eTable 1. Frequency of Procedure Current Procedural Terminology (CPT) Codes
eTable 2. Number and Unadjusted Frequency of 30-Day Complications by NSQIP mFI Category in Patients After Ambulatory and 23-Hour-Stay Hernia, Breast, Thyroid, or Parathyroid Surgery in the 2007-2010 ACS NSQIP PUF
Customize your JAMA Network experience by selecting one or more topics from the list below.
Seib CD, Rochefort H, Chomsky-Higgins K, et al. Association of Patient Frailty With Increased Morbidity After Common Ambulatory General Surgery Operations. JAMA Surg. 2018;153(2):160–168. doi:10.1001/jamasurg.2017.4007
Is frailty associated with perioperative morbidity in patients undergoing ambulatory hernia, breast, thyroid, or parathyroid surgery?
In this cohort study, an increasing National Surgical Quality Improvement Program modified frailty index was associated with a stepwise increase in the incidence of 30-day complications; an intermediate modified frailty index (2-3 frailty traits) and a high modified frailty index (≥4 frailty traits) were significantly associated with any type of complication and with serious complications, respectively, in multivariable analysis adjusting for other risk factors and clustering by operation. Anesthesia with local and monitored anesthesia care was the only modifiable covariate associated with decreased odds of serious 30-day complications.
Frailty is independently associated with perioperative morbidity in patients undergoing common ambulatory general surgery and should be considered in patient selection and counseling.
Frailty is a measure of decreased physiological reserve that is associated with morbidity and mortality in major elective and emergency general surgery operations, independent of chronological age. To date, the association of frailty with outcomes in ambulatory general surgery has not been established.
To determine the association between frailty and perioperative morbidity in patients undergoing ambulatory general surgery operations.
Design, Setting, and Participants
A retrospective cohort study was conducted of 140 828 patients older than 40 years of age from the 2007-2010 American College of Surgeons National Surgical Quality Improvement Program Participant Use File who underwent ambulatory and 23-hour-stay hernia, breast, thyroid, or parathyroid surgery. Data analysis was performed from August 18, 2016, to June 21, 2017.
Main Outcomes and Measures
The association between the National Surgical Quality Improvement Program modified frailty index and perioperative morbidity was determined via multivariable logistic regression with random-effects modeling to control for clustering within Current Procedural Terminology codes.
A total of 140 828 patients (80 147 women and 60 681 men; mean [SD] age, 59.3 [12.0] years) underwent ambulatory hernia (n = 71 455), breast (n = 51 267), thyroid, or parathyroid surgery (n = 18 106). Of these patients, 2457 (1.7%) experienced any type of perioperative complication and 971 (0.7%) experienced serious perioperative complications. An increasing modified frailty index was associated with a stepwise increase in the incidence of complications. In multivariable analysis adjusting for age, sex, race/ethnicity, anesthesia type, tobacco use, renal failure, corticosteroid use, and clustering by Current Procedural Terminology codes, an intermediate modified frailty index score (0.18-0.35, corresponding to 2-3 frailty traits) was associated with statistically significant odds ratios of 1.70 (95% CI, 1.54-1.88; P < .001) for any complication and 2.00 (95% CI, 1.72-2.34; P < .001) for serious complications. A high modified frailty index score (≥0.36, corresponding to ≥4 frailty traits) was associated with statistically significant odds ratios of 3.35 (95% CI, 2.52-4.46; P < .001) for any complication and 3.95 (95% CI, 2.65-5.87; P < .001) for serious complications. Anesthesia with local and monitored anesthesia care was the only modifiable covariate associated with decreased odds of serious 30-day complications, with an adjusted odds ratio of 0.66 (95% CI, 0.53-0.81; P < .001).
Conclusions and Relevance
Frailty is associated with increased perioperative morbidity in common ambulatory general surgery operations, independent of age, type of anesthesia, and other comorbidities. Surgeons should consider frailty rather than chronological age when counseling and selecting patients for elective ambulatory surgery.
