Risk factors for postoperative delirium in elderly patients undergoing gastroenterological surgery: A single‐center retrospective study

Abstract

Aim

To investigate the risk factors for postoperative delirium among elderly patients undergoing elective surgery for gastroenterological cancer.

Methods

From May 2020 to March 2022, patients ≥75 years old with gastroenterological cancer who underwent radical surgery were enrolled. The geriatric assessment, including evaluations of frailty, physical function, nutrition status, and cognitive function, was conducted preoperatively. The confusion assessment method was used to diagnose postoperative delirium. A multivariate logistic regression analysis was used to determine risk factors for postoperative delirium.

Results

A total of 158 patients were eligible for inclusion in this study. Of these 53 patients (34%) developed postoperative delirium. In the univariate analysis, the age, regular use of sleeping drugs and benzodiazepine, Charlson Comorbidity Index score, performance status, Fried's frailty score, Vulnerable Elders Survey‐13 score, grip weakness, Short Physical Performance Battery (SPPB) score, Mini Nutritional Assessment Short‐Form score, and Mini‐Mental State Examination score were statistically associated with the incidence of postoperative delirium. In the multivariate analysis, a SPPB score ≤9, Mini Nutritional Assessment score ≤11, a Mini‐Mental State Examination score ≤24, and regular use of benzodiazepine were found to be independent preoperative risk factors for postoperative delirium.

Conclusion

Certain findings during the preoperative geriatric assessment, especially low SPPB, Mini Nutritional Assessment Short‐Form and Mini‐Mental State Examination scores, and regular use of benzodiazepine were risk factors for postoperative delirium in elderly patients undergoing gastroenterological surgery.

Short abstract

This single‐center retrospective study analyzes 158 patients over 75 years old who underwent radical gastroenterological cancer and identifies preoperative risk factors for postoperative delirium. This study revealed that SPPB score ≤9, Mini Nutritional Assessment Short‐Form score ≤11, Mini‐Mental State Examination score ≤24, and regular use of benzodiazepine were independent preoperative risk factors for postoperative delirium, respectively. This study is the first to show that the preoperative SPPB score is associated with postoperative delirium in the field of gastroenterological surgery.

1. INTRODUCTION

The aging population in Japan is rapidly increasing, currently sitting at 37 million and projected to reach 30% of the population by 2025, ¹ and similar phenomena are being seen in Western countries. Furthermore, an estimated 19.3 million new cancer cases were reported worldwide in 2020, and the number of elderly patients with gastroenterological cancer who undergo surgery is increasing. ²

Postoperative delirium is a severe complication after surgery, with a reported incidence rate of 8%–46%. ³ , ⁴ It is characterized by changes in cognition, attention, and levels of consciousness within 30 days after surgery. ⁵ Postoperative delirium not only adversely affects patients, it also increases the burden on clinicians, as it increases the length of hospital stay, medical costs, nursing burden, risk of complications, and mortality. ⁶ As a result of the increasing population of elderly patients, the incidence of postoperative delirium is expected to increase as well, presenting a major problem in the field of elderly surgery.

A way to detect elderly patients at a high risk of postoperative delirium is needed in order to prevent this issue altogether or apply appropriate treatment during the perioperative period. Reported typical preoperative risk factors for delirium are an older age, cognitive impairment, and comorbidities (cardiovascular disease, heart failure, renal failure, chronic obstructive pulmonary disease, and diabetes mellitus). ⁵ , ⁶ In this era of an aging society, geriatric assessments are used in various settings to evaluate medical, social, and environmental factors that influence patients' overall well‐being and to determine the functional status, fall risk, medication status, nutrition, vision, cognition, mood, and toileting status. ⁷

Geriatric assessments have been shown to be correlated with the postoperative outcomes in various types of surgery. ⁸ However, there are few quantifiable indicators concerning the degree to which deterioration of the physical, nutritional, and mental function with aging contributes to the development of postoperative delirium.

The present study clarified the preoperative risk factors for postoperative delirium in elderly patients who underwent radical surgery for gastroenterological cancer, based on geriatric assessments.

