Degenerative spinal diseases are increasing worldwide [1], as is the number of surgeries; degenerative spinal surgeries increased by 62.3% from 2004 to 2015 in the USA [2] and by 54% from 1999 to 2013 in Norway [3]. Moreover, the number of older people who have undergone spinal surgery has steadily risen over the years [1, 3, 4]. Postoperative complications after spinal surgery include cardiovascular, respiratory, and wound-related complications; inpatient falls; and postoperative delirium [4]. It is worth noting that the incidence of postoperative complications of spinal surgery has remained stable from 2006 to 2016, whereas that of postoperative delirium has risen in the USA [4].

Delirium is a state characterized by a sudden change in the consciousness level, which may be accompanied by loss of brain function with attention deficit, memory loss, and abnormal behavior [5]. Previous studies have revealed that 8% to 40% of patients who had undergone spinal surgery had postoperative delirium at age 70 [6]. Postoperative delirium was associated with complications including wound infection, deep vein thrombosis, pneumonia, and pressure ulcers [7]. Furthermore, in the long term, it can lead to impaired functional recovery, decreased quality of life, declined cognitive function, and increased mortality [8].

Age is a significant risk factor for postoperative delirium and the likelihood of developing delirium increases in older patients when one or two risk factor precipitants are added [8]. The predisposing factors for delirium after spinal surgery in older adults include opioid use, hypertension, cerebrovascular disease, and pulmonary disease. The precipitating factors include the site, method, duration of surgery, and infused intravenous fluid volume [6]. A comprehensive geriatric assessment and patient-specific interventions based on preoperative assessment can reduce the incidence of postoperative delirium, related complications, and length of hospital stay [9, 10]. Therefore, it is necessary to identify delirium risk factors to stratify the risk groups and tailor interventions to the patient’s characteristics.

In particular, pain, frailty, activities of daily living, and the nutritional status of older adults undergoing orthopedic surgery affect postoperative complications [11]. However, previous studies of spinal surgery patients have only investigated fragmentary risk factors for postoperative delirium, such as disease-related and laboratory data and surgery-related factors [6]. Therefore, it is necessary to determine whether certain characteristics of older adults who have had spinal surgery affect delirium. The treatment of delirium is centered on preoperative assessment of the patient [11] and prevention and management of the disease [10]. However, the pathophysiology of postoperative delirium remains unclear [8], and there is no solid evidence of medical treatment [10]. Hence, investigating biomarkers is crucial as they can help diagnose and understand this condition.

Previous studies that have explored the biomarkers of postoperative delirium were based on the relatively new hypothesis that the condition originated from brain injury [12, 13, 14]. Digital enzyme-linked immunosorbent assay is highly sensitive and can accurately quantify low concentrations of proteins or peptides with excellent reproducibility, which makes it advantageous for identifying biomarkers [15, 16]. Plasma tau [13, 14] and neurofilament light (NFL) [12] have been identified as biomarkers of delirium using Single-Molecule Array (Simoa) analysis. However, collecting blood samples from older adults and patients experiencing hyperactive delirium is challenging because of low blood pressure, dehydration, and loss of vein patency and elasticity. Thus, biomarkers must be identified using samples obtained through minimally invasive techniques, such as urine sampling.

Extracellular vesicles (EVs) are 40~100 nm extracellular endoplasmic reticulum that retains cell characteristics and information; they have recently garnered attention as potential biomarkers of cancer and neurodegenerative diseases [17, 18]. EVs can easily be obtained from blood [18], saliva [19], cerebrospinal fluid [17], and urine [20] samples. They further have the advantages of stability, specificity, and immediate reactivity. In addition, EVs can cross the blood–brain barrier and exchange substances through the cell membrane [17], eliciting details about the state of the brain and body during delirium. This study aimed to explore the risk factors for postoperative delirium in patients aged ≥ 70 years who underwent spinal surgery. In doing so, it identified biomarkers from urinary EVs using digital immunoassay technology.

This prospective cross-sectional study attempted to identify the risk factors and biomarkers for postoperative delirium using EVs in patients aged ≥ 70 years who underwent spinal surgery. The incidence of delirium among the study participants was 19.8%, which was higher than that reported by another South Korean study [33]. This difference may be attributed to the prospective nature of the present study and the fact that delirium was not diagnosed until discharge. Most participants in our study developed delirium within four postoperative days. However, the timing of onset varied, consistent with the findings of a previous study [8]. It suggests that nurses should pay attention to any signs of delirium in older patients after surgery until discharge. In an American study [25] conducted on the same target population, the incidence of delirium was reported to be as high as 40%. The incidence of delirium in our study was much lower because we strictly excluded patients with a high possibility of delirium, such as patients with cognitive impairment and cancer. Therefore, this study explored the risk factors for postoperative delirium exclusively.

