Comparison of the clinical frailty score (CFS) to the National Emergency Laparotomy Audit (NELA) risk calculator in all patients undergoing emergency laparotomy

Abstract Aim There is evolving evidence that preoperative frailty predicts outcomes of older adults undergoing emergency laparotomy (EmLap). We assessed frailty scoring in an emergency surgical population that included patients of all ages and then compared this to an established perioperative prognostic score. Method Data from the prospective Emergency Laparoscopic and Laparotomy Scottish Audit (ELLSA; November 2017–October 2018) was used. All adults over 18 were included. Frailty was measured using 7‐point clinical frailty score (CFS). Outcome measures: 30‐day mortality, hospital length of stay (LOS), 30‐day readmission. Areas under the receiver‐operating characteristic (ROC) curves were calculated for CFS (1–7) and compared to the National Emergency Laparotomy Audit (NELA) score with Forest plots used to compare 30‐day mortality across CFS and NELA categories. Results A total of 2246 patients (median age 65 years [IQR 51–75]; female 51%) underwent EmLap (60% for colorectal pathology). A total of 10.6% were frail preoperatively (≥CFS 5). As CFS increased so did 30‐day mortality (2.1% CFS1 to 25.3% CFS6 and 7; ꭓ 278.2, p < 0.001) and median LOS (10 days CFS1 to 20 days CFS6 and 7; p < 0.001). Readmission rates did not differ significantly across CFS. ROC (95% CI) for mortality was 0.71 (0.65–0.77) for CFS and 0.84 (0.78–0.89) for NELA. Addition of CFS to NELA did not increase ROC value. Conclusion This study supports the prognostic role of frailty in the emergency surgical setting, finding increasing frailty to be associated with increased mortality and longer LOS in adults of all ages. Although NELA performed better, CFS remained predictive and has the advantage of being calculated preoperatively to aid decision‐making and treatment planning.


INTRODUC TI ON
Traditionally, clinical decision-making in emergency surgery was guided by specific factors such as age, comorbidities and current clinical status and presentation. With none of these providing absolute clinical certainty for outcomes, risk-scoring systems were developed including P-POSSUM (Portsmouth Physiological and operative severity score for the enumeration of mortality and morbidity), that has now been replaced with the National Emergency Laparotomy Audit (NELA) risk calculator score [1][2][3][4]. Developed from the world's largest prospective emergency surgery database, the NELA risk calculator predicts 30-day mortality from twenty pre-and intraoperative variables and is routinely collected across England and Wales.
Recently, frailty has been reported in the emergency surgical population [5,6]. Frailty is associated with increasing age and it can be defined as a "biological syndrome of decreased reserve and resistance to stressors, resulting from cumulative decline across multiple physiological systems and causing vulnerability to adverse outcomes [7]". Using the clinical frailty score (CFS), the UK Emergency Laparotomy and Frailty (ELF) study reported that 20% of older adults ≥65 years undergoing emergency surgery were frail on admission [6,8]. Starting from a CFS of 1 (fit and healthy) and extending up to CFS 7 (severely frail), increasing frailty was found to predict mortality (30 and 90-day), morbidity and discharge destination [6,9].
Although promising for clinical integration, the role of frailty has not been validated in another emergency surgical cohort, nor in younger adults where frailty has been previously reported uncertain. [5,10] Our primary aim was to validate the prognostic role of frailty in a population undergoing emergency surgery that included adults of all ages. The secondary aim was to directly compare the CFS to the NELA risk calculator to assess CFS as a potential point-of-care prognosticator for mortality. indicating at least severe systemic disease. [13] CFS was classified as 1 = very fit, 2 = well, 3 = well with treated comorbid disease, 4 = apparently vulnerable, 5 = mildly frail, 6 = moderately frail and 7 = severely frail with CFS ≥ 5 considered frail [8]. Due to insufficient numbers of patients in CFS 6&7 categories, they were grouped together.

ME THODS
The NELA score was calculated by ELLSA auditors based on preand perioperative data. The percentage outcomes were stratified into low risk (<5%), intermediate risk (5%-10%) and high risk (>10%) as per the Royal College of Surgeons of England "The High-Risk General Surgical Patient: Raising the Standard" report. [14] Operative and post-operative data was recorded by type of operation, postoperative care setting, total hospital length of stay (LOS), 30-day readmission and 30-day mortality.

What does this paper add to the literature?
This is the first paper to show that frailty scoring is prognostic in adults of all ages undergoing emergency surgery.
Although NELA score performs better, the clinical frailty score can be done preoperatively providing guidance for perioperative pathways and decision-making in this highrisk patient population.