2. METHODS

2.1. Patients

This retrospective study was approved by our institutional review board (approval no. R04‐139). All patients ≥75 years old who underwent elective radical surgery for gastroenterological cancer at Tsukuba University Hospital (Ibaraki, Japan) between May 2020 and March 2022 were included in this study. Exclusion criteria included: (1) palliative surgery and exploratory laparotomy; (2) severe hearing impairment; (3) inability to communicate because of severe dementia or psychiatric illness; and (4) the presence of delirium at baseline.

2.2. Preoperative factors

A preoperative evaluation was conducted to determine patients' data, including their age, sex, tumor location, performance status (PS), and American Society of Anesthesiologists‐physical status (ASA‐PS), body mass index (BMI), drinking habit, smoking habit, and regular use of sleeping drugs, including benzodiazepine and others. We assessed patients' comorbidities using the Charlson Comorbidity Index (CCI) including heart disease, peripheral vascular disease, brain disease, respiratory disease, collagen disease, diabetes mellitus, and renal disease. Each patient's CCI score was calculated by summing the scores weighted by comorbidity and age. ⁹

2.3. Preoperative geriatric assessment

All patients underwent geriatric assessments before surgery. The patients were evaluated by Fried's frailty score for frailty status, ¹⁰ the Vulnerable Elders Survey‐13 (VES‐13) for social frailty, ¹¹ grip strength and the Short Physical Performance Battery (SPPB) for the physical function, ¹² the Mini Nutritional Assessment Short‐Form (MNA‐SF) and Controlling Nutritional Status (CONUT) score for the nutritional status, ⁵ , ¹³ , ¹⁴ and the Mini‐Mental State Examination (MMSE) for the cognitive function. ¹⁵

Fried's frailty is a tool for assessing frailty, consisting of five criteria: unintentional weight loss, poor endurance and energy, low physical activity level, slow gait speed, and weakness. ¹⁰ The VES‐13 is composed of eight questionnaires, which are intended to assess the age, health status, and activities of daily living. The SPPB consists of three timed components (balance, the 5X Sit‐to‐Stand Test and usual‐pace gait speed) that measure the lower limb function. The MNA‐SF comprises six questionnaires and is used to classify elderly individuals as well‐nourished, at risk of malnutrition, or malnourished. The CONUT score is a screening tool used to assess the nutritional status of patients, consisting of three clinical variables: total peripheral lymphocyte counts, total cholesterol levels, and serum albumin levels. The MMSE is most often used to detect cognitive disability among elderly individuals.

These preoperative geriatric assessments were recorded before surgery.

2.4. Perioperative and postoperative factors

We evaluated the type of surgery (esophagectomy, gastrectomy, colectomy, rectectomy, hepatectomy, extrahepatic bile duct resection, and pancreatectomy) and operating procedure (robot‐assisted, laparoscopic, and open surgery). Regarding esophagectomy, procedures in which a laparoscopic approach was used for both the mediastinum and the abdomen were classified as laparoscopic surgery, while those in which an open surgical approach was used for either the mediastinum or the abdomen were classified as open surgery. Regarding the operating procedure, we classified both robot‐assisted and laparoscopic surgeries as laparoscopic surgery. We also evaluated operating times, blood loss, transfusion, and whether or not patients had postoperative complications using the Clavien–Dindo classification system (grade ≥ II counted).

2.5. The diagnosis and treatment of postoperative delirium

We diagnosed postoperative delirium according to the definition of delirium as indicated by the confusion assessment method (CAM) criteria by a gastroenterologist or nurse. ⁶ The CAM algorithm was used to determine the delirium diagnosis according to the presence or absence of four features: an acute change in mental status with a fluctuating course, inattention, and either disorganized thinking or an altered level of consciousness. The patients' condition was observed, and findings corresponding to the CAM algorithm were written on their medical record three times a day.

The treatment of postoperative delirium is described in Table 2. Medication and physical restraint were used when improving the patient's orientation in cases in which reducing sensory deprivation failed to improve the symptoms of postoperative delirium.

