Abstract
Background and Aims:
Advancing age is associated with poor physiological reserve to combat stressors of surgery and anaesthesia, which is termed frailty, and it leads to postoperative complications. Frailty has been found to have a strong association with postoperative delirium (POD) and postoperative cognitive dysfunction (POCD) in cardiac surgeries. This study aims to determine the association of frailty with POD and POCD in patients undergoing non-cardiac surgeries.
Methods:
A prospective observational, cohort study was done at a tertiary-level hospital on 130 patients aged 55 years and above undergoing surgery under general anaesthesia. Preoperative frailty was classified using a 5-factor modified frailty index (5mFI) and sarcopenia [hand grip strength (HGS)]. POD and POCD were assessed with the revised Delirium Rating Scale (DRS) and Addenbrooke’s Cognitive Examination-III at 24 h, 72 h and 30 days after surgery. Association between 5mFI and HGS with postoperative POD and POCD were calculated using linear regression model.
Results:
Of 117 patients analysed, 58% were identified as frail (5mFI score ≥0.2), exhibiting 2.9 times higher risk of POD [odds ratio (OR) 2.933, 95% confidence interval (CI): 1.001, 8.600, P = 0.050] and 5.8 times higher risk of POCD (OR: 5.380, 95% CI: 1.718, 16.685, P = 0.004) compared to non-frail counterparts. The correlation between 5mFI and postoperative revised DRS-98 scores was statistically significant (P < 0.001), indicating a moderate positive association. However, sarcopenic patients displayed higher but statistically insignificant incidence rates of POD (OR: 1.967, 95% CI: 0.771, 5.014, P = 0.157) and POCD (OR: 1.070, 95% CI: 0.442, 2.589, P = 0.880) than non-sarcopenic patients. Patients with 5mFI scores >0.4 showed a notably increased risk of adverse events within 30 days post-surgery.
Conclusions:
Our study proves the hypothesis that frailty, apart from age, contributes to POD and POCD. Using 5mFI as a predictor in pre-anaesthetic checkups can help identify vulnerable patients early and implement necessary interventions to decrease the burden of cognitive decline.
Keywords: Aged, emergence delirium, frailty, hand strength, postoperative cognitive complications, sarcopenia
INTRODUCTION
Frailty is a geriatric syndrome that depletes the homeostatic reserve of patients, making them more vulnerable to surgical stress.[1] Although the initial definition of frailty has been revised to include an exhaustive list of features and physical tests, the use of the 5-factor modified frailty index (5mFI) is a newer and practically easier criterion to assess frailty.
Sarcopenia is the loss of skeletal muscle mass and strength, which causes functional and physical impairment with advancing age. The gold standard for its assessment is computed tomography (CT) assessment of psoas muscle area or dual-energy X-ray absorptiometry (DEXA) scan, but these are expensive, time-consuming and expose patients to radiation. Poor hand grip strength (HGS) and slow walking speed are easier alternatives to classify sarcopenia in any patient.[2]
Frailty and sarcopenia are associated with poor post-surgical outcomes such as extended length of hospital stay, higher readmission rates and mortality.[2] The assessment of preoperative frailty using Fried criteria predicted postoperative delirium (POD) within 72 h and even up to 7 days in patients undergoing cardiac surgery.[3] However, there is no established data regarding the same in non-cardiac surgery in the Indian population. We hypothesised that frailty and sarcopenia measured by 5mFI and HGS, respectively, would be associated with POD and postoperative cognitive dysfunction (POCD) as well as poor postoperative outcomes in patients undergoing major non-cardiac surgery. The primary objective was to identify the predictive ability of 5mFI to predict POD. Secondary outcomes were the ability of 5mFI to predict POCD, the ability of sarcopenia (HGS) to predict POD and POCD, and the association of 5mFI and HGS with other adverse postoperative outcomes.
