Secondary Logo

Journal Logo

Geriatric Anesthesia: Original Clinical Research Report

The Association of Frailty With Outcomes and Resource Use After Emergency General Surgery: A Population-Based Cohort Study

McIsaac, Daniel I. MD, MPH, FRCPC*†‡§; Moloo, Husein MD, FRCSC, MSc§‖; Bryson, Gregory L. MD, MSc, FRCPC*‡§; van Walraven, Carl MD, FRCPC, MSc§¶

Author Information
doi: 10.1213/ANE.0000000000001960


Conditions requiring emergency general surgery (EGS) are an important public health issue.1,2 Patients undergoing EGS experience high rates of postoperative morbidity and mortality, and consume a substantial amount of health care resources.1,3 Care of our ageing population is a public health priority that intersects with emergency surgery care, because rapid population growth in older demographic groups has a direct effect on the perioperative health care system.4 Individuals older than 65 years of age have surgery more often than any other age group5 and, in addition to the excess risk of having emergency (compared with elective) general surgery, age is an independent predictor of adverse postoperative outcomes and high resource use.6–8

Frailty is an aggregate expression of susceptibility to poor outcomes owing to age-related and disease-related deficits that accumulate within multiple domains.9,10 Preoperative frailty is a well-established risk factor that defines a high risk and high resource use strata of the older surgical population.11,12 At a population level, the prevalence of frailty increases exponentially with age.13 Therefore, as our population ages, caring for frail older patients is likely to become more common in the EGS setting. Although the effect of frailty on outcomes after elective surgery is increasingly well described in single-center11 and population-based studies,14,15 the association of frailty with outcomes and resource use after EGS is not. This is due in part to a lack of a standardized definition of EGS procedures and in part due to a lack of standard of frailty definitions, including frailty definitions that are validated in health administrative data to allow for population-level study of perioperative frailty.

Recent epidemiologic advances have produced a core set of EGS procedures that account for most admissions, complications, and costs in EGS at a national level.1 In Ontario, Canada, population-based health administrative data exist that allow identification of frail patients using validated methods.16,17 Therefore, we sought to measure the association between preoperative frailty and postoperative outcomes and health care resource utilization in the older EGS population.


Following approval by the Sunnybrook Health Sciences Centre Research Ethics Board, we conducted a population-based cohort study in Ontario, Canada, where hospital and physician services are provided to all residents through a publicly funded health care system and recorded in health administrative datasets that are collected using standardized methods.18,19 All data were linked deterministically using encrypted patient-specific identifiers at the Institute for Clinical Evaluative Sciences (ICES; ie, all links across databases were exact based on the anonymized patient identifiers, as opposed to probabilistic linkage that relies on making matches between records that are likely to be from the same individual, but possibly not exact). ICES is an independent research institute that houses the health administrative data for the province of Ontario. Datasets used for the study included the following: the Discharge Abstract Database (DAD), which captures all hospitalizations; the Ontario Health Insurance Plan database, which captures physician service claims; the Continuing Care Reporting System, which records details of long-term and respite care; and the Registered Persons Database (RPDB), which captures all death dates for residents of Ontario.

The analytic data set was created by a trained data analyst independent from the study team. Because the analytic data were generated from data normally collected at ICES, no further data processing was required. Analysis was performed by the lead author and overseen by the senior author. The study protocol was registered at (NCT02835053), and this manuscript is reported per the STrengthening the Reporting of OBservational studies in Epidemiology and the REporting of studies Conducted using Observational Routinely-collected health Data guidelines.20,21


We identified all residents of Ontario who were older than 65 years of age on the date of their first EGS procedure between April 2002 (to coincide with the introduction of International Classification of Diseases [ICD], 10th Revision coding system to identify diagnoses) and March 2014 (the last date for which follow-up data were complete). Patients residing in long-term care facilities before hospital admission were excluded from the study because their living situation was likely to bias study outcomes such as length of stay (LOS), costs of care, and discharge disposition. EGS procedures were defined based on recent recommendations for a core procedure set based on codes obtained from the ICD, 9th Revision.1 These codes were translated into corresponding Canadian Classification of Intervention codes (see Supplemental Digital Content, Supplementary Appendix A,, which have been shown to be accurate based on reabstraction studies.22 Included procedures were large bowel surgery; small bowel surgery; appendectomy; cholecystectomy; control of hemorrhage and suture of duodenal ulcer; lysis of adhesions; and laparotomy. We only included patients who were admitted to hospital on an urgent basis to avoid misclassification of elective surgical patients. Patients without valid provincial health insurance were not included in our data and were therefore ineligible for inclusion.