TABLE 2.

Details of patients with postoperative delirium.

Patients and preoperative factors	n = 53
Onset day of postoperative delirium (POD, median)	2 (0–10)
Duration of postoperative delirium (days, median)	2 (1–8)
Medication	31 (57.4%)
Type of medications (with duplicates)
Typical antipsychotics	25
Atypical antipsychotics	7
Antihistamine	2
Non‐benzodiazepine sleeping drugs	1
Physical restraint	21 (38.9%)

Abbreviation: POD, postoperative day.

2.6. Statistical analyses

Statistical analyses were performed using R version 4.2.0 (R Project for Statistical Computing). Differences in the age, sex, drinking habit, regular use of sleeping drugs, benzodiazepine and others, CCI score, each comorbidity, PS, ASA‐PS, Fried's frailty, VES‐13 score, grip strength, SPPB, MNA‐SF, CONUT, MMSE scores, type of surgery, operating procedure, transfusion, and postoperative complications were analyzed using the chi‐squared test. Differences in tumor locations and smoking habits were analyzed using Fisher's exact test. Differences in operating time and BMI were analyzed using a two‐sample t‐test. Differences in blood loss were analyzed using Wilcoxon's two‐sample test. All p‐values ≤0.05 were considered statistically significant. The cut‐off values of the age, CCI score, PS, ASA‐PS, Fried's frailty, VES‐13, grip strength, SPPB, MNA‐SF, CONUT, and MMSE scores were determined using a receiver operating characteristic curve (Figure S1), and the patients were divided into a delirium group and a non‐delirium group. Variables analyzed in the multivariate analysis were selected from among the age, regular use of benzodiazepine, comorbidity (CCI score), frailty (VES‐13 score), physical function (SPPB score), nutritional status (MNA‐SF score), and cognitive function (MMSE score). Multivariate analyses were conducted using a multiple logistic regression model to adjust for multiple risk factors.

3. RESULTS

A total of 158 patients were analyzed during the study period. The median age of these patients was 79 (range, 75–91) years old. There were 111 males and 47 females. The median BMI was 22.5 (14.2–32.4) kg/m². Sixty‐two of the 158 patients had a drinking habit. Sixteen of the 158 patients were current smokers, and 61 were past smokers. Seventeen of the 158 patients were regular users of sleeping drugs (benzodiazepines, n = 9; and others, n = 8). The procedures of the 158 patients were as follows: esophagectomy (n = 16), gastrectomy (n = 28), colectomy (n = 38), rectectomy (n = 18), hepatectomy (n = 10), extrahepatic bile duct resection (n = 3), and pancreatectomy (n = 45). The comorbidities of the 158 patients were as follows: heart disease (n = 22), peripheral vascular disease (n = 18), brain disease (n = 13), respiratory disease (n = 12), collagen disease (n = 3), liver disease (n = 14), diabetes mellitus (n = 41), and renal disease (n = 4). The median CCI was 7 (6–15). The average PS and ASA‐PS were 0.27 (0–2) and 2.51 (1–4), respectively. The median grip strength of the male patients was 30 (range, 15–46) kg, and that of the female patients was 19 (range, 6.7–44) kg. The median SPPB score of these patients was 11 (range, 3–12), and the average MMSE was 26.4 (7–30). Table 1 shows the patient characteristics.

TABLE 1.

Preoperative factors.