METHODS
This prospective observational, cohort study was carried out after obtaining Institute Ethics Committee for Postgraduate Research approval (vide approval number IECPG-23/27.01.2021, OT-09/23.06.2021, dated 23 June 2021) and prospectively registering it in Clinical Trials Registry-India (vide registration number CTRI/2021/08/036082, dated 31 August 2021, accessible at https://ctri.nic.in/). The study was conducted in accordance with the ethical principles for medical research involving human subjects, outlined in the Helsinki Declaration of 1975 (revised 2013) and Good Clinical Practice guidelines.
Patients aged above 55 years who could provide informed consent, had a baseline Addenbrooke’s Cognitive Examination-III (ACE-III) score more than 88 and had been scheduled for elective surgery under general anaesthesia between 2021 and 2023 at a tertiary care hospital were included.[3] Surgeries included were both open and laparoscopic abdominal surgeries of an anticipated duration of 1 h or more. Patients were excluded from the study if they had a history of dementia, active alcohol abuse, chronic opioid or other substance abuse, cerebral surgeries, stroke, psychiatric diseases or use of central nervous system active medications (e.g., antidepressants, antipsychotics, sedatives).
Patients were admitted a day before surgery. After obtaining informed written consent in Hindi or English for their participation and use of data in research and educational purposes, a baseline assessment was done to measure height, weight, body mass index (BMI) and a detailed clinical history focusing on pre-existing conditions. Additional parameters explicitly obtained included the following:
Sarcopenia estimation: By measurement of HGS using a digital hand grip dynamometer [Figure 1] and the highest value recorded based on the American Society of Hand Therapists protocol. Low muscle strength of HGS in the Asian Working Group for Sarcopenia is defined as <26 kg in men and <18 kg in women.[4]
5mFI was noted based on the history by assigning one point to each of the five comorbid variables ascertained from the history of the patient: history of diabetes mellitus, history of congestive heart failure, history of chronic obstructive pulmonary disease or current pneumonia, history of hypertension requiring medication and non-independent functional status defined as requiring assistance for any activity of daily living including bathing, feeding, dressing and mobilisation.[5]
The baseline score of the revised Delirium Rating Scale -98 (DRS)/DRS-R-98 was recorded to rule out pre-existing delirium. Severity ratings range from 0 (no impairment) to 3 (severe impairment), and a severity score >15 or a total score of >18 is indicative of delirium.
Baseline neurocognitive function was scored by ACE-III out of 100, and scores less than 88 were excluded. Scores of 82 or less were taken as POCD.[3]
Figure 1.
Digital hand grip dynamometer for sarcopenia estimation
Intraoperative and postoperative management catered to the patient’s needs, and surgery was performed as per routine institutional protocol. Surgery in all cases was done under balanced general anaesthesia, in which intravenous (IV) fentanyl, propofol and atracurium with tracheal intubation were used for induction, airway management and isoflurane and oxygen/nitrous oxide mixture were used for maintenance. Tracheal extubation was done at the end of surgery after reversal with IV neostigmine and glycopyrrolate. Analgesia was maintained with IV paracetamol, morphine, ketorolac and dexamethasone in laparoscopic surgeries and with thoracic epidural catheter placement in open surgeries. Postoperative analgesia was provided with scheduled IV paracetamol and ketorolac for 3 days in laparoscopic surgeries and with added epidural morphine in open cases. Fluid management was done using the Holliday–Segar formula. Postoperative evaluation was done at intervals of 24 h, 72 h and 30 days after surgery. Those who developed complications were shifted to the high dependence unit and managed further.
The primary outcome parameter, POD, was assessed using the DRS-R-98 score 24 h after surgery. A follow-up evaluation was done at the end of 72 h after surgery with ACE-III to assess for POCD. A set of morbidity outcomes secondary to frailty was evaluated within 30 days, as given below, over a telephonic or in-person evaluation. These included cardiac, pulmonary and renal complications, deep vein thrombosis, surgical site infection, length of hospital and intensive care unit (ICU) stay, readmission rates and 30-day mortality.
Definitions critical to the study include frailty, identified with a 5mFI score of 1/5 or more, and sarcopenia, defined by specific cutoff points in HGS measurements.