Frailty status may be ascertained by the use of scales, phenotypes, or through the identification of individuals with frailty-defining diagnoses.23 We identified frailty using the Johns Hopkins Adjusted Clinical Groups (ACG) frailty-defining diagnoses indicator, an instrument designed for use in health administrative data16,17,24 The ACG frailty-defining diagnoses indicator is a binary variable that uses 12 clusters of frailty-defining diagnoses (See Supplemental Digital Content, Supplementary Appendix B, and has been used to study frailty-related health care resource use25,26 and surgical outcomes.14,15,27 No ordinal form of the variable is available because it indicates the presence, or absence, of clusters of frailty-defining diagnoses. Because of the proprietary nature of the ACG system, specific diagnostic codes used are not available for dissemination. We used administrative data available in the 3 years before admission to identify pertinent frailty-defining diagnoses. Identifying diagnostic codes via a lookback window longer than the index hospitalization or admission plus the year before admission has been shown to improve the accuracy of exposure measures in administrative data.28–31

Because there is no gold standard frailty instrument,32 the ACG frailty-defining diagnoses indicator has been tested externally in a comparative analysis with the Vulnerable Elderly Scale (VES), which was recorded from a comprehensive geriatric assessment,16 a gold-standard multidimensional assessment for older people who may be frail.23 The construct validity of the ACG indicator was demonstrated by the significantly greater VES scores in people with a frailty-defining diagnosis compared with those without (P < .005). Patients identified as frail using the ACG indicator had characteristics consistent with multidimensional frailty, including a greater prevalence of falls, lower cognitive scores, and worse global functional scores than nonfrail patients, demonstrating content validity. Using a VES score of ≥3 as a cutoff, the ACG frailty-defining diagnoses indicator had low-to-moderate discrimination between VES frail and VES nonfrail (c-statistic 0.62) patients. The low level of discrimination likely reflects the limited agreement typically found between frailty instruments, where more than 90% of comparisons between instruments result in only fair-to-moderate agreement.32


The primary outcome was death within 365 days of surgery, which was identified from the RPDB (the gold-standard measure of mortality in Ontario). Secondary outcomes included LOS (calculated as the number of inpatient days after surgery from the DAD), postoperative intensive care unit (ICU) admission (identified as a new ICU admission date from the DAD on or after the date of surgery), institutional discharge (defined as a discharge disposition in the DAD to a long-term care facility), and patient-level total costs of care incurred by the provincial health care system from the date of surgery up to 365 days after surgery (using standard algorithms, normalized to 2014 Canadian dollars).33


Demographics were identified from the RPDB. Validated algorithms were used to calculate a Deyo-Charlson Comorbidity Index (CCI) based on ICD-9 and ICD-10 diagnostic codes in ambulatory and in-patient encounters that occurred in the 3 years preceding the index admission.34 We also identified whether each patient was admitted via ambulance or not and whether they were in the ICU before surgery.


Characteristics were compared between frail and nonfrail groups by the use of descriptive statistics. Absolute mortality rates and surgery-specific mortality rates were calculated. Adjusted and unadjusted hazard ratios (HR) measuring the association of frailty with mortality were computed using Cox proportional hazards regression. Because frailty is an aggregate representation of risk, for which medical comorbidities may be on the causal pathway, we did not control for specific medical comorbidities in our primary model35,36; the possibility of comorbidities not being on the causal pathway was addressed in a sensitivity analysis, as described in the Sensitivity Analyses section. We did control for potentially important confounders in the frailty–mortality relationship9,37 including age, sex, socioeconomic status, year of surgery, hospital admission via ambulance, and preoperative ICU admission. Age and year of surgery were treated as continuous variables represented by restricted cubic splines (5 knots for age, 3 for year). Socioeconomic status was modeled as a 5-level categorical variable by the use of neighborhood income quintiles (based on the smallest unit of the national census, representing 400–700 individualsa). Surgery type was modeled as a 7-level categorical variable.