Patients and preoperative factors	n = 158
Age (years)	79 (75–91)
Sex (male/female)	111/47
BMI (kg/m2)	22.5 (14.2–32.4)
Drinking habit (+/−)	62/96
Smoking habit (current/past/absent)	16/61/81
Regular use of sleeping drug (+/−)	17/141
Benzodiazepine (+/−)	9/149
Others (+/−)	8/150
Type of surgery
Esophagectomy	16
Gastrectomy	28
Colectomy	38
Rectectomy	18
Hepatectomy	10
Extrahepatic bile duct resection	3
Pancreatectomy	45
CCI	7 (6–15)
Heart disease (+/−)	22/136
Peripheral vascular disease (+/−)	18/140
Brain disease (+/−)	13/145
Respiratory disease (+/−)	12/146
Collagen disease (+/−)	3/155
Liver disease (+/−)	14/144
Diabetes mellitus (+/−)	41/117
Renal disease (+/−)	4/154
PS	0.27 (0–2)
ASA‐PS	2.51 (1–4)
Fried's frailty	1.89 (0–5)
VES‐13	1.80 (1–10)
Grip strength (kg)
Male	30 (15–46)
Female	19 (6.7–44)
SPPB	11 (3–12)
Balance	3.6 (0–4)
Gait speed (4 m/s)	4.8 (2.25–23.9)
Chair standing speed (5 times/s)	12.3 (5–49.4)
MNA‐SF	10.11 (3–14)
CONUT	2.52 (0–11)
MMSE	26.4 (7–30)

Abbreviations: ASA‐PS, American Society of Anesthesiologists‐physical status; BMI, body mass index; CCI, Charlson Comorbidity Index; CONUT, Controlling Nutritional Status; MMSE, Mini‐Mental State Examination; MNA‐SF, Mini Nutritional Assessment Short‐Form; PS, performance status; SPPB, Short Physical Performance Battery; VES‐13, Vulnerable Elders Survey‐13.

The overall incidence of postoperative delirium was 33.5% (53 out of 158 patients) among all patients ≥75 years old who underwent radical surgery for gastroenterological cancers. Table 2 shows the details of postoperative delirium patients. The median timing of the onset of postoperative delirium was postoperative day 2 (0–10). The median duration of postoperative delirium was 2 (1–8) days. Medication was administered to 31 of 53 patients with postoperative delirium (58.5%). The most commonly prescribed medications were typical and atypical antipsychotics. Antihistamine and non‐benzodiazepine sleeping drugs were also administered to a few patients. Physical restraints were used for 21 of 53 patients (39.6%).

In Table 3, the univariate analysis shows the differences in the preoperative factors between the two groups. The age, regular use of sleeping drugs and benzodiazepines, CCI score, and PS were significantly different between the delirium and non‐delirium groups (p < 0.05).

TABLE 3.

A comparison of preoperative factors between patients with and without postoperative delirium.

	Postoperative delirium (+) (n = 53)	Postoperative delirium (−) (n = 105)	Total (n = 158)	p‐value
Age (≥79/≤78 years old)	34/19	47/58	81/77	0.02
Sex (male/female)	35/18	76/29	111/47	0.41
BMI (kg/m2)	22.2	22.7	22.5	0.19
Drinking habit (+/−)	16/37	46/59	62/96	0.10
Smoking habit (current/past/absent)	6/19/28	10/42/53	16/61/81	0.85
Regular use of sleeping drug (+/−)	13/40	4/101	17/141	<0.01
Benzodiazepine (+/−)	8/45	1/104	9/149	<0.01
Others (+/−)	5/48	3/102	8/150	0.08
CCI (≥8/≤7)	27/26	31/74	58/100	0.01
Heart disease	8/45	14/91	22/136	0.76
Peripheral vascular disease	5/48	13/92	18/140	0.58
Brain disease	6/47	7/98	13/145	0.31
Respiratory disease	4/49	8/97	12/146	0.99
Collagen disease	2/51	1/104	3/155	0.22
Liver disease	6/47	8/97	14/141	0.44
Diabetes mellitus	13/40	28/77	41/117	0.77
Renal disease	2/51	2/103	4/154	0.48
PS (≥1/0)	22/31	10/95	32/126	<0.01
ASA‐PS (≥3/2)	28/25	52/53	80/78	0.70

Abbreviations: ASA‐PS, American Society of Anesthesiologists‐physical status; BMI, body mass index; CCI, Charlson Comorbidity Index; PS, performance status.

In Table 4, the univariate analysis shows the differences in the preoperative geriatric factors between the two groups. The Fried's frailty score, VES‐13 score, grip strength, SPPB score, MNA‐SF score, and MMSE score were significantly different between the delirium and non‐delirium groups (p < 0.05).

TABLE 4.