Keeping the incidence of POD in frail patients as the primary outcome, 38 patients were required to have a 95% chance of detecting an increase in the primary outcome measure from 2.6% in the control group to 47.1% in the experimental group, with a significance of 5%.[6] Sample size calculation was done online using the website https://clincalc.com/Stats/SampleSize.aspx. Adjusting for five independent covariates involved in the incidence of POD (old age, undiagnosed pre-existing psychiatric or neurological illness, intraoperative hypotension, blood transfusion and unreported substance use) in addition to the factor under study, the sample size was determined to be 114 to reject the null hypothesis of frailty not being associated with POD.
Statistical Package for the Social Sciences statistics software version 26.0 (International Business Machines Corporation, Armonk, NY, USA) was used for the statistical analysis. The Shapiro–Wilk test was used to assess the normality, and internal consistency was assessed by estimating Cronbach’s alpha (0.73). Descriptive statistics were provided in absolute numbers and percentages for categorical variables and mean and standard deviation for continuous variables. We also used bivariate analyses using a linear regression model to determine the relationship between POD (postoperative DRS scores) and POCD (postoperative ACE-III scores) with the predictive factors 5mFI and HGS. P values <0.05 were considered statistically significant.
RESULTS
A total of 130 patients were evaluated for the study. However, since the sociodemographic data for three patients were incomplete, four were not extubated at the end of surgery, and six were discharged 3 days after surgery; these patients were excluded from the analysis. Most patients belonged to American Society of Anesthesiologists classes I and II, and there was a higher female preponderance. The following are the characteristics of patients: median age 64 years [interquartile range (IQR): 58–69], weight 65 kg (IQR: 56–70) and undergoing surgery for a median duration of 2.75 h (IQR: 2.00–4.00). HGS, a measure of sarcopenia, was less than 27 kg in 16.9% of males and less than 16 kg in 23.7% of females. The 5mFI score was calculated out of 5, and it was noted that the maximum percentage (41.5%) scored 0/5 and was classified as ‘not frail’. In addition, 58% of the participants scored 0.2 or more, indicating some degree of frailty, with 25% exhibiting severe frailty of 0.4 and above. Preoperative and postoperative DRS scores were noted, and a noticeable increase was found postoperatively. Similarly, ACE-III scores showed a decrease postoperatively, indicating cognitive changes [Table 1].
Table 1.
Descriptive statistics of baseline parameters of the study population
| Variables | Distribution |
|---|---|
| Age (years) | 64 (57.8–68.9) |
| BMI (kg/m2) | 24.9 (21.8–27.4) |
| Gender (male/female) | 38/62 |
| ASA: I/II/III/IV | 37/55/7/1 |
| 5mFI out of 5 (0/1/2/3/4) | 42/33/15/8/2 |
| Hand grip strength (kg) | 21.4 (15.2–29.9) |
| Duration of surgery (min) | 166 (121–242) |
| Preoperative DRS | 3.2 (0.9–6.4) |
| Postoperative DRS | 7.6 (2.6–14.4) |
| Preoperative ACE-III | 94.2 (89.8–97.9) |
| Postoperative ACE-III | 89.1 (82.9–94.9) |
Data expressed as percentages or median (25th–75th percentile). 5mFI=5-factor modified frailty index, ACE-III=Addenbrooke’s Cognitive Examination-III score, ASA=American Society of Anesthesiologists, BMI=body mass index, DRS=Delirium Rating Scale
A positive correlation suggests that as 5mFI increases, postoperative DRS scores increase. While 10.2% of non-frail patients had POD, frail patients had 10.2% versus 29.4% versus 70% versus 50% incidence of POD for 5mFI 0.2 versus 0.4 versus 0.6 versus 0.8, respectively. The correlation between 5mFI and postoperative DRS scores is statistically significant (P < 0.001), indicating a weak positive association [Table 2]. Patients with 5mFI >0 had 2.9 times higher risk of POD than non-frail patients (5mFI = 0) [odds ratio (OR): 2.933, 95% confidence interval (CI): 1.001, 8.600, P = 0.050], but the difference was not statistically significant. But for higher degrees of frailty of 5mFI score >0.2, the risk of POD was significantly higher (OR: 7.132, 95% CI: 2.609, 19.494, P < 0.001) [Table 2].