Cox proportional hazards regression assumes that the ratio of hazards (ie, the risk of outcome at a given time) is proportional between exposure levels (in this case, frail and not frail) over the time period studied. Because previous research demonstrated that the risk of postoperative mortality was much greater in the early postoperative period,14 we hypothesized that this violation of the proportional hazards assumption also might be present in the EGS population. Therefore, we tested an interaction term between frailty and postoperative day. To identify an appropriate continuous representation for days of postoperative survival, fractional polynomials (with exponent values of −0.5, −1, 1, 2, 3, and log-transformed) were iteratively tested.38 The linear function was found to provide the best model fit using the Akaike Information Criterion. The adjusted HR for mortality in frail versus nonfrail patients was then calculated for each postoperative day. Previous research also suggested that the effect of frailty on death also differed between different procedures.11,39 To test for this type of effect modification, we asses an interaction term between frailty and procedure by multiplying our binary frailty term by our 7-level categorical procedure variable within our adjusted model.

Secondary outcomes were analyzed with the use of regression techniques appropriate for each outcome. The association of frailty with both institutional discharge and ICU admission were analyzed using logistic regression. The association between frailty and costs was analyzed using a generalized linear model with a log link and γ-distributed errors to account for the skewed distribution of cost data.40 Because in-hospital mortality was a competing risk for hospital discharge, we analyzed LOS as time to discharge based on the subdistributional hazard function according to the methods of Fine and Gray.41 In this analysis, HRs > 1 are associated with a decreased LOS. All secondary analyses were adjusted with the same covariates as the primary analysis.

Sensitivity Analyses

Although comorbidities are on the causal pathway to frailty and were therefore excluded from adjustment in our primary and secondary outcome models, we did complete preplanned sensitivity analyses of all outcomes that included adjustment for CCI (as a 5-level categorical variable based on a CCI of 0, 1, 2, 3, or ≥4) in addition to the covariates included in our initial adjusted outcome models. On the basis of our data, we cannot know the chronological onset of comorbidities and frailty, and therefore felt it appropriate to estimate the degree to which comorbidity might influence outcome risk, in addition to frailty status, to provide insights into different possible causal mechanisms.

SAS version 9.4 (SAS Institute, Cary, NC) for Windows was used for all analyses.

Missing Data

Exposure and outcome status for all participants was complete. Rural residency status was missing for 0.08% of the cohort, and income quintile for 0.4%. Missing values were imputed with the most common values for rural residency status (not rural), and the median value for income quintile (quintile 3). No other data were missing.

Sample Size

Compared with elective surgery, emergency surgery is associated with an approximately 2-fold increase in the risk of postoperative mortality.3 Based on 1-year mortality rates for frail patients in the elective setting of 13.6% and 4.8% in nonfrail patients,14 we anticipated mortality rates of approximately 20% and 10% in frail and nonfrail EGS patients, respectively. Therefore, using a .05 level of significance for the primary outcome, we would have approximately 100% power to detect a true difference if one exists (to achieve 80% power would have required 200 participants in each group). Using a conservative approach to adjustment for multiple primary comparisons (13 in total: 1 primary outcome, 4 secondary outcomes, 1 time interaction model, and a procedure interaction model with 7 procedures), a Bonferroni corrected P value of .05/13 = .004 would signify statistical significance.


Table 1.
Table 1.:
Baseline Characteristics of Study Population by Frailty Status
Figure 1.
Figure 1.:
Yearly absolute number and proportion of patients undergoing emergency general surgery with a frailty-defining diagnosis.

We identified 77,240 patients older than 65 years of age who had EGS during our study period; 56 were excluded because they lived in a long-term care facility at the time of their surgical admission. Of the 77,184 patients analyzed, 19,779 (25.6%) were frail. The absolute number and proportion of the total cohort with a frailty-defining diagnosis increased during the study period (n = 1279 [19.50%] in 2002–2003; n = 1881 [28.13%] in 2013–2014; Figure 1). Frail patients were older, more likely to be female, had a greater CCI score, and were more likely to come to hospital via ambulance and be in ICU before surgery than nonfrail patients (Table 1). Large bowel surgery was the most common procedure in both groups, although the distribution of procedures differed significantly between frail and nonfrail patients.