The present study’s risk factors for delirium included a history of benzodiazepine use. These findings were consistent with those of previous studies [34, 35]. In particular, a cognitive disorder in older adults increased in long-term benzodiazepine users than in short-term users [36]. As of the many adverse effects of benzodiazepine use, such as cognitive decline and dependence, the South Korean government began to restrict benzodiazepine prescriptions in 2010; however, long-term prescriptions remain prevalent [37]. Deciding whether to continue taking benzodiazepines or discontinue them before surgery is important in postoperative delirium. Omich et al. [34] reported that the group that discontinued benzodiazepines before surgery had a higher risk of delirium than those that never used or continued using them. However, most drugs, including benzodiazepines, are abruptly stopped before surgery, and only essential drugs such as anti-hypertensive are administered; thus, in most cases, there is a possibility that the side effects of abruptly stopping benzodiazepines manifested postoperatively in patients who had been taking these for a prolonged period. Given the results, nurses need to carefully evaluate the types and patterns of medications taken by patients, particularly psychiatric medications.

Moreover, among orthopedic surgery patients, the incidence of postoperative delirium increased with the severity of depression, and the depression symptoms were associated with the pathology of Alzheimer’s disease as detected in the cerebrospinal fluid [38]. Depression has been associated with decreased overall perception, cognitive decline, episodic memory, performance, and processing speed [39]. Therefore, if delirium occurs after surgery in a patient with depression, the chances of them developing cognitive dysfunction should be considered, and regular follow-up examinations should be performed. Furthermore, preoperative anxiety has been reported to increase agitation after surgery [40]. A previous study in Japan reported that preoperative anxiety treatment effectively reduced postoperative delirium [41]. Contrarily, a study of older patients who had cardiac surgery reported that postoperative delirium was not associated with anxiety and depression symptoms [42]. Additionally, as the number of participants with each disorder was small in this study, the individual disorders could not be analyzed separately. Further validation studies are required to better understand the relationship between postoperative delirium and psychiatric disorders.

Using urine EVs, this study explored the biomarkers associated with delirium through a concentration analysis of tau, UCH-L1, NFL, and GFAP proteins, which indicate nerve damage. EV biomarkers of Alzheimer’s disease have previously been discovered through next-generation deep sequencing or microarrays of serum and cerebrospinal fluid [17, 18]. The Simoa analysis technique used in this study for postoperative delirium was similar to that used by Fong et al. [12] and Ballweg et al. [14]. Most participants with cognitive impairment, including delirium, have comparatively weaker blood vessels or require minimally invasive treatment to prevent infection. As delirium occurs episodically, no difficulties should be encountered in terms of the time of sample collection. Therefore, we determined that the readily accessible urine sample was suitable for biomarker screening. Moreover, urine undergoes little protein degradation in the urinary tract or bladder and can be stored for a longer period of time [43]. As long as a urine catheter is inserted in patients, urine can be collected at any time, which can help detect the changes in the patient’s cells at the time of the onset of delirium. Therefore, EV samples in urine were ideal for identifying the biomarkers of delirium after spinal surgery.

UCH-L1 concentration was high among the participants with the top 30% of delirium severity (p < .001). UCH-L1 plays a vital role in the proteasome system by removing misfolded or oxidized proteins [44]. Contrary to our findings, a study conducted on 427 intensive care unit patients showed that high plasma UCH-L1 concentrations were related to a lower incidence of delirium [45]. In addition, Ballweg et al. [14] reported no relationship between postoperative delirium and UCH-L1 concentration. Differences in the results may be attributed to the dissimilarities between the participants of the present study and those of Hayhurst et al. [45] and Ballweg et al. [14]. Furthermore, the relationship between UCH-L1 and delirium remains controversial [45]. Therefore, more research is necessary to verify the relationship between UCH-L1 and postoperative delirium. Previous studies on patients with brain injuries [46, 47] have shown a positive correlation between UCH-L1 concentration and brain injury after surgery. UCH-L1 has been reported to better predict the clinical prognosis of patients after undergoing brain damage, as compared to S100B, which is an indicator of traumatic brain injury [46]. Unlike delirium caused by other factors, postoperative delirium may cause brain injury due to surgery-induced tissue damage, systemic inflammatory reaction, and oxidative metabolism [48]. Therefore, further repetition studies are needed based on the premise that surgery might cause brain damage.