The population of patients older than 65 years of age continues to increase and alter the demographics of those seeking surgical care. In 2006, patients older than 65 years of age made up 32% of patients undergoing ambulatory surgery in the United States, and with this group projected to double in number from 2010 to 2050, this percentage is likely to increase accordingly.1,2 As the baby boomer generation ages and life expectancy increases, surgical services will have to focus on providing safe and specialized care for this aging population in the ambulatory setting.
Frailty is a measure of decreased physiological reserve that results from impairments in multiple organ systems and can be distinguished from the aging process and comorbidity. Multiple studies have shown that frailty is associated with increased risk of complications in a wide range of elective and emergency operations.3-8 These effects appear to be independent of chronological age. However, there is evidence that older patients are less likely than younger patients to receive standard of care operative management for benign and malignant conditions, even after adjusting for health status and preference.9-12 In addition, higher-than-anticipated complication rates for younger frail patients13 suggest that operative risks should be carefully weighed in this group and that informed consent must take into account more than just chronological age. Preoperative assessment of frailty is critical to ensuring that indicated operations are performed on the right patients and that patients are well informed of their surgical risk. In addition, there is evidence that preoperative targeted interventions to optimize medical comorbidities, nutrition, and physical fitness in at-risk patients can improve outcomes and postoperative functional status and reduce complications and length of stay (LOS).14-17
There are few studies addressing the association of frailty with outcomes in patients undergoing ambulatory and 23-hour-stay surgery. Although operations performed in this setting are generally considered to be low risk, studies of surgical outcomes that stratify patients based on age may be subject to selection bias and fail to address the association of impaired physiological reserve with complications and quality of life. In addition, results from single-center studies of elderly patients are often superior to those in population-based studies, suggesting that continued investigation is needed to inform decision making in this group.3 The aim of this study is to better understand the association of frailty with perioperative outcomes of patients undergoing common ambulatory general surgery operations. Further information on this association will allow us to appropriately counsel elderly patients on the risks of surgery; improve access to care for older, nonfrail patients; and intervene preoperatively to improve patient fitness and surgical outcomes.
We used the 2007-2010 American College of Surgeons National Surgical Quality Improvement Program (NSQIP) Participant Use File for our analysis. This database contains prospective, multi-institutional information on patient characteristics and 30-day morbidity and mortality outcomes for a systematic sample of major inpatient and outpatient surgical procedures at participating institutions. Trained surgical clinical reviewers obtain complete 30-day follow-up data on selected patients with regular interrater reliability audits.18 The collection of NSQIP data has been previously described.19 This study was exempt from approval by the University of California, San Francisco Institutional Review Board because it involved analysis of deidentified patient data. Informed consent was not required because data were deidentified.
Patients undergoing ambulatory surgery were identified using variables for outpatient procedures and total LOS. Surgical procedures listed as emergencies or that did not use anesthesia were excluded. Ambulatory hernia surgery, breast surgery, thyroid surgery, and parathyroid surgery were identified using Current Procedural Terminology codes (eTable 1 in the Supplement). We did not include thyroid or parathyroid surgery requiring sternotomy.
Frailty was assessed using the NSQIP modified frailty index (mFI), a measure of frailty based on 11 NSQIP variables that is adapted from the Canadian Study of Health and Aging Frailty Index20 and has been validated to correlate with frailty in all surgical specialties.7,13,21,22 The variables included in the mFI are listed in the Box. The mFI scores are obtained by adding the total number of present variables and dividing by 11, the total possible points. The mFI was tested in multivariable models as a continuous predictor and then broken down into the following multilevel categories to facilitate the clinical application of study findings: low mFI (score ≤0.18, corresponding to 0-1 frailty traits), intermediate mFI (score 0.18-0.35, corresponding to 2-3 frailty traits), and high mFI (score ≥0.36, corresponding to ≥4 frailty traits). Additional covariates included in multivariable models based on clinical relevance were age, sex, race/ethnicity, type of anesthesia, tobacco use, renal failure, and corticosteroid use.