Table 2.
Associations between frailty and sarcopenia with POD and POCD
| Frailty | No POD | POD | No POCD | POCD | ||
|---|---|---|---|---|---|---|
| Non-frail (5mFI=0) | 44 | 5 | 45 | 4 | ||
| Frail (5mFI >0.2) | 51 | 17 | 46 | 22 | ||
| Odds ratio (P) | 2.933 (P=0.050) | 5.380 (P=0.004) | ||||
| Severely frail (5mFI >0.4) | 16 | 13 | 15 | 14 | ||
| Odds ratio (P) | 7.132 (P<0.001) | 5.911 (P<0.001) | ||||
| Non-sarcopenic | 59 | 10 | 54 | 15 | ||
| Sarcopenic | 36 | 12 | 37 | 11 | ||
| Odds ratio (P) | 1.967 (P=0.157) | |||||
5mFI=5-factor modified frailty index, POCD=postoperative cognitive dysfunction, POD=postoperative delirium
Figure 2 displays results from a linear regression plot, examining predictors of delirium. 5mFI demonstrated an optimistic estimate, indicating that a higher score was associated with increased odds of delirium (R2 = 0.186, P < 0.001). 5mFI, sarcopenia, duration of surgery, age and BMI were evaluated for their association with delirium occurrence. Similarly, duration of surgery and age showed non-significant associations with delirium.
Figure 2.
Regression plot of 5mFI with revised DRS with fit line (x-axis: 5mFI, y-axis: postoperative DRS score). The fit line shows a positive slope with R2 > 0 (R2 = 0.186, P < 0.001), suggesting a significant mild positive relationship between the 5mFI values and postoperative DRS scores. 5mFI=5-factor modified frailty index, DRS=Delirium Rating Scale
Frail patients by 5mFI were at substantially increased risk of POCD (OR: 5.380, 95% CI: 1.718, 16.6855, P = 0.004) [Table 2]. The negative correlation shows that with higher 5mFI, postoperative ACE-III scores decrease and are statistically significant (Spearman’s Rho correlation coefficient −0.443, P < 0.001) [Figure 3].
Figure 3.
Regression plot of 5mFI with ACE-III score with fit line (x-axis: 5mFI, y-axis: postoperative ACE-III). The fit line shows a negative slope with R2 > 0, suggesting a mild negative relationship between the 5mFI values and postoperative ACE-III scores. 5mFI=5-factor modified frailty index, ACE-III=Addenbrooke’s Cognitive Examination-III
Sarcopenic patients by HGS had a higher incidence of POD (OR: 1.967, 95% CI: 0.771, 5.014, P = 0.157) and POCD (OR: 1.070, 95% CI: 0.442, 2.589, P = 0.880) than non-sarcopenic patients. Still, the difference is not statistically significant to hold sarcopenia as an independent risk factor for POD and POCD [Table 2].
The predictive accuracy, specificity, and sensitivity of 5mFI for predicting POD and POCD are reported. While 5mFI demonstrated moderate accuracy (0.720), it exhibited low specificity (1–0.911) but high sensitivity (0.846), suggesting that it correctly predicts delirious cases but may falsely identify non-delirious patients as positive [Figure 4].
Figure 4.
ROC curve for 5mFI in predicting postoperative delirium. 5mFI=5-factor modified frailty index, DRS=Delirium Rating Scale, ROC=receiver operating characteristic
Patients were followed up in the hospital for readmission or outpatient consult or contacted via telephone 30 days after surgery. It was noted that the most commonly encountered adverse event was a requirement of ICU stay, occurring in almost 10.7% of patients, followed by repeat surgery, postoperative pulmonary complications and wound infection in 2.4% of patients each. Out of 21 (17.9%) patients who developed any adverse event in the 30 days after surgery, 5mFI scores of 0 versus 1 versus 2 versus 2+ were 14.3% versus 10.3% versus 17.6% versus 58.33%, respectively, indicating significantly increasing risk (P = 0.018) with increasing 5mFI score from 2/5 and more.