Frailty and Postoperative Survival

A total of 6626 (33.5%) frail patients died in the year after surgery, compared with 11366 (19.8%) nonfrail patients (crude HR 1.82, 95% confidence interval [CI] 1.77–1.88; P< .0001). One-year crude survival estimates for all surgeries together are provided in Figure 2. The adjusted associations between frailty and outcomes are provided in Table 2.

Table 2.
Table 2.:
Adjusted Associations Between Frailty and Study Outcomes
Figure 2.
Figure 2.:
Crude 1-year survival after emergency general surgery in frail and not frail older patients.

The interaction term between frailty and days of postoperative survival was significant (P< .0001). The daily risk of death for frail versus nonfrail patients is provided in Figure 3. The risk of death in the early postoperative period was substantially greater for frail than nonfrail patients, with an interaction adjusted HR 23.1 (95% CI 22.3–24.1) on postoperative day 1, 22.5 (95% CI 21.5–23.6) on postoperative day 3, and 17.4 (95% CI 16.7–18.2) on day 21. The daily risk of death continued to be greater for frail patients until 9 months after surgery when it approached the null value (ie, HR = 1).

Figure 3.
Figure 3.:
Time-dependent adjusted relative hazard of mortality in frail versus not frail patients. This plot shows the hazard ratio (adjusted for age, sex, year, income quintile, and procedure) and 95% confidence intervals for the association between frailty and mortality in the year after surgery. Hazard ratios above 1 indicate an increased risk of death in frail patients compared with not frail patients.
Figure 4.
Figure 4.:
Forest plot of the relative hazard of 1-year mortality for frail versus nonfrail patients by procedure. Effect measures with point estimates and 95% confidence intervals right of the solid vertical line indicate a significantly increased risk of death for frail patients.

The interaction term between frailty and procedure was also significant (P < .0001). The risk of death for frail patients was significantly greater in all surgery types except for laparotomy and control of hemorrhage and suture of duodenal ulcer (Figure 4).

Secondary Outcomes

Before adjustment, frailty was associated adversely with all secondary outcomes (LOS HR 0.62, 95% CI 0.61–0.63; ICU admission odds ratio [OR] 2.14, 95% CI 2.07–2.21; institutional discharge OR 8.30, 95% CI 7.91–8.71; and 1-year costs of care incidence rate ratio 1.62, 95% CI 1.61–1.63). After adjusting our regression models for potential confounders, including the sociodemographic and surgical factors listed in the Methods section, we found that frailty continued to be adversely associated with all primary and secondary outcomes. The effect estimates for the independent association between frailty and all secondary outcomes are listed in Table 2.

Results of our sensitivity analyses, where all outcome models were additionally adjusted for comorbidity burden, revealed that although all estimated measures of association were attenuated to some extent, only the association with mortality decreased substantively when comorbidities were included in the regression model (mortality HR 1.05, 95% CI 1.02–1.09; LOS HR 0.80, 95% CI 0.79–0.82; ICU admission OR 1.34, 95% CI 1.29–1.39; costs of care incidence rate ratio 1.43, 95% CI 1.41–1.44; institutional discharge OR 5.28, 95% CI 5.01–5.55).

Table 3.
Table 3.:
Rates of ICU Admission by Frailty Status and Procedure

Post hoc, we calculated the proportion of patients admitted to the ICU after surgery for frail and nonfrail patients. These rates are provided in Table 3.


Our findings in a population of older patients undergoing EGS provide several important insights into the epidemiology of perioperative frailty. First, frailty, as measured with the ACG frailty-defining diagnoses indicator, appears to be far more prevalent in EGS patients than in elective surgery patients (25% vs 3%14). Next, the association of frailty with mortality was strongest in the EGS surgeries that were most commonly performed, and in the early postoperative period, which suggest that even in common and less invasive procedures like appendectomy and cholecystectomy, a unique approach to the acute care of frail patients may be required. Finally, frailty was strongly associated with many other adverse postoperative outcomes, and in particular, a 5-fold increase in the odds of institutional discharge for older patients who were community dwelling before admission.