A comparison of preoperative geriatric factors between patients with and without postoperative delirium.

	Postoperative delirium (+) (n = 53)	Postoperative delirium (−) (n = 105)	Total (n = 158)	p‐value
Fried's frailty (≥3/≤2)	27/26	21/84	48/110	<0.01
VES‐13 (≥2/≤1)	31/22	34/71	65/93	<0.01
Grip strength	38/15	45/60	83/75	<0.01
Male (<32/≥32 kg)
Female (<15.5/≥15.5 kg)
SPPB (≤9/≥10)	24/29	16/89	40/118	<0.01
MNA‐SF (≤11/≥12)	41/12	62/43	103/55	0.02
CONUT (≥3/≤2)	26/27	35/70	61/97	0.06
MMSE (≤24/≥25)	24/29	16/89	40/118	<0.01

Abbreviations: CONUT, Controlling Nutritional Status; MMSE, Mini‐Mental State Examination; MNA‐SF, Mini Nutritional Assessment Short‐Form; SPPB, Short Physical Performance Battery; VES‐13, Vulnerable Elders Survey‐13.

Table 5 shows the perioperative and postoperative factors of the delirium and non‐delirium groups. There were no significant differences between the two groups.

TABLE 5.

A comparison of perioperative and postoperative factors between patients with and without postoperative delirium.

	Postoperative delirium (+) (n = 53)	Postoperative delirium (−) (n = 105)	Total (n = 158)	p‐value
Type of surgery				0.14
Esophagectomy	8	8	16
Gastrectomy	12	16	28
Colectomy	13	25	38
Rectectomy	2	16	18
Hepatectomy	4	6	10
Extrahepatic bile duct resection	2	1	3
Pancreatectomy	12	33	45
Operating procedure				0.96
Laparoscopic/open	26/27	52/53	78/80
Operating time (min)	310 (87–594)	330 (71–604)	323 (71–604)	0.31
Blood loss (mL)	263 (0–1760)	324 (0–4380)	303 (0–4380)	0.47
Transfusion (+/−)	13/40	13/92	26/132	0.051
Postoperative complication (+/−)	26/27	45/60	71/87	0.46
Clavien–Dindo grade ≥2

A multivariate logistic regression analysis showed that, after adjusting for risk factors, susceptibility to postoperative delirium was indicated by an SPPB score ≤9 (odds ratio [OR] = 2.48; 95% confidence interval [CI], 1.03–6.04), an MNA‐SF score ≤11 (OR = 2.66; 95% CI, 1.11–6.90), an MMSE score ≤24 (OR = 3.76; 95% CI, 1.61–9.07), and regular use of benzodiazepines (OR = 14.3; 95% CI, 2.10–295.5) (p < 0.05) (Table 6).

TABLE 6.

Results of a multivariate analysis of preoperative factors.

	Multivariate OR (95% CI)	p‐value
Age (≥79 years old)	1.52 (0.67–3.47)	0.31
CCI (≥8)	1.40 (0.60–3.21)	0.43
VES‐13 (≥2)	1.84 (0.82–4.14)	0.14
SPPB (≤9)	2.48 (1.03–6.04)	0.04
MNA‐SF (≤11)	2.66 (1.11–6.90)	0.03
MMSE (≤24)	3.76 (1.61–9.07)	<0.01
Regular use of benzodiazepine (+)	14.3 (2.10–295.5)	0.02

Abbreviations: CCI, Charlson Comorbidity Index; MMSE, Mini‐Mental State Examination; MNA‐SF, Mini Nutritional Assessment Short‐Form; SPPB, Short Physical Performance Battery; VES‐13, Vulnerable Elders Survey‐13.

The results of the comparison of SPPB (≤9/≥10), MNA‐SF (≤11/≥12), MMSE (≤24/≥25), and regular use of benzodiazepine use (+/−) for each type of surgery between patients with and without postoperative delirium are shown in Tables [Link], [Link].