DISCUSSION
Our prospective cohort study found that frailty, to any degree, as per 5mFI at preoperative evaluation, was associated with a higher risk of POD and POCD. Delayed neurocognitive recovery indicates a new-onset cognitive decline within 30 days after surgery and could replace the term early POCD. Since this diagnosis was based on the presence of deterioration from baseline and not the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition criteria, here we continue to use the term POCD for early-onset cognitive dysfunction and postoperative neurocognitive disorders (NCD) when referring to previous studies describing cognitive decline occurring between 30 days and 12 months after surgery.
The results of our study are comparable to other studies conducted with complex frailty or sarcopenia criteria like CT-guided sarcopenia estimation, complex mathematical formula with multiple anthropometric measures, bioimpedance analysis and DEXA, proposing a replacement of these indices with the easy and reproducible 5mFI at pre-anaesthetic checkup.[7,8,9,10] This was the first study to use such a simplified scale to predict POD and POCD with cost-effective measures, no radiation exposure and no additional investigation in the preoperative period.
A recent meta-analysis showed that Fried’s phenotype predicts frail patients are 2.7 times more likely to experience POD among the older population that underwent non-cardiac surgery.[11] The use of 5mFI showed a comparable 2.9-fold increase in the risk of POD and provides an ordinal scale that additionally indicates a higher risk of POD and POCD with higher 5mFI scores.
The Anaesthesia, Cognition, Evaluation study in older patients who underwent total hip replacement showed that frailty or prefrailty was associated with cognitive decline at 3 and 12 months with 1.5–2 times the risk.[12] The study concluded that preoperative screening for frailty to risk-stratify patients and implement preventive strategies would improve postoperative outcomes for older individuals. This suggests extending our follow-up period to more than 30 days to assess if the positive association holds true beyond early cognitive decline in our study population.
In our study, HGS showed a positive association with POD and POCD, but it was not statistically significant enough to recommend it as a predictor of NCD. This could be due to HGS’s nonspecific nature, as it integrates information about the nervous, musculoskeletal, cardiac and respiratory systems, among others. The study did not stratify patients with different BMIs before performing HGS, and this can be a suggestive step to be included in further studies.
A recent systematic review evaluated frailty and postoperative outcomes like increased mortality, functional decline and major adverse cardiac and cerebrovascular events in cardiac surgery.[13] A similar positive correlation between frailty and increased duration of ICU stay, longer length of hospital stay, wound infection and repeat surgery seen in our study at higher degrees of frailty (5mFI score above 2/5) suggests that frailty, irrespective of the type of surgery, contributes to poor postoperative outcomes.
This study recognises simple tools like 5mFI and HGS and a one-time investment equipment like hand grip dynamometer for frailty assessment. Both are extremely simple, quick, inexpensive and reproducible with minimal training.[14] If these two measures can provide most of the additional prognostic information related to the risk of delirium, they could easily be implemented into standard preoperative assessments. The strength of the study lies in the follow-up of patients till 30 days, ensuring delayed cognitive decline is not missed. The limitations could be the study being a single-centre trial, lack of homogeneity with the nature of surgeries, which can sometimes contribute to NCDs, and lack of stratification of patients based on BMI before HGS is instructed to assess sarcopenia. However, further studies can be mapped to compare the effectiveness of limited versus comprehensive testing in predicting other perioperative outcomes beyond delirium, considering further contributing factors that might have been overlooked.
CONCLUSION
Frailty, as assessed by 5mFI, can predict POD and POCD in elderly patients undergoing non-cardiac surgery. Therefore, routine frailty assessment should be considered in such patients.
Study data availability
De-identified data may be requested with reasonable justification from the authors (email to the corresponding author) and shall be shared upon request.
Conflicts of interest
There are no conflicts of interest.
Funding Statement
Nil.