History of diabetes
Impaired functional status
History of chronic obstructive pulmonary disease or pneumonia
History of congestive heart failure
History of myocardial infarction within 6 months
History of percutaneous coronary intervention
Cardiac surgery or angina
Receipt of hypertensive medications
Peripheral vascular disease or rest pain
History of transient ischemic attack or cerebrovascular accident with persistent residual deficit
The NSQIP contains comprehensive data on 30-day perioperative complications and mortality. Primary outcomes were any type of 30-day complication and serious 30-day complications. Any type of complication included pneumonia, unplanned intubation, ventilator dependence, cardiac arrest or myocardial infarction, stroke or coma for longer than 24 hours, acute or progressive renal failure, bleeding, sepsis, surgical site infections, wound dehiscence, venous thromboembolism, and urinary tract infections. Serious complications excluded urinary tract infections and superficial surgical site infections, which do not have uniform clinical significance. We chose to include in our primary outcomes complications that are not life threatening but that may significantly affect the quality of life, recovery, and long-term functional outcomes of frail patients undergoing elective surgery. Additional 30-day outcomes included mortality and LOS.
Statistical analysis was performed from August 18, 2016, to June 21, 2017. χ2 Statistics and analysis of variance were used to compare differences in perioperative outcomes between mFI categories. Multivariable logistic regression was used to calculate odds ratios (ORs) for the association between mFI category and primary outcomes adjusting for prespecified covariates. Clustering by Current Procedural Terminology codes using multilevel modeling with random effects was used to account for intraprocedure variation in the overall model. Adjusted ORs were also calculated within surgery type. Multivariable linear regression was used to compare LOSs in each group. The significance level for all comparisons was P < .05, and the tests were 2-tailed. Analysis was performed using STATA statistical software, version 13.0 (StataCorp LP).
We identified 140 828 patients older than 40 years of age who underwent ambulatory hernia (n = 71 455), breast (n = 51 267), thyroid, or parathyroid surgery (n = 18 106). Baseline demographic and comorbidity characteristics are listed in Table 1. A total of 118 831 patients had low mFI scores, 21 036 had intermediate mFI scores, and 961 had high mFI scores. As expected in the ambulatory setting, more than 99% of patients (n = 139 732) had independent functional status on admission, and most had no or mild systemic disease (102 539 [72.8%] patients with American Society of Anesthesiologists class I or II). However, 192 patients (20.0%) with high mFI scores had dependent functional status prior to admission compared with 255 (0.2%) patients with low mFI scores. The most prevalent comorbidities contributing to frailty in patients with intermediate and high mFI scores were hypertension (20 201 [96.0%] with intermediate mFI scores and 946 [98.4%] with high mFI scores), insulin-dependent and non–insulin-dependent diabetes (11 046 [52.5%] with intermediate mFI scores and 699 [72.7%] with high mFI scores), and coronary artery disease (7286 [34.6%] with intermediate mFI scores and 760 [79.1%] with high mFI scores). A total of 73 033 patients (51.9%) had no frailty traits, 45 798 (32.5%) had 1 frailty trait, 21 036 (14.9%) had 2 to 3 frailty traits, and 959 (0.7%) had 4 or more frailty traits (Table 2).
An increasing mFI score was associated with a stepwise increase in the unadjusted incidence of complications (Table 2). Overall complication rates were low, and 30-day mortality was rare. Any type of postoperative complication occurred in 2457 patients (1.7%), and serious postoperative complications occurred in 971 patients (0.7%). An increasing mFI score was associated with a statistically significant increase in the occurrence of any complication (1828 of 118 831 [1.5%] in patients with low mFI scores and 56 of 961 [5.8%] in those with high mFI scores; P < .001), serious complications (671 of 118 831 [0.6%] in patients with low mFI scores and 29 of 961 [3.0%] in those with high mFI scores; P < .001), and 30-day mortality (52 of 118 831 [0.04%] among patients with low mFI scores and 3 of 961 [0.3%] among those with high mFI scores; P < .001). The percentage of patients with an LOS longer than 1 day was greater among patients with intermediate mFI scores (663 of 21 036 [3.2%]) and patients with high mFI scores (52 of 961 [5.4%]) than among those with low mFI scores (2297 of 118 831 [1.9%]; P < .001). Data on the unadjusted frequencies of specific complications are provided in eTable 2 in the Supplement. The most common complications were superficial incisional, deep incisional, and organ or space surgical site infections, followed by urinary tract infections, pulmonary complications, and sepsis.