Tau is a protein indicating neurodegeneration; an increase in tau concentration has been associated with dementia, stroke, and traumatic brain injury [49]. Further, on comparing the top 30% of participants with severe delirium and those without delirium, tau concentration was found to be higher among the participants with delirium (p < .002), which was consistent with previous studies [13, 14]. From preoperative to postoperative day 1, plasma tau increased more significantly in patients with postoperative delirium and correlated with delirium severity [14]. In cardiac surgery, preoperative tau levels were elevated in patients with delirium than that in those without delirium [13]. Based on our findings and the results of previous studies, it can be concluded that patients with severe delirium may have a higher risk of neurodegeneration and brain injury. However, since the urine used in the analysis was collected only after surgery on the same day, differences in tau concentrations before and after surgery were not confirmed. Thus, surgery does not necessarily cause brain injury. Since a causal relationship could not be established due to the experimental design, further studies are needed to confirm this aspect. In addition, since tau is already used as a biomarker in the research on cognitive dysfunction such as dementia, it may potentially serve as a significant biomarker in determining the relationship between postoperative delirium and cognitive dysfunction.

This study provides new insights into the influencing factors of postoperative delirium through a comprehensive assessment and urine exosome analysis of spinal surgery patients aged ≥ 70 years. Prior to surgery, nurses should conduct a thorough drug assessment and help patients on benzodiazepines to consider a prescription change or encourage them to choose non-pharmacological treatment methods such as psychological support and tapering education [50]. In addition, nurses can use the history of benzodiazepine use and psychiatric disorders to classify patients at risk for postoperative delirium. Since the onset time of delirium varied in this study, it is necessary to consider the possibility of delirium in high-risk patients until discharge and to monitor these patients closely. This study confirmed that postoperative delirium might be related to brain damage through digital immunoassay of urine exosomes. Therefore, nurses must explain to patients the possibility of cognitive decline in postoperative delirium so that they can continue to follow up on cognitive function testing. Along with this nursing implication, training geriatric nurses to professionally use comprehensive assessments may be valuable. Comprehensive geriatric assessments, non-pharmacological interventions, and early detection of delirium by the nurse are essential in managing postoperative delirium in older adults. Furthermore, a multidisciplinary approach with anesthesiologists and surgeons should be included to remove or modify the risk factors of postoperative delirium during the perioperative period [6].

This study has some limitations. First, because the number of patients diagnosed with depression, anxiety, and panic disorder was small, the data were merged and analyzed as a psychiatric disorder. Therefore, it was not possible to explore the relevance of delirium according to the exact diagnosis. Second, since protein analysis in urine EVs did not involve measurement of the preoperative concentration, a pre and postoperative comparison could not be made. Therefore, a causal relationship to explain whether the difference was surgically induced, could not be established. Third, the distribution of delirium severity scores differed for each participant, and the number of patients with severe delirium was relatively small. Fourth, an analysis of neuronal EVs isolated from urine was not possible because the amount of extracted EVs was small. Further studies analyzing preoperative samples are necessary to provide the basis for causal reasoning by confirming whether delirium was due to brain damage, which can occur before or after surgery.

In summary, this study was conducted to investigate the risk factors associated with postoperative delirium in older adults aged ≥ 70 years who underwent spinal surgery and to identify the biomarkers associated with delirium by measuring tau, UCH-L1, GFAP, and NFL protein concentrations in urine EVs using digital immunoassay analysis. Our findings can assist healthcare providers in accurately investigating risk factors associated with delirium before surgery and help identify high-risk patients. In addition, tau and UCH-L1 may be used as potential biomarkers for postoperative delirium in elderly patients prior to surgery. However, because this study presents the results from only 91 patients who underwent spinal surgery in South Korea, a follow-up study with larger sample size and other populations is needed to generalize these to the global population.

Supplementary data to this article can be found online at https://doi.org/10.4040/jkan.22146.

Supplementary Figure 1

Exosome characteristics of urinary extracellular vesicles. (A) EVs isolated from urine examined by size and concentration using Nanoparticle Tracking Analysis. The red line represents the error of means of the triplicate measurements; the black line represents the curve fitting. (B) Quantification of EV size and concentration.

Supplementary Figure 2

Box plot of Tau, UCH-L1, GFAP, and NFL of urinary extracellular vesicles. (A) Tau, (B) UCH-L1, (C) GFAP, and (D) NFL assays were measured twice per sample and outliers were removed.

Supplementary Table 1

Concentrations of Tau, UCH-L1, NFL, and GFAP