REFERENCES
- 1.Fried LP, Tangen CM, Walston J. Frailty in older adults: Evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56:M146–56. doi: 10.1093/gerona/56.3.m146. [DOI] [PubMed] [Google Scholar]
- 2.de Salles ICD, Sernik R, da Silva JLP, Taconeli C, Amaral AA, de Brito CMM, et al. Sarcopenia, frailty, and elective surgery outcomes in the elderly: An observational study with 125 patients (the SAFESOE study) Front Med (Lausanne) 2023;10:1185016. doi: 10.3389/fmed.2023.1185016. doi: 10.3389/fmed.2023.1185016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Vu T, Smith JA. An update on postoperative cognitive dysfunction following cardiac surgery. Front Psychiatry. 2022;13:884907. doi: 10.3389/fpsyt.2022.884907. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chen LK, Woo J, Assantachai P, Auyeung TW, Chou MY, Iijima K, et al. Asian Working Group for Sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J Am Med Dir Assoc. 2020;21:300–7. doi: 10.1016/j.jamda.2019.12.012. [DOI] [PubMed] [Google Scholar]
- 5.Subramaniam S, Aalberg JJ, Soriano RP, Divino CM. New 5-factor modified frailty index using American College of Surgeons NSQIP data. J Surg Res. 2017;208:171–8. doi: 10.1016/j.jamcollsurg.2017.11.005. [DOI] [PubMed] [Google Scholar]
- 6.Wei L, Liu M, Zhang S. Association of preoperative frailty with risk of postoperative delirium in older patients undergoing craniotomy: A prospective cohort study. BMC Surg. 2024;24:272. doi: 10.1186/s12893-024-02573-2. doi: 10.1186/s12893-024-02573-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Knoedler S, Schliermann R, Knoedler L, Wu M, Hansen FJ, Matar DY, et al. Impact of sarcopenia on outcomes in surgical patients: A systematic review and meta-analysis. Int J Surg. 2023;109:4238–62. doi: 10.1097/JS9.0000000000000688. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kalyon S, Gumuskaya PO, Ozsoy N, Ozcan M, Yoldeir SA, Toprak ID, et al. The relationship between sarcopenia and cognitive dysfunction in bladder tumour patients. J Surg Med. 2020;6:537–42. [Google Scholar]
- 9.Shema AB, Forget MF, Bouchard M, Bellemare L, Afilalo J, Mullie LA, et al. Comparison between sarcopenia and frailty as predictors of postoperative complications: 65 years and older non-cardiac surgery cohort study. J Gerontol A Biol Sci Med Sci. 2020;75:1734–42. [Google Scholar]
- 10.Marques A, Queirós C, Santy-Tomlinson J. Frailty and sarcopenia. In: Hertz K, Santy-Tomlinson J, editors. Fragility Fracture and Orthogeriatric Nursing. Perspectives in Nursing Management and Care for Older Adults. Cham: Springer; 2024. [Google Scholar]
- 11.Gracie TJ, Caufield-Noll C, Wang NY, Sieber FE. The association of preoperative frailty and postoperative delirium: A meta-analysis. Anesth Analg. 2021;133:314–23. doi: 10.1213/ANE.0000000000005609. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Evered LA, Vitug S, Scott DA, Silbert B. Preoperative frailty predicts postoperative neurocognitive disorders after total hip joint replacement surgery. Anesth Analg. 2020;131:1582–8. doi: 10.1213/ANE.0000000000004893. [DOI] [PubMed] [Google Scholar]
- 13.Peeler A, Moser C, Gleason KT, Davidson PM. Frailty as a predictor of postoperative outcomes in invasive cardiac surgery: A systematic literature review. J Cardiovasc Nurs. 2022;37:231–47. doi: 10.1097/JCN.0000000000000821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Müller J, Nowak S, Vogelgesang A, von Sarnowski B, Rathmann E, Schmidt S, et al. Evaluating Mechanisms of Postoperative Delirium and Cognitive Dysfunction Following Elective Spine Surgery in Elderly Patients (CONFESS): Protocol for a Prospective Observational Trial. JMIR Res Protoc. 2020;9(2):15488. doi: 10.2196/15488. doi: 10.2196/15488. [DOI] [PMC free article] [PubMed] [Google Scholar]