The provision of EGS care is substantively different than elective surgical care. Patients routinely present with acute-on-chronic illnesses, and there is often little time to achieve patient optimization prior to surgery. Although improving the outcomes of frail patients in the elective setting may be easier to conceive because of the increased amount of time available for optimization and decision making, the greater prevalence of frailty in the EGS population suggests that EGS-specific strategies to address the needs and risk profile of frail older patients are needed. Therefore, beyond strategies advocated for frail patients undergoing elective surgery (such as exercise, nutrition, and drug therapies12), a substantial focus is likely needed to address optimization of the health care system itself. On the basis of our data, we suggest that preoperative decision making by patients and families in conjunction with the perioperative team about whether to proceed with surgery, how to mitigate frailty-specific risk, and the need for acute postoperative care as major areas that merit further attention.

Effect modification of the impact of frailty on mortality by surgery type was present in our study, and surprisingly, it was following more common, and lower absolute risk of mortality procedures (like appendectomy and cholecystectomy) that the association of frailty with mortality in our study was most pronounced. This could be because these procedures are not regarded as particularly high risk, and therefore, the presence of a frailty defining diagnosis was not regarded with as much concern when deciding to proceed with a surgical intervention (ie, a willingness by the perioperative team to take a patient with a greater baseline risk to the operating room for a smaller procedure). A similar pattern has been reported in elective surgery, where the presence of frailty before elective surgery had a smaller association with mortality after liver and pancreatic surgery, than after total joint replacements.14 Therefore, routine assessment of frailty and integration of frailty status into preoperative risk stratification, which is already recommended by multiple specialty societies,42,43 should be seriously considered by all perioperative clinicians.

Based on findings from this study, as well as a previous study in elective surgery,14 there is a strong suggestion that the physiologic stress of surgery is poorly tolerated by frail patients. In the elective setting, the adjusted hazard of death was more than 35 times greater for a frail compared with a nonfrail individual on postoperative day 3. In the current study, we found a similar association, with a hazard of death 22 times higher on postoperative day 3 for frail compared with nonfrail EGS patients. Frailty denotes a vulnerability to stressors; therefore, it is not surprising that this vulnerability translates into an increased risk of early mortality after surgery. One strategy to address this early mortality risk may be enhanced postoperative care and monitoring of frail patients. As discussed in the previous paragraph, frailty was more strongly associated with relative risk of mortality after procedures with a lower absolute risk of mortality; patients were also less likely to be admitted to ICU after these surgeries. In fact, in the 2 procedures in which frailty was not associated with an increased risk of death, more than 50% of frail and nonfrail patients alike were admitted to ICU after surgery. Therefore, although discharge directly to the general surgery ward from the postanesthesia unit may be appropriate for younger and healthier patients undergoing appendectomy or cholecystectomy, this may not be appropriate for older frail patients having these same surgeries.

Current evidence supports the role of orthogeriatric care for emergency hip fracture patients,44 and evidence to support other approaches to enhanced care for high-risk elderly patients after acute care surgery is pending.45 In the interim, our findings suggest that frail, older patients undergoing EGS may require more careful monitoring even after lower risk of mortality procedures, and suggests that available guidelines, such those provided by the American Geriatrics Society/American College of Surgeons be considered to enhance contemporary postoperative care.46

Interestingly, in both elective14 and the current EGS studies, secondary analyses that adjusted for comorbidities (which are on the causal pathway to frailty) attenuated the frailty-mortality association. This suggests that comorbidity-related mechanisms may underlie much of the physiologic vulnerability to surgical stress that we hypothesize contributes to early mortality risk. Therefore, future efforts to reduce postoperative mortality will need to either risk stratify older patients more accurately based on their comorbid and/or physiologic status to allow enhancement their physiologic status in a reasonable timeframe before surgery, as well as guiding the enhanced care and monitoring that we discussed in the previous paragraph.