Statistical analysis
This was performed using IBM SPSS Statistics version 25.0. Age, NELA score, CFS (six categories: 1, 2, 3, 4, 5, 6 and 7) and LOS were reported using median and interquartile ranges (IQR). The remaining data was reported using numerical figures and percentages. Primary outcome measures were 30-day mortality rate, total LOS and hospital readmission within 30 days. Specifically for LOS, the median for the entire cohort was calculated to allow LOS to be categorised as short or long LOS.
For secondary analysis, areas under the receiver operating characteristic (ROC) curves for the outcome of 30-day mortality were calculated for CFS and NELA score to allow for direct performance comparison along with a Forest plot comparing 30-day mortality between CFS and NELA using adjusted odds ratios with a 95% confidence interval. The proportions of younger (under 65 years) and older people with frailty (CFS 5 or above) were calculated, along with their respective ASA scores.

Comparison of CFS to NELA score in predicting 30-day mortality
The areas under the receiver-operating characteristic (ROC) curve in relation to 30-day mortality with 95% CI for CFS and NELA

DISCUSS ION
This study validates the role of frailty in patients undergoing emergency laparotomy in that preoperative frailty score is associated with greater 30-day mortality and longer total LOS [6]. By including all ages, not just older adults 65 years and above, this study confirms the existence of frailty in the younger emergency surgical population and that frailty negatively impacts on their clinical outcomes [5,10]. Younger and older patients with frailty had similar mean ASA scores. This reinforces the importance of frailty over age as a prognostic marker, though both are important. Comparison of the prognostic usefulness between CFS and the routinely applied NELA Score has not been previously performed and we report that CFS compares favourably to NELA for 30-day mortality. Overall, with the clinical advantages of being free, easy to apply and most importantly, applicable preoperatively, frailty scoring should be routinely implemented into the acute adult surgical setting. [5,6,10] Preoperatively, the NELA score estimates its perioperative parameters, adding a degree of estimation and possible error. This means that the CFS carries significant advantage for surgeons and patients. By helping to correctly identify frail patients quickly, surgeons can then counsel patients and their families fully about risks, engage the appropriate frailty specific clinical pathway and liaise with appropriate allied health professionals to deliver the best, most appropriate, patient-centred care. There may be settings where application of the CFS is not applicable, for example during the COVID-19 pandemic, the National Institute for Health and Care Excellence (NICE) stated that CFS is not appropriate for clinical assessment of patients with stable long-term disabilities, learning disabilities and autism 16 . However, NICE also excluded younger adults, an exclusion that the authors feel is challenged by this work.  is an element of difficulty in directly applying the findings of this Scottish study from a 96% Caucasian base population to these groups [18][19][20]. Cases with any missing covariates were excluded in statistical analysis, hence the final models had reduced sample size and thus prone to type II error which may explain some of the nonsignificant results observed such as 30-day readmission.
Finally, no data validation was performed on the individual data submitted by each local site.

CON CLUS ION
Frailty scoring preoperatively provides prognostic information that can be applied to all adults being considered for emergency surgery.
Although the NELA score performed better, CFS has the clear advantage of being able to be applied rapidly in a time-pressured situation. By virtue of its reliability, simple nature and quick prospective application, CFS can guide shared decision making between clinicians and patients.

ACK N OWLED G M ENTS
We wish to thank David McDonald, Jennifer Edwards and Neil Pekins of ELLSA for their support of this work. We also want to highlight the enthusiasm from all participating Scottish sites in supporting the ELLSA initiative.

CO N FLI C T O F I NTE R E S T
All of the authors declare no conflict of interests.

F I G U R E 2
Forest plot comparing 30-day mortality between CFS and NELA using adjusted odds ratio with 95% CI. CFS categorised into 1, 2, 3, 4, 5, 6 and 7. NELA score categorised using low, intermediate and high. CFS 1 and NELA low score used as reference categories. Blue hashed line is aOR of 1.00. Comparability between CFS 6 and 7 versus CFS 1 and NELA high versus low in predicting mortality based on adjusted odds ratio. CFS, clinical frailty score; CI, 95% confidence interval; NELA, National Emergency Laparotomy Audit CFS 6&7 vs CFS 1 30-day mortality comparison between CFS and NELA score using adjusted odds ratio CFS Abbreviations: CI, 95% confidence interval; NELA, National Emergency Laparotomy Audit. a Adjusted by age linearly, sex (male, female), ASA (1, 2, 3, 4, 5) and sepsis antibiotic provision (not given antibiotics, given antibiotics).