap construction. PNI can be calculated using serum albumin and lymphocyte count, which are often collected preoperatively as routine blood and liver function tests. Similarly, 5-mFI can also be easily evaluated. This model can effectively help physicians optimise perioperative management, refine clinical pathways and improve clinical prognosis without increasing their effort or workload. Further research is required to understand how and to what extent low nutritional status can be improved in patients with oral cancer.

Limitations

This study has several limitations. Primarily, all the data were retrospective and largely reliant on single-centre cohorts, which could introduce selection bias or bias due to incomplete data. Fortunately, the database we used implements effective measures to ensure data quality, and the sample size of our study was relatively large. Furthermore, this study only considered postoperative complications during hospitalisation. Complications that arose after discharge were not included in the analysis, which may have led to an underestimation of the true incidence. Lastly, our study population included different cancer subsites, various neoplasm stages and different free flap types. We did not perform further subgroup analysis on neoplasm stage, which could have influenced our results. However, as the aim of the study was to evaluate and compare the predictive abilities of two widely used indicators—5-mFI and PNI, along with their combination—we believe this study remains significant, although further comparative prospective trials are needed.

Conclusion

In this study, we demonstrated that both 5-mFI and PNI are effective predictive tools for assessing the likelihood of short-term complications in older adult patients following oral cancer resection with free flap reconstruction. By integrating these two assessments into a comprehensive risk prediction model, we were able to predict postoperative adverse events with greater accuracy and objectivity. The use of this multivariate assessment technique significantly contributes to the early identification of patients at increased risk of postoperative complications, enabling optimised preoperative preparations and reducing the likelihood of postoperative adverse events.

Data availability statement

Data are available upon reasonable request.