Although comorbidity is thought to be on the causal pathway to frailty (and was therefore excluded from our primary adjusted analysis), we cannot know for certain in our data the timing of comorbidity and frailty onset. Therefore, it is important to consider the impact that comorbidity adjustment may have on other outcome measures. Interestingly, the frailty-outcome association for our nonmortality outcomes was attenuated far less when additional adjustment for comorbidity was performed. This suggests that other dimensions of frailty may be the major drivers of prolonged hospital LOS, ICU admissions, costs, and institutional discharge. In particular, the rate of discharge to an institution for frail individuals in our study, all of whom were community dwelling before surgery, is striking. Almost 30% of frail patients were discharged to institutional care after EGS, and the relative increase in the odds of institutional discharge virtually was unchanged between our secondary analysis without comorbidity adjustment and our sensitivity analysis with comorbidity adjustment (OR 5.82 vs 5.28). We hypothesize that noncomorbid aspects of frailty, such as functional, cognitive, or psychosocial status, may be more likely to drive the loss of independence experienced by many frail patients undergoing EGS. This merits future prospective evaluation to support the development of specific interventions to limit loss of independence after surgery. Furthermore, loss of independence at discharge is associated with the development of further adverse events later in the postoperative course.47 Therefore, in discussions of preoperative risk with frail patients and their families, anesthesiologist and surgeons should highlight outcomes such as discharge disposition and possible loss of independence.

This study features several strengths. We used a well-defined grouping of EGS procedures to define our study population, in addition to a validated frailty instrument to define exposure. Our primary outcome, mortality, was captured by use of the gold standard source in Ontario, and secondary outcomes were derived by the use of defined methods with proven accuracy and reliability. Furthermore, our study protocol was registered a priori, which minimizes the risk of multiple outcome testing.

Limitations also must be considered. Although the ACG frailty-defining diagnoses indicator is validated, it is a binary measure that does not allow for ascertainment of the degree of frailty at the individual level, or reveal the specific frailty defining diagnoses present. Furthermore, we used health administrative data that was not initially collected for research purposes. Despite using a suggested set of procedures to define our cohort, some procedures were relatively rare, and our ability to provide precise findings for these procedures may be limited. Finally, we were able to measure only associations between frailty and outcomes due to the observational nature of our study.


In older patients having EGS, frailty is associated with an increased risk of mortality, especially early in the postoperative period, in addition to increased LOS, ICU admissions, and costs of care. Importantly, frail patients are at an exceptionally high risk of loss of independence through institutional discharge. Efforts are needed to improve the care and outcomes of older frail surgical patients in the unique setting of EGS.


Name: Daniel I. McIsaac, MD, MPH, FRCPC.

Contribution: This author helped conceive and design the study, analyze and interpret the data, and write and revise the manuscript.

Name: Husein Moloo, MD, FRCSC, MSc.

Contribution: This author helped conceive and design the study, interpret the data, and revise the manuscript.

Name: Gregory L. Bryson, MD, MSc, FRCPC.

Contribution: This author helped conceive and design the study, interpret the data, and revise the manuscript.

Name: Carl van Walraven, MD FRCPC MSc.

Contribution: This author helped conceive and design the study, analyze and interpret the data, and revise the manuscript.

This manuscript was handled by: Robert Whittington, MD.


aAvailable at: Accessed July 14, 2016.