In multivariable logistic regression models, the mFI was strongly associated with the occurrence of complications. In multivariable analysis, intermediate mFI scores were associated with ORs of 1.70 (95% CI, 1.54-1.88; P < .001) for any complication and 2.00 (95% CI, 1.72-2.34; P < .001) for serious complications (Table 3). High mFI scores were associated with ORs of 3.35 (95% CI, 2.52-4.46; P < .001) for any complication and 3.95 (95% CI, 2.65-5.87; P < .001) for serious complications. In the adjusted model, being 80 to 89 years of age (OR, 1.36; 95% CI, 1.04-1.77; P = .03) or 90 years of age or older (OR, 2.27; 95% CI, 1.21-4.24; P = .01) was also associated with increased odds of serious complications. The only modifiable covariate associated with decreased odds of serious complications in the multivariable model was anesthesia with local and monitored anesthesia care with an OR of 0.66 (95% CI, 0.53-0.81, P < .001). On multivariable linear regression, intermediate and high mFI scores were associated with increases in LOS of 0.099 days (95% CI, 0.054-0.144; P < .001) and 0.261 days (95% CI, 0.072-0.450; P = .007), respectively. In prespecified subgroup multivariable analysis of hernia, breast, thyroid, or parathyroid surgery, the mFI was still associated with any type of 30-day complications and serious 30-day complications in most analyses (Table 4). However, for cervical endocrine surgery, the mFI was not associated with increased odds of serious complications in the adjusted analysis.
Our understanding of the association of frailty with outcomes of patients undergoing the acute stress of surgery has expanded significantly in recent years. It is well supported that frail geriatric patients experience increased complications, loss of functional status, and prolonged recovery after a wide range of elective and emergency surgical procedures.3-8 However, there has been little focus on how frailty affects patients undergoing common ambulatory general surgery. The assessment of frailty in a geriatric patient can be relevant to both preoperative risk stratification and risk modification to improve outcomes. Therefore, our goal for frailty assessment in the preoperative setting should be to improve patient selection and the informed consent process for frail patients, regardless of chronological age, and to remove barriers to treatment for older nonfrail patients. The results of this study will improve our ability to do this for patients undergoing common general surgery operations.
Our data show that frailty is associated with increased adjusted odds of 30-day morbidity after hernia, breast, thyroid, and parathyroid surgery. Although complication rates were low overall, the relative risk of complications was increased, with patients with 2 to 3 frailty traits (intermediate mFI score) having more than 2 times the odds of serious complications, after adjusting for poor prognostic factors. Hypertension, diabetes, and coronary artery disease were the most common comorbidities contributing to frailty, so the presence of 2 or more of these conditions should be noted in preoperative evaluations. In addition, our findings show that the assessment of preoperative frailty should affect the informed consent process in an elective setting because the risks of operative intervention may outweigh the benefits for certain frail patients.
Inguinal and ventral hernias can have a significant negative effect on quality of life and have the potential to cause life-threatening complications. Although elective inguinal hernia repair is performed routinely on an outpatient basis with low morbidity among young and old patients,23-25 published surgical outcomes are subject to selection bias and do not include patients who are not referred for or are denied hernia repair based on age and comorbidities. Older patients make up a disproportionate number of those who undergo emergency hernia repair owing to strangulation or bowel obstruction, which suggests that access to care or surgeon decision making in the elective setting may leave more elderly patients at risk for these complications.26 Emergency hernia repair in this group is associated with increased morbidity and mortality26-28 and often results in reoperations and hernia recurrence, leading to a cycle of high-risk surgery, complications, and cost.29 Our results should serve as evidence that older patients with few frailty traits have low risks of complications after elective hernia repair and can be safely offered these surgical procedures. Using preoperative assessment tools that support elective hernia repair for nonfrail elderly individuals may minimize morbidity and cost. However, complication rates and adjusted ORs do increase for patients older than 80 years of age; for this group of patients, more emphasis should be placed on patient selection based on surgical indication and risk modification.