4. DISCUSSION

The present study showed that one of the independent preoperative predictive factors for postoperative delirium in elderly patients who underwent gastroenterological surgery was a decreasing physical function, especially that of the lower limbs, as indicated by an SPPB score ≤9. Another important risk factor for postoperative delirium in elderly patients was undernutrition, as indicated by an MNA‐SF score ≤11. In addition, cognitive impairment, as indicated an MMSE score ≤24, and regular use of benzodiazepines were also revealed to be independent preoperative risk factors for postoperative delirium.

This study is the first to show that a low preoperative SPPB score, representing a reduced extremity physical function, is an independent preoperative risk factor for postoperative delirium in elderly patients who undergo gastroenterological surgery. SPPB has emerged as one of the most promising tools for evaluating the functional capability and frailty and providing a quantitative measure of the biological age of older patients. ¹⁶ A lower SPPB score was previously reported to be associated with the incidence of falls, ¹⁷ and an SPPB score ≤9 was predictive of all‐cause mortality in a meta‐analysis. ¹⁸ Compared to SPPB scores of 10–12, values of 0–3 (OR 3.25, 95% CI 2.86–3.79, p < 0.001), 4–6 (OR 2.14, 95% CI 1.92–2.39, p < 0.001), and 7–9 (OR 1.50, 95% CI 1.32–1.71, p < 0.001) were each associated with an increased risk of all‐cause mortality. ¹⁸ In areas of cardiac and transplant surgery, an association between the preoperative SPPB score and postoperative delirium has been reported. ¹⁹ , ²⁰ Compared to an SPPB score of 10–12, values of 4–6 (OR 8.26, 95% CI 2.23–30.64, p = 0.040) were associated with an increased risk of postoperative delirium in cardiac surgery. ¹⁹ Dalton et al. also reported that increased SPPB scores were associated with a decreased incidence of postoperative delirium (OR 0.55, 95% CI 0.31–0.96, p = 0.036) in transplant surgery. ²⁰ Therefore, the preoperative measurement of SPPB is important for predicting postoperative delirium not only in cardiac and transplant surgery but also in gastroenterological surgery.

In the field of gastroenterological surgery, there are no reports that describe the association between the preoperative SPPB score and the incidence of postoperative delirium. Our results showed that preoperative measurement of the SPPB score was able to predict postoperative delirium in elderly patients.

This study also revealed that a low MNA‐SF score, reflecting undernutrition, was an independent preoperative risk factor for postoperative delirium in elderly patients who underwent gastroenterological surgery. Although the MNA is an established nutritional screening tool for identifying subjects with malnutrition or at risk of malnutrition, it consists of 18 questionnaires, so its completion is time‐consuming. ²¹ The MNA‐SF is a simplification of the MNA, consisting of just six questionnaires, and is used as a substitute for the MNA. ¹⁴ In general, the MNA‐SF score has been reported to strongly correlate with the MNA score. ¹⁴ Therefore, the MNA‐SF is also considered a valid tool for assessing the nutritional status in elderly patients. Undernutrition is a reported preoperative risk factor for postoperative delirium. Chu et al. reported that, compared to MNA‐SF scores of 12–14, values of 0–11 (OR 2.85, 95% CI 1.19–6.87, p < 0.05) were associated with an increased risk of postoperative delirium in orthopaedic surgery. ²¹ Zhao et al. similarly reported that, compared to MNA‐SF scores of 12–14, values of 0–7 (OR 4.06, 95% CI 1.62–10.18, p = 0.003) were associated with an increased risk of postoperative delirium in non‐cardiac surgery. ²² Our study also showed an MNA‐SF score ≤11 in elderly patients as an independent preoperative predictive factor for postoperative delirium, which is in line with previous reports (Table 5). Not only the MNA‐SF score but also other nutritional indices, such as the serum albumin level, are reported to be preoperative risk factors for postoperative delirium. Matsuki et al. reported that, compared to a serum albumin level of ≥3.6 g/dL, a value <3.5 g/dL (OR 3.12, 95% CI 1.25–7.83, p = 0.015) was associated with an increased risk of postoperative delirium in urological surgery. ²³ The serum albumin level has already been established as a useful tool for assessing the nutritional condition of patients, but changes in the level occur later than decreases in the MNA score. ²⁴ Therefore, the MNA and MNA‐SF are useful for managing surgery for elderly patients, as they allow clinicians to identify patients who are malnourished and at risk of malnutrition more quickly than the serum albumin level.