1. Scott JW, Olufajo OA, Brat GA, et al. Use of national burden to define operative emergency general surgery. JAMA Surg. 2016;2115:e160480.
2. Gale SC, Shafi S, Dombrovskiy VY, Arumugam D, Crystal JS. The public health burden of emergency general surgery in the United States: a 10-year analysis of the Nationwide Inpatient Sample—2001 to 2010. J Trauma Acute Care Surg. 2014;77:202–208.
3. Haider AH, Obirieze A, Velopulos CG, et al. Incremental cost of emergency versus elective surgery. Ann Surg. 2015;262:260–266.
4. Etzioni DA, Liu JH, Maggard MA, Ko CY. The aging population and its impact on the surgery workforce. Ann Surg. 2003;238:170–177.
5. Etzioni DA, Liu JH, O’Connell JB, Maggard MA, Ko CY. Elderly patients in surgical workloads: a population-based analysis. Am Surg. 2003;69:961–965.
6. Hamel MB, Henderson WG, Khuri SF, Daley J. Surgical outcomes for patients aged 80 and older: morbidity and mortality from major noncardiac surgery. J Am Geriatr Soc. 2005;53:424–429.
7. Turrentine FE, Wang H, Simpson VB, Jones RS. Surgical risk factors, morbidity, and mortality in elderly patients. J Am Coll Surg. 2006;203:865–877.
8. Ingraham AM, Cohen ME, Raval MV, Ko CY, Nathens AB. Variation in quality of care after emergency general surgery procedures in the elderly. J Am Coll Surg. 2011;212:1039–1048.
9. Rockwood K, Song X, MacKnight C, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ. 2005;173:489–495.
10. Fried LP, Tangen CM, Walston J, et al.; Cardiovascular Health Study Collaborative Research Group. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56:M146–M156.
11. Beggs T, Sepehri A, Szwajcer A, Tangri N, Arora RC. Frailty and perioperative outcomes: a narrative review. Can J Anaesth. 2015;62:143–157.
12. Partridge JS, Harari D, Dhesi JK. Frailty in the older surgical patient: a review. Age Ageing. 2012;41:142–147.
13. Gale CR, Cooper C, Sayer AA. Prevalence of frailty and disability: findings from the English Longitudinal Study of Ageing. Age Ageing. 2014;0:1–4.
14. McIsaac DI, Bryson GL, van Walraven C. Association of frailty and 1-year postoperative mortality following major elective noncardiac surgery: a population-based cohort study. JAMA Surg. 2016;151:538–545.
15. McIsaac DI, Beaule PE, Bryson GL, van Walraven C. The impact of frailty on outcomes and healthcare resource utilization after total joint arthroplasty: a population-based cohort study. Bone Joint J. 2016;98-B:799–805.
16. Sternberg SA, Bentur N, Abrams C, et al. Identifying frail older people using predictive modeling. Am J Manag Care. 2012;18:e392–e397.
17. Lieberman R, Abrams C, Weiner J. Development and Evaluation of the Johns Hopkins University Risk Adjustment Models for Medicare+Choice Plan Payment. 2003.Baltimore, MD: Centers for Medicare and Medicaid Services, U.S. Department of Health and Human Services.
18. Vision and Mandate. Canadian Institute of Health Information. Available at: Accessed January 25, 2017.
19. About Us. Institute for Clinical and Evaluative Sciences. Available at: Accessed January 25, 2017.
20. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 2007;335:806–808.
21. Benchimol EI, Smeeth L, Guttmann A, et al.; RECORD Working Committee. The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement. PLoS Med. 2015;12:e1001885.
22. Juurlink DN, Croxford R, Chong A, Austin P, Tu J, Laupacis A. Canadian Institute for Health Information Discharge Abstract Database: A Validation Study. ICES Investigative Report June 2006 Canadian Institute for Health Information Discharge Abstract Database. 2006.Toronto: Institute for Clinical Evaluative Science.
23. Turner G, Clegg A; British Geriatrics Society; Age UK; Royal College of General Practioners. Best practice guidelines for the management of frailty: a British Geriatrics Society, Age UK and Royal College of General Practitioners report. Age Ageing. 2014;43:744–747.
24. Technical Reference Guide 9.0. 2009. Baltimore, MD;
25. Bronskill S, Carter M, Costa A, et al. Aging in Ontario: An ICES Chartbook of Health Service Use by Older Adults— A Technical Report. 2010.Toronto, ON, Canada: Institute for Clinical Evaluative Sciences.
26. Bronskill S, Camacho X, Gruneir A, Ho M. Health System Use by Frail Ontario Seniors: An In-Depth Examination of Four Vulnerable Cohorts. 2011.Toronto, ON, Canada: Institute for Clinical Evaluative Sciences.