Age appears to be a barrier to appropriate oncologic management of women with breast cancer. A large proportion of surgical procedures for breast cancer are now performed in the ambulatory setting,30 and morbidity for breast-conserving therapy, mastectomy, and aesthetic breast surgery is uncommon.31,32 However, there is evidence that, as a whole, older women with breast cancer are less likely to receive standard oncologic surgical management of their disease.33 Studies have shown that patients 85 years of age or older still have lower odds of undergoing surgery for breast cancer after controlling for patient preference and comorbidities.9 Although the diagnosis of early-stage breast cancers (stage I) may not significantly affect the life expectancy of older patients, overall survival is decreased among those who receive less-aggressive surgical and medical therapy, regardless of the stage of the cancer.10 Our findings should help minimize the role of chronological age in patient selection for ambulatory treatment of breast cancer and should encourage frailty assessment to identify patients at risk for complications or functional decline so that appropriate discussion of medical therapy can be had with patients in this group.
Many single-center studies show that thyroid surgery and parathyroid surgery are safe for older patients, with overall and endocrine-specific complication rates similar to those in their younger counterparts.34-37 In population-based studies, however, older age is independently associated with a longer LOS and morbidity,38,39 possibly associated with the effect of low-volume surgeons.38,40-42 There is evidence that a patient’s age also negatively affects decisions about the management of differentiated thyroid cancer. The prevalence of thyroid nodules and the risk of associated malignant neoplasms increases with age.43,44 Differentiated thyroid cancer in older patients has also been associated with more aggressive disease, such as extrathyroidal extension and distant metastases, in addition to increased recurrence and disease-specific mortality.11,45,46 However, large database studies have shown that patients older than 65 years of age with differentiated thyroid cancers larger than 1 cm are less likely to be treated according to guidelines with thyroidectomy, lymph node dissection, and/or radioactive iodine ablation, despite their more advanced disease and the survival benefit of these therapies.11 Therefore, preoperative assessment of elderly patients requiring thyroidectomy is critical to make appropriate management decisions, and high-volume surgeons may be better equipped to select and operate on appropriate patients.
A patient’s age affects the likelihood of managing primary hyperparathyroidism with surgical therapy. The prevalence of primary hyperparathyroidism increases with age and is estimated to be 1.5% among patients 65 years of age or older and 3.4% among postmenopausal women.47 Surgery is the only definitive treatment for primary hyperparathyroidism, and long-term observational studies demonstrate improvement in bone mineral density and decreased overall mortality owing to cardiovascular events after successful parathyroidectomy.48-53 Because many elderly patients with primary hyperparathyroidism are asymptomatic with mild hypercalcemia, surveillance is often advocated. The older a patient is, the less likely he or she is to undergo a parathyroidectomy, even when the patient meets consensus criteria for operative management and after adjusting for comorbidity and biochemical parameters.12 In addition, older patients experience significantly longer delays to surgical referral, exposing them to additional negative effects on the skeletal, renal, and cardiovascular systems.12,37 Our data show that a preoperative frailty assessment will improve our ability to identify patients with primary hyperparathyroidism who are at higher risk of complications and those who will benefit from a parathyroidectomy.