The present study revealed that a cognitive decline, as measured by an MMSE score ≤24, is an independent preoperative risk factor for postoperative delirium in elderly patients who undergo gastroenterological surgery. Cognitive impairment has been reported to be a major risk factor for postoperative delirium. ⁶ To assess the cognitive function of patients, the MMSE and Hasegawa Dementia Scale‐Revised (HDS‐R) are widely used. An MMSE score <23 is reported as an independent preoperative predictive factor for postoperative delirium in older patients who undergo colorectal surgery (OR 3.50, 95% CI 1.36–9.02, p = 0.009), ²⁵ and an HDS‐R score <20 is reported to be an independent preoperative predictive factor for postoperative delirium in elderly patients who undergo urological surgery (OR 3.50, 95% CI = 1.36–9.02, p = 0.009). ²³ Our study also showed an MMSE ≤24 to be an independent preoperative predictive factor for postoperative delirium, which is consistent with previous reports (Table 5). Cognitive impairment is also reported to be a predictor of mortality. ⁸ Therefore, conducting the MMSE preoperatively to clarify the cognitive impairment status is important when planning gastroenterological surgery for elderly patients.

In this study, the regular use of benzodiazepines was identified as an independent preoperative risk factor for postoperative delirium in patients undergoing gastroenterological surgery. The short‐term use of benzodiazepines is considered safe and effective for the management of insomnia and anxiety ²⁶ ; however, perioperative use is reported to be a risk factor for postoperative delirium. ⁵ , ⁶ , ²⁶ Omichi et al. reported that postoperative delirium occurred in 13.8% (four out of 29) of patients who continued using benzodiazepines during hospitalization and in 49.0% (24 out of 49) of those who discontinued benzodiazepines. ²⁶ Therefore, if elderly patients regularly take benzodiazepines, it is advisable to gradually reduce the dose of benzodiazepines or to change to non‐benzodiazepine drugs before hospitalization.

The overall incidence of delirium in the present study was 33.5% (53 of 158 patients), a higher frequency than previously reported. Chu et al. reported that the incidence of postoperative delirium was 9.6% in patients ≥60 years old who underwent orthopaedic surgery. ²¹ Arita et al. reported that the incidence of postoperative delirium was 13.3% in patients ≥70 years old who underwent colorectal cancer surgery. ²⁵ This difference in the incidence of postoperative delirium between the present study and previous reports may be due to our setting the target age higher than in other studies, as we targeted patients ≥75 years old in our study. An older age is reported to be an extremely important risk factor for the development of postoperative delirium in predisposed patients. ⁶ Therefore, the older target age of patients in the present study may have contributed to the higher incidence of postoperative delirium than in other studies.

In the multivariate analysis, frailty was not extracted as a predictive factor for postoperative delirium, although other reductions in the function of geriatric patients, such as physical, nutritional, and cognitive decline (SPPB ≤9, MNA‐SF ≤11, and MMSE ≤24) were specified as predictive factors for postoperative delirium in elderly patients who received gastroenterological surgery. However, Susano et al. reported that, compared to Fried's frailty score of 0, scores of 3–5 (OR 6.6, 95% CI 1.96–21.9, p < 0.002) were associated with an increased risk of postoperative delirium in spine surgery, ²⁷ and Yang et al. also reported that the frequency of social activities was associated with the incidence of delirium. ²⁸ Frailty is defined as a clinically recognizable state of increased vulnerability resulting from aging‐associated decline and an increased risk of poor health outcomes. ¹⁰ Therefore, further investigations are needed to identify additional risk factors related to geriatric assessment for postoperative delirium other than physical, nutritional, and cognitive weakness in elderly patients undergoing gastroenterological surgery.