27. Neuman HB, Weiss JM, Leverson G, et al. Predictors of short-term postoperative survival after elective colectomy in colon cancer patients ≥ 80 years of age. Ann Surg Oncol. 2013;20:1427–1435.
28. Chen G, Lix L, Tu K, et al. Influence of using different databases and “look back” intervals to define comorbidity profiles for patients with newly diagnosed hypertension: implications for health services researchers. PLoS One. 2016;11:e0162074.
29. Preen DB, Holman CD, Spilsbury K, Semmens JB, Brameld KJ. Length of comorbidity lookback period affected regression model performance of administrative health data. J Clin Epidemiol. 2006;59:940–946.
30. Zhang JX, Iwashyna TJ, Christakis NA. The performance of different lookback periods and sources of information for Charlson comorbidity adjustment in Medicare claims. Med Care. 1999;37:1128–1139.
31. Nedkoff L, Knuiman M, Hung J, Sanfilippo FM, Katzenellenbogen JM, Briffa TG. Concordance between administrative health data and medical records for diabetes status in coronary heart disease patients: a retrospective linked data study. BMC Med Res Methodol. 2013;13:121.
32. Theou O, Brothers TD, Mitnitski A, Rockwood K. Operationalization of frailty using eight commonly used scales and comparison of their ability to predict all-cause mortality. J Am Geriatr Soc. 2013;61:1537–1551.
33. Wodchis W, Bushmeneva K, Nikitovic M, McKillop I. Guidelines on Person-Level Costing Using Administrative Databases in Ontario. 2013.Toronto, ON, Canada: Health System Performance Research Network.
34. Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43:1130–1139.
35. Hernán M a, Hernández-Díaz S, Werler MM, Mitchell AA. Causal knowledge as a prerequisite for confounding evaluation: an application to birth defects epidemiology. Am J Epidemiol. 2002;155:176–184.
36. Fried LP, Ferrucci L, Darer J, Williamson JD, Anderson G. Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care. J Gerontol A Biol Sci Med Sci2004;59:255–263.
37. Rockwood K, Song X, Mitnitski A. Changes in relative fitness and frailty across the adult lifespan: evidence from the Canadian National Population Health Survey. CMAJ. 2011;183:E487–E494.
38. Sauerbrei W, Royston P. Building multivariable prognostic and diagnostic models—transformation of the predictors by using fractional polynomials. J R Stat Soc Ser A Stat Soc. 2002;165:399–400.
39. McIsaac DI, Wijeysundera DN, Huang A, Bryson GL, van Walraven C. Association of the hospital volume of frail surgical patients cared for with outcomes after elective, major noncardiac surgery: a retrospective population-based cohort study. Anesthesiology. 2017;126:602–613.
40. Austin PC, Ghali WA, Tu JV. A comparison of several regression models for analysing cost of CABG surgery. Stat Med. 2003;22:2799–2815.
41. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496.
42. Griffiths R, Beech F, Brown A, et al.; Association of Anaesthetists of Great Britain and Ireland. Peri-operative care of the elderly: Association of Anaesthetists of Great Britain and Ireland. Anaesthesia. 2014;69(Suppl 1):81–98.
43. Chow WB, Rosenthal RA, Merkow RP, Ko CY, Esnaola NF; American College of Surgeons National Surgical Quality Improvement Program; American Geriatrics Society. Optimal preoperative assessment of the geriatric surgical patient: a best practices guideline from the American College of Surgeons National Surgical Quality Improvement Program and the American Geriatrics Society. J Am Coll Surg. 2012;215:453–466.
44. Prestmo A, Hagen G, Sletvold O, et al. Comprehensive geriatric care for patients with hip fractures: a prospective, randomised, controlled trial. Lancet. 2015;385:1623–1633.
45. Khadaroo RG, Padwal RS, Wagg AS, Clement F, Warkentin LM, Holroyd-Leduc J. Optimizing senior’s surgical care—elder-friendly approaches to the surgical environment (EASE) study: rationale and objectives. BMC Health Serv Res. 2015;15:338.
46. Mohanty S, Rosenthal RA, Russell MM, Neuman MD, Ko CY, Esnaola NF. Optimal perioperative management of the geriatric patient: a best practices guideline from the American College of Surgeons NSQIP and the American Geriatrics Society. J Am Coll Surg. 2016;222:930–947.
47. Berian J, Mohanty S, Ko CY, Rosenthal RA, Robinson TN. Association of loss of independence with readmission and death after discharge in older patients after surgical procedures. JAMA Surg. 2016:151:e161689.
Copyright © 2017 International Anesthesia Research Society