In the present study, the incidence of POD in elderly patients with oral cancer was 26.5%, which was higher than that reported in previous studies [10, 20]. This discrepancy may be related to the fact that the population in this study was elderly and required free flap reconstruction with extensive and prolonged surgery. Most of the POD cases (93.7%) were observed within 1–5 days postoperatively. This suggests that clinicians should pay special attention to the possibility of POD during this time. Notably, in order to reduce the incidence of POD, it is crucial to have an early, validated and accurate tool to assess adverse outcomes.
We retrospectively analysed the clinical data and laboratory indicators of 359 elderly oral cancer patients. Through a series of statistical methods, we identified age, sex, alcohol consumption history, marriage, preoperative anxiety, preoperative sleep disorder, and the length of ICU stay as independent influences on POD in elderly oral cancer patients. We constructed an intuitive and accurate nomogram and performed internal validation, demonstrating the good discriminatory and clinical applicability of the model.
In this study, it was found that the older the elderly patients with oral cancer, the more likely they were to develop POD, and this result was consistent with the study of Zhao et al [6]. The organism undergoes a series of changes in the brain during aging, such as changes in stress-regulated neurotransmitters, decreased cerebral blood flow, decreased cerebral vascular density, neuronal apoptosis, and alterations in cellular signal transduction systems. Therefore, the risk of delirium is higher [21, 22].
Male is an independent predictor of POD in elderly oral cancer patients. Our findings may suggest that men do not handle postoperative psychological stress well, as well as more postoperative complications such as infections and negative emotions [23]. Crawford et al[24] reported that a study of head and neck oral cancer in the West of Scotland found that male patients were more likely to develop POD than female patients. A systematic evaluation of risk factors for POD in Stanford type A aortic coarctation suggested that the risk of POD in male patients was 1.33 times higher than in women [25]. However, the gender factor is still controversial. Hasegawa et al. showed no association between males and POD in oral cancer patients [10]. Future studies could further explore the relationship between POD and gender.
The drinking problem is more prominent in China. Monitoring data revealed an upward trend in alcohol consumption among the Chinese adult population from 2015 to 2017, with its drinking rate reaching 43.7%[2]. Alcohol consumption history is one of the independent risk factors for POD in elderly oral cancer patients. This finding has been confirmed by previous studies [26]. In the case of alcohol, the toxic by-product acetaldehyde can bind to DNA, disrupt cell replication and increase the body's susceptibility to other carcinogens. What's more, there is a synergistic effect of tobacco and alcohol. The dangers of the two combined are exponentially greater. In addition, these two behavioural risks are often co-existing and interrelated [27].
We found that widowed and unmarried elderly oral cancer patients are vulnerable to POD, consistent with previous studies [28]. Marriage may be particularly important for patients with head and neck cancers, given that surgery and/or radiotherapy for head and neck cancers are often associated with local disease recurrence and significant long-term dysfunction in speech, communication and swallowing. Recent studies have also confirmed that for patients with oral cancers, marriage is associated with early stage, aggressive treatment and superior survival rates. [29].
Preoperative anxiety and preoperative sleep quality are also important risk factors for POD in elderly oral cancer patients. Preoperative anxiety has been suggested as a possible predisposing factor [30]. Anxiety may be associated with higher glucocorticoid concentrations. And metabolic disorders are also well known mechanisms leading to POD [31]. In addition, there is evidence that anxiety can promote the production of pro-inflammatory cytokines, which has been proven to be a potential marker of POD [32]. What's more, anxiety may lead to sleep disturbances. Preoperative sleep quality have long been associated with the development of POD as well [33]. Liu et al. identified that patients with preoperative sleep disorders undergoing craniotomy were 2.7 times more likely to develop POD than patients without sleep disorders[34], which is similar to the results of this study. Sleep disorders, particularly sleep fragmentation and poor sleep quality, are common among older adults. There is growing evidence that sleep disorders are associated with impairments in spatial memory, verbal fluency, attention, and executive function[35]. Therefore, the relationship between preoperative anxiety, preoperative sleep quality, and POD warrants continued research.
A study of elderly gastric cancer patients reported a positive correlation between prolonged ICU stay and POD [36]. This is supported by our findings. ICU is a psychologically challenging environment. ICU patients may be frightened by the occasional shrill alarm. If a patient has a tracheotomy, then they are unable to communicate. Many patients have a urinary catheter in their urethra and are physically restrained [37].
The strengths of this study are threefold. Firstly, a wider range of independent risk factors were included in this study. All these factors were available in a timely and direct manner after hospital admission, which ensured the simplicity and timeliness of the model. Second, the use of simple and objective clinical data to construct predictive models facilitates their application to clinical practice. Finally, POD is a common problem that has been well explored in Western countries. However, there are still few reports on POD in China. We hope that our findings will fill the gaps in the incidence and risk factors of POD in elderly oral cancer patients.
Several limitations should be considered. This study is a single-centre cross-sectional investigation. Therefore, the results obtained need to be further confirmed by the results of multi-centre and large-sample studies to look for more risk factors and to take early measures to prevent the occurrence of POD and slow down its development. In addition, only elderly oral cancer patients were analysed in this study. Further studies are needed to determine if this is applicable to other populations. Although internal validation assessed the robustness of the model, the nomogram model was not validated against an external dataset, which may limit the generalisability of our findings.