This study identified low SPPB, MNA‐SF, and MMSE scores and the regular use of benzodiazepines as risk factors for postoperative delirium. Prehabilitation may be effective for patients with a poor physical function, nutritional status, or cognitive function. Prehabilitation is a preoperative multidisciplinary intervention that includes exercise and nutritional support. ²⁹ Mima et al. reported that preoperative rehabilitation and nutritional intervention improve surgical outcomes and reduce the incidence of postoperative complications in gastroenterological surgery. ²⁹ Therefore, the prehabilitation of vulnerable patients may reduce the incidence of postoperative delirium. Although geriatric assessments are not routinely performed in the daily clinical setting, it is important to identify frail patients using geriatric assessments such as SPPB, MNA‐SF, and MMSE, and to intervene intensively before surgery.

Several limitations associated with the present study warrant mention. First, this study was retrospective and performed at a single institute. Second, psychiatrists were not involved in the diagnosis of postoperative delirium in the present study. However, the diagnosis rate of postoperative delirium was similar between different evaluators, for example, 8.0% by psychiatrists and 9.6% by generalists after orthopaedic surgery. ⁴ , ²¹ Third, the diagnosis of postoperative delirium was made using the CAM, not the 5th edition of the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM‐5). ⁶ A more accurate diagnosis of delirium might have been obtained had we used the DSM‐5. However, Shi et al. compared the utility of the CAM and DSM‐4 for diagnosing postoperative delirium and reported that CAM presents higher specificity than sensitivity. ³ Therefore, an overdiagnosis was probably avoided in the present study. Furthermore, Rolfson et al. reported a study in which 23 of 71 patients were diagnosed with postoperative delirium using CAM, and in which DSM‐4 was retrospectively applied to all cases; the diagnosis remained unchanged in all cases. ³⁰

With the development of surgical and anesthesia techniques, the indication of radical surgery for gastroenterological cancer has been increasingly considered for elderly patients. In the near future, surgeons are likely to have an increased chance of encountering postoperative delirium patients, as elderly patients are more likely than nonelderly to develop postoperative delirium. Therefore, identifying the risk factors for postoperative delirium preoperatively would be useful for prevention and early intervention. Our study revealed that measuring the physical, nutrition, and cognitive status and confirming the regular use of benzodiazepines can lead to the identification of patients at a high risk of postoperative delirium. These findings are thought to be useful in the management of postoperative delirium after gastroenterological surgery in the era of an aging society.

5. CONCLUSION

A preoperative decline in the physical function and lower extremity muscle strength, represented by an SPPB score ≤9, undernutrition, as represented by an MNA‐SF score ≤11, a decline in the cognitive function, represented by an MMSE score ≤24, and regular use of benzodiazepines were independent risk factors for postoperative delirium in elderly patients who underwent gastroenterological surgery.

AUTHOR CONTRIBUTIONS

Tomoyuki Sugi and Tsuyoshi Enomoto contributed equally to this work; Tomoyuki Sugi and Tsuyoshi Enomoto obtained the clinical information, performed the literature research, and wrote the manuscript and tables under the supervision of Yusuke Ohara, Kinji Furuya, Daichi Kitaguchi, Shoko Moue, Yoshimasa Akashi, Koichi Ogawa, Yohei Owada, and Tatsuya Oda.

FUNDING INFORMATION

The authors received no funding for this article.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest for this article.

ETHICS STATEMENTS

Approval of the research protocol: The protocol for this research project has been approved by a suitably constituted Ethics Committee of the institution and it conforms to the provisions of the Declaration of Helsinki (Committee of University of Tsukuba hospital, approval no. R04‐139).

Informed Consent: N/A.

Registry and the Registration No. of the study/trial: N/A.

Animal Studies: N/A.

Supporting information

Figure S1:

Table S1:

Table S2:

Table S3:

Table S4:

Sugi T, Enomoto T, Ohara Y, Furuya K, Kitaguchi D, Moue S, et al. Risk factors for postoperative delirium in elderly patients undergoing gastroenterological surgery: A single‐center retrospective study. Ann Gastroenterol Surg. 2023;7:832–840. 10.1002/ags3.12676