The only factor associated with decreased odds of complications in our analysis was anesthesia with local and monitored anesthesia care. These results are consistent with those of single-center studies of elderly patients undergoing inguinal hernia repair that have shown low complication rates, comparable to those of younger, healthier patients, with the use of local anesthesia.54,55 For frail patients who choose to undergo hernia repair, local and monitored anesthesia care should be used whenever possible. Reports have documented the successful use of local anesthesia in breast-conserving therapy and mastectomy, although most studies are small, have focused on feasibility without conversion to general anesthesia, and have not shown significant differences in complications or oncologic outcomes.56-60 The use of local with monitored anesthesia care may be challenging in complex surgical procedures for breast cancer, such as modified radical mastectomy or axillary dissection, but it should be considered for patients with increased anesthesia risk who are undergoing ambulatory breast surgery. Most studies of local anesthesia in thyroidectomy and parathyroidectomy report using it in combination with regional superficial and deep cervical plexus blocks. Most large studies are from high-volume, tertiary referral centers, with many being single-surgeon series.61-63 Although their results suggest this technique is feasible, it is not likely that outcomes are generalizable, especially given the inability to use nerve monitoring, which has become routine at many centers, and the difficulty converting a cervical operation to general endotracheal anesthesia, which is required in 3% to 10% of operations.61-63
Many limitations of our study are inherent to those that use large administrative databases and are due to the nature of data collection and reporting. Outcomes reported in the NSQIP are limited to 30 days, which is a short period to assess the physiological and functional effects of surgery on older patients, given evidence of the increased risk of progressive functional decline and death at 1 year and longer after major general surgery.64 In addition, the NSQIP prior to 2011 does not include data on readmissions, which are associated with persistent functional deficits in elderly patients and are a critical perioperative outcome in frail populations.65 Owing to the data available in the NSQIP, our ability to assess frailty was limited, and we chose to use the mFI because of its validation and use in prior studies.13,21,66 Many variables included in the mFI are comorbidities, and although they map to variables from the Canadian Study of Health and Aging Frailty Index and contribute to frailty as accumulated deficits that make patients vulnerable to acute stressors, they provide an incomplete picture of patient frailty. An evaluation of functional status, cognition, and mobility would have allowed a more comprehensive assessment. The NSQIP does not include several adverse outcomes documented in elderly patients after surgery, such as postoperative depression, anxiety, loss of independence, and accelerated cognitive decline, so we may not be capturing the magnitude of the negative effect these operations have on this vulnerable patient population.64 In addition, large academic teaching institutions are overrepresented in the NSQIP, which may affect the generalizability of our results to community hospitals that perform a large proportion of common ambulatory general surgery procedures. Finally, the NSQIP stopped collecting data on many variables in the validated mFI after 2010,67 so we were unable to include recent years in our analysis.
Frailty is associated with worse perioperative outcomes in patients undergoing ambulatory hernia, breast, thyroid, or parathyroid surgery, and its preoperative assessment is still necessary in these seemingly low-risk operations. Our findings contribute to the expanding literature highlighting the relevance of frailty rather than chronological age in preoperative decision making and preparation. Informed consent should be adjusted based on frailty to ensure that patients have an accurate assessment of their risk when making decisions about whether to undergo surgery. A patient’s age should not be used as a barrier to appropriate surgical management of conditions that affect quality of life and long-term risk of associated complications, such as in hernia repair or parathyroidectomy, or treatment of malignant neoplasms, such as in breast surgery and thyroid surgery. Best practices guidelines68 should be followed to ensure an adequate preoperative assessment of frailty in geriatric patients planned for any inpatient or outpatient operation to appropriately care for and minimize morbidity in this unique patient population.
Corresponding Author: Carolyn D. Seib, MD, MAS, Department of Surgery, University of California, San Francisco, 1600 Divisadero St, Box 1674, San Francisco, CA 94115 (firstname.lastname@example.org).
Accepted for Publication: June 25, 2017.
Published Online: October 11, 2017. doi:10.1001/jamasurg.2017.4007
Author Contributions: Dr Seib had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: Seib, Finlayson.
Drafting of the manuscript: Seib.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Seib.
Administrative, technical, or material support: Seib, Chomsky-Higgins, Suh, Duh, Finlayson.
Study supervision: Suh, Duh, Finlayson.
Conflict of Interest Disclosures: None reported.
Funding/Support: Dr Finlayson is supported by grant R01AG044425 from the National Institute on Aging.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the American College of Surgeons National Surgical Quality Improvement Program are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.
Meeting Presentation: This paper was presented at the 88th Annual Meeting of the Pacific Coast Surgical Association; February 20, 2017; Indian Wells, California.
Create a personal account or sign in to: