Geriatric Vulnerability and the Burden of Disability after Major Surgery

Abstract

Background:

Strong epidemiologic evidence linking indicators of geriatric vulnerability to long-term functional outcomes after major surgery is lacking. The objective of this study was to evaluate the association between geriatric vulnerability and the burden of disability after hospital discharge for major surgery.

Methods:

From a prospective longitudinal study of 754 nondisabled community-living persons, aged 70 years or older, 327 admissions for major surgery were identified from 247 participants who were discharged from the hospital from March 1997 to December 2017. The indicators of geriatric vulnerability were ascertained immediately prior to the major surgery or during the prior comprehensive assessment, which was completed every 18 months. Disability in 13 essential, instrumental and mobility activities was assessed each month.

Results:

The burden of disability over the 6 months after major surgery was considerably greater for non-elective than elective surgery. In multivariable analysis, 10 factors were independently associated with disability burden: age 85 years or older, female sex, Black race or Hispanic ethnicity, neighborhood disadvantage, multimorbidity, frailty, one or more disabilities, low functional self-efficacy, smoking, and obesity. The burden of disability increased with each additional vulnerability factor, with mean values (credible intervals) increasing from 1.6 (1.4–1.9) disabilities for 0–1 vulnerability factors to 6.6 (6.0–7.2) disabilities for 7 or more vulnerability factors. The corresponding values were 1.2 (0.9–1.5) and 5.9 (5.0–6.7) disabilities for elective surgery and 2.6 (2.1–3.1) and 8.2 (7.3–9.2) disabilities for non-elective surgery.

Conclusions:

The burden of disability after hospital discharge for major surgery increases progressively as the number of geriatric vulnerability factors increases. These factors can be used to identify older persons who are particularly susceptible to poor functional outcomes after major surgery, and a subset may be amenable to intervention, including frailty, low functional self-efficacy, smoking, and obesity.

INTRODUCTION

Major surgery is a common event in the lives of older persons, with a nationally-representative incidence (per 100 person-years) of 8.8, including 5.2 for elective and 3.7 for nonelective surgeries.¹ The 5-year cumulative risk of major surgery is 13.8%, representing nearly 5 million older persons in the United States (US).¹ With the projected doubling of persons aged 65 years or older between 2014 and 2060 from 46 to 98 million,² the number of major surgeries in this population will increase substantially.

Geriatricians are increasingly being enlisted to co-manage older patients who are having major elective and non-elective surgery.^{3, 4} In quality improvement initiatives, geriatric co-management has been shown to improve short-term surgical outcomes, including postoperative complications, hospital length of stay, discharge to home, and hospital readmission.⁵ Supportive evidence from clinical trials, however, is relatively sparse. Although not co-management, per se, the Tailored, Family-Involved Hospital Elder Life Program was recently shown in a clustered randomized trial to reduce post-operative delirium among older patients, to shorten hospital length of stay, and to maintain or improve physical and cognitive functions for up to 30 days after hospital discharge.⁶ Whether geriatric co-management or comparable interventions improves longer-term functional outcomes is unknown. One impediment to designing trials to answer this question is the relative paucity of strong epidemiologic evidence linking indicators of geriatric vulnerability to longer-term functional outcomes after major surgery.

The objective of the current study was to evaluate the association between geriatric vulnerability and the burden of disability after hospital discharge for major elective and non-elective surgery. We used high quality data from a unique longitudinal study that includes multiple indicators of geriatric vulnerability, assessed every 18 months, and monthly assessments of functional status among a large cohort of community-living older persons who have been followed for more than 20 years. The results of this study may provide insights on how longer-term functional outcomes after major surgery can be improved in vulnerable older persons.

METHODS

Study Population

Participants were members of an ongoing longitudinal study of 754 community-living persons, aged 70 years or older, who were initially nondisabled in their essential activities of daily living.^{7, 8} Potential participants were members of a large health plan in South Central Connecticut and were excluded for significant cognitive impairment with no available proxy,⁹ life expectancy less than 12 months, plans to move out of the area, or inability to speak English. Only 4.6% of persons refused screening, and 75.2% of those eligible agreed to participate and were enrolled from March 1998 to October 1999. The study was approved by the Yale Human Investigation Committee, and all participants provided informed consent.

Data Collection

Comprehensive assessments were completed by trained nurse researchers at baseline and every 18 months, while telephone interviews were completed monthly by a separate team of researchers through June 2018. For participants who had significant cognitive impairment or were otherwise unavailable, a proxy informant was interviewed using a rigorous protocol.⁹ Deaths were ascertained by review of local obituaries and/or from an informant during a subsequent telephone interview, with a completion rate of 100%. Six hundred eighty-eight (91.3%) participants died after a median of 108 months, while 43 (5.7%) withdrew from the study after a median of 27 months. Data were otherwise available for 94.8% of 4,760 comprehensive assessments and 99.2% of 85,531 monthly interviews. The cohort has been linked to Medicare data.¹⁰

Geriatric Vulnerability

In addition to demographic characteristics, we considered factors from six domains that have been linked to disability in prior studies.¹¹ The socioeconomic factors included years of education, living situation, social support,¹² Medicaid eligibility,¹⁰ and neighborhood disadvantage, as assessed by the area deprivation index (ADI),^{13, 14} which has been updated through 2017.¹⁵ The health-related factors included multimorbidity¹⁶, based on nine self-reported, physician-diagnosed chronic conditions,¹⁷ and frailty.¹⁸ The functional factors included disability in 1 or more of 13 daily activities, as described below, and cognitive impairment.¹⁹ The sensory factors included visual²⁰ and hearing impairment,²¹ each based on objective testing. The psychological factors included depressive symptoms²² and low functional self-efficacy.²³ The behavioral factors included smoking and obesity.²⁴

To enhance clinical interpretability, values for the vulnerability factors were dichotomized using accepted cut-points. Additional operational details are provided in Table 1.

Table 1.

Geriatric Vulnerability Across Multiple Domains Prior to Major Elective and Non-elective Surgery^a

		All Surgeries (N=327)	Elective Surgery (N=197)	Non-elective Surgery (N=130)
Characteristic	Measurement Details
			n (%)
Demographic
Age ≥ 85 years		127 (38.8)	55 (27.9)	72 (55.4)
Female		216 (66.1)	129 (65.5)	87 (66.9)
Black or Hispanic	Self identifiedb	33 (10.1)	26 (13.2)	7 (5.4)
Lives in nursing home		10 (3.1)	2 (1.0)	8 (6.2)
Socioeconomic
Years of education ≤ 12		207 (63.3)	128 (65.0)	79 (60.8)
Lives alone		151 (46.2)	97 (49.2)	54 (41.5)
Low social support	Score on MOS ≤ 18c	72 (22.0)	43 (21.8)	29 (22.3)
Medicaid eligible	Ascertained from Medicare data	23 (7.0)	16 (8.1)	7 (5.4)
Neighborhood disadvantage	Highest quintile of Area Deprivation Index16,17	84 (25.7)	60 (30.5)	24 (18.5)
Health related
Multimorbidity	> 2 chronic conditionsd	119 (36.4)	72 (36.2)	47 (36.2)
Frailty	≥ 3 criteria from Fried phenotypee	106 (32.4)	49 (24.9)	57 (43.9)
Functional
One or more disabilities	From 13 daily activitiesf	248 (75.8)	142 (72.1)	106 (81.5)
Cognitive impairment	Score on Folstein MMSE < 24	42 (12.8)	14 (7.1)	28 (21.5)
Sensory
Visual impairment	> 26%, assessed with a Jaeger card	70 (21.4)	28 (14.2)	42 (32.3)
Hearing impairment	4 tones missed out of 4g	89 (27.2)	42 (21.3)	47 (36.2)
Psychological
Depressive symptoms	Score on CES-D ≥ 20	53 (16.2)	36 (18.3)	17 (13.1)
Low functional self-efficacy	Score ≤ 27h	145 (44.3)	76 (38.6)	69 (53.1)
Behavioral
Current or former smoker		205 (62.7)	125 (63.5)	80 (61.5)
Obesity	Body mass index ≥ 30 kg/m2j	72 (22.0)	46 (23.4)	26 (20.0)

Abbreviations: CES-D, Center for Epidemiological Studies-Depression Scale; MMSE, Mini-Mental State Examination; MOS, Medical Outcomes Study Social Support Scale.

^aCharacteristics were assessed at the beginning of each 18-month person interval, with the exception of lives in nursing home, Medicaid eligible, and disability, which were ascertained immediately prior to the hospitalization for major surgery. The 327 observations were contributed by 247 participants.

^bParticipants were asked by a trained nurse researcher to identify their race/ethnicity. These data were collected primarily for descriptive purposes and to fulfill federal regulations regarding the inclusion of minority participants in studies funded by the National Institutes of Health.

^cCut-point demarcates the worst quartile, based on the first 356 enrolled participants who had been selected randomly from the source population.

^dChronic conditions included hypertension, myocardial infarction, heart failure, stroke, diabetes mellitus, arthritis, hip fracture, chronic lung disease, and cancer (other than minor skin cancers); the cut-point was selected based on the American Geriatrics Society Expert Panel on the Care of Older Adults with Multimorbidity.¹⁶

^eBased on the five standard criteria: weight loss, exhaustion, low physical activity, muscle weakness, and slow walking speed.¹⁸

^fIncludes four essential activities (bathing, dressing, walking, and transferring), five instrumental activities (shopping, housework, meal preparation, taking medications, and managing finances), and four mobility activities (walk ¼ mile, climb flight of stairs, lift/carry ten pounds, and driving), as described in the Methods.

^gBased on 1000 and 2000 HZ measurements for the left and right ears.

^hMaximum score is 40, based on level of confidence in performing 10 activities (each scored 0 to 4): dressing, bathing/showering, transferring, going up and down stairs, walking around the neighborhood, house cleaning, preparing simple meals, simple shopping, reaching into cabinets or closets, and hurrying to answer the telephone. Cut-point demarcates the worst quartile, based on the first 356 enrolled participants who had been selected randomly from the source population.

ⁱInformation was not collected on total pack-years of smoking.

^jBased on established cut-point.⁴⁰

Disability Assessments

Complete details regarding the assessment of disability are provided elsewhere.^{8, 9} Each month, participants were asked, “At the present time, do you need help from another person to (complete the task)?” for each of four essential activities (bathing, dressing, walking, and transferring), five instrumental activities (shopping, housework, meal preparation, taking medications, and managing finances), and three mobility activities (walk ¼ mile, climb flight of stairs, and lift/carry ten pounds). For these 12 activities, disability was operationalized as the need for personal assistance. Participants were also asked about a fourth mobility activity, “Have you driven a car during the past month?” Participants who responded “No” were considered to be “disabled” in driving.²⁵ To address the small amount of missing data on disability (1% of observations), multiple imputation was used with 100 random draws per missing observation.²⁶

Ascertainment of Major Surgeries

As previously described,²⁷ Centers for Medicare and Medicaid Services (CMS) records and monthly interview data on self-reported surgeries, verified by chart review, were used to identify participants who had undergone an operation. Major surgery was defined as any procedure in an operating room requiring the use of general anesthesia for a non-percutaneous, non-endoscopic, invasive operation (open or laparoscopic) during a hospital admission. This is consistent with other definitions of high-risk major surgery in older persons.²⁸ We categorized each major surgery as musculoskeletal, abdominal, vascular, cardiothoracic, neurologic, or other (including thyroidectomies, major breast operations, extensive lymph node dissections, burn debridements, and skin grafts).²⁷ Major surgeries identified from CMS records were categorized as elective or non-elective by an indicator variable; non-elective surgeries included both urgent and emergent operations.²⁹ Major surgeries identified by self-report and chart review were categorized as elective or non-elective based on the history in the chart; any admission for major surgery originating from the emergency department was categorized as non-elective, as were unscheduled operations due to a time-sensitive condition. For each admission, length of hospital stay was determined.

Assembly of analytic sample

Major surgeries were included through December 2017. Participants could contribute more than one major surgery to the analysis as long as each represented the first hospital admission within an 18-month interval, since prior hospitalizations may have altered key participant characteristics from the preceding comprehensive assessment. Of the 579 major surgeries, 348 represented the first hospitalization within an 18-month interval: 278 (79.9%) from Medicare claims and 70 (21.1%) from self-report and chart review. Twenty-one observations were excluded—8 were already disabled in all 13 activities and 13 died in the hospital or prior to the first post-hospital follow-up interview, leaving 327 major surgeries from 247 participants.

Statistical Analysis

The primary analyses included all major surgeries, although results are also provided separately for elective and non-elective surgeries. Descriptive statistics were calculated for the geriatric vulnerability factors by domain. These data were obtained from the comprehensive assessment that preceded the major surgery except for age, living in nursing home, Medicaid eligibility, and disability, which were based on participant status at the time of the monthly interview immediately prior to the major surgery. The adjusted least square means and standard errors were calculated for the number of disabilities (out of 13) for each of the six months after hospital discharge for major surgery, with the unadjusted values from the month prior to hospitalization serving as the reference point.

Bayesian negative binomial random effects models were used to determine the bivariate and multivariable associations between each of the candidate vulnerability factors and the burden of disability, defined as the number of disabilities for each of the six months after hospital discharge for major surgery. These models generate rate ratios that represent the proportional change in disability burden for each of the vulnerability factors, with statistical significance denoted by the highest posterior density credible interval exclusive of 1.0. All coefficients and random effects had highly dispersed, non-informative normal prior distributions. Person-specific random intercepts were included to account for multiple observations per participants. As in prior studies,^{30, 31} the multivariable models were adjusted for type of surgery, non-elective surgery, length of hospital stay, number of months to major surgery from start of 18-month interval, and number of the specific 18-month interval. Formal tests of statistical interaction were run to determine whether the associations differed between elective and non-elective surgeries.

In a final set of analyses, the adjusted least square means for the number of disabilities, accompanied by credible intervals, were calculated according to the number of statistically significant vulnerability factors from the multivariable models; and the presence of a linear association between the number of vulnerability factors and burden of disability was evaluated.

All analyses were conducted using SAS version 9.4 (SAS Institute, Inc.; Cary, NC).

RESULTS

The participant characteristics for the 327 major surgeries are provided in Table 1, organized by domain of geriatric vulnerability. Overall, nearly 40% were 85 years or older, two-thirds were female, 1 out of 10 were Black or Hispanic, and only a small number were living in a nursing home. Participants who had elective surgery were generally less vulnerable than those who had non-elective surgery, as evidenced by their younger age and lower prevalence of living in a nursing home, frailty, disability, impairments in cognition, vision and hearing, and low functional self-efficacy. However, participants who had elective surgery were more likely to be Black or Hispanic, to live alone or in a disadvantaged neighborhood, and to have depressive symptoms. The surgical characteristics are provided in Table S1. The three most common types of surgery, which differed considerably between elective and non-elective procedures, were musculoskeletal, abdominal and vascular. Hospital length of stay was nearly twice as long for the non-elective versus elective surgeries. A complete list of the 327 major surgeries by operation type is provided in Table S2.

Figure 1 shows the course of disability over 6 months after hospital discharge for major surgery. At each time point, the number of disabilities was greater for non-elective than elective surgeries. In addition, while the number of disabilities at month 6 was comparable to that at month 0 (i.e. pre-hospitalization) for elective surgeries, it was considerably greater than that at month 0 for non-elective surgeries. The bivariate and multivariable associations between the geriatric vulnerability factors and the subsequent burden of disability are provided in Table 2. In the bivariate analysis, statistically significant associations were observed for many of the vulnerability factors, with rate ratios ranging from 1.12 for neighborhood disadvantage to 2.84 for one or more disabilities. The factors that remained statistically significant in the multivariable analysis included age 85 years or older, female sex, Black race or Hispanic ethnicity, neighborhood disadvantage, multimorbidity, frailty, one or more disabilities, low functional self-efficacy, smoking, and obesity. Considering the diminished power, the multivariable results for elective and non-elective surgery were generally comparable, with a few exceptions, as shown in Table 3. Of ten independent risk factors from the overall model, two—female sex and multimorbidity—were independently associated with the subsequent burden of disability for elective, but not for non-elective surgery; and formal tests of interaction were statistically significant for both factors. New statistically significant associations were observed for Medicaid eligible, but only for non-elective surgery, and cognitive impairment, but only for elective surgery; however, formal tests of interaction were statistically significant only for cognitive impairment.

Figure 1.

Course of Disability over 6 Months after Hospital Discharge for Major Surgery. Values (ranging from 0 to 13) represent the adjusted least-square mean number of disabilities, accompanied by standard errors. As a reference point, the unadjusted pre-hospitalization disability counts are included at month 0. The 327 observations were contributed by 247 participants. The number of observations is provided below the figure at each month of follow-up. There was no attrition for reasons other than death.

Table 2.

Associations Between Geriatric Vulnerability Factors and Burden of Disability over 6 Months after Hospital Discharge for Major Surgery (N=327)^a

	Bivariate	Multivariable
Characteristic	Rate Ratio (credible interval)
Demographic
Age ≥ 85 years	1.45 (1.33, 1.58)	1.32 (1.22, 1.43)
Female	1.40 (1.27, 1.54)	1.15 (1.06, 1.25)
Black or Hispanic	1.14 (0.98, 1.32)	1.22 (1.08, 1.39)
Socioeconomic
Years of education ≤ 12	1.02 (0.93, 1.11)	0.97 (0.90, 1.05)
Lives alone	1.02 (0.93, 1.12)	0.99 (0.92, 1.07)
Low social support	1.02 (0.92, 1.14)	0.97 (0.89, 1.07)
Medicaid eligible	1.43 (1.20, 1.68)	1.03 (0.91, 1.18)
Neighborhood disadvantage	1.12 (1.01, 1.23)	1.15 (1.05, 1.25)
Health related
Multimorbidity	1.32 (1.20, 1.44)	1.09 (1.01, 1.17)
Frailty	2.19 (2.00, 2.39)	1.19 (1.09, 1.29)
Functional
One or more disabilities	2.84 (2.54, 3.14)	1.58 (1.45, 1.74)
Cognitive impairment	1.82 (1.60, 2.06)	1.06 (0.96, 1.19)
Sensory
Visual impairment	1.61 (1.45, 1.79)	1.05 (0.96, 1.14)
Hearing impairment	1.34 (1.21, 1.48)	1.02 (0.94, 1.11)
Psychological
Depressive symptoms	1.37 (1.22, 1.54)	1.03 (0.94, 1.13)
Low functional self-efficacy	2.18 (2.00, 2.37)	1.47 (1.35, 1.60)
Behavioral
Current or former smoker	0.87 (0.80, 0.96)	1.11 (1.04, 1.21)
Obesity	1.25 (1.13, 1.39)	1.22 (1.12, 1.32)

^aThe 327 observations were contributed by 247 participants. The multivariable models were adjusted for type of surgery, non-elective surgery (yes/no), length of hospital stay, number of months to major surgery from start of 18-month interval, and number of the specific 18-month interval. The prevalence of lives in nursing home was too low to evaluate as a vulnerability factor.

Table 3.

Multivariable Associations Between Geriatric Vulnerability Factors and Burden of Disability over 6 Months after Hospital Discharge for Major Elective and Non-elective Surgery^a

	Elective	Non-elective
Characteristic	Rate Ratio (credible interval)
Demographic
Age ≥ 85 years	1.15 (1.01, 1.33)	1.19 (1.09, 1.31)
Femaleb	1.23 (1.08, 1.41)	1.02 (0.93, 1.12)
Black or Hispanic	1.15 (0.96, 1.39)	1.19 (1.01, 1.43)
Socioeconomic
Years of education ≤ 12	0.96 (0.86, 1.09)	0.91 (0.84, 0.99)
Lives alone	1.02 (0.90, 1.15)	0.89 (0.82, 0.98)
Low social support	0.95 (0.83, 1.09)	0.95 (0.86, 1.04)
Medicaid eligible	1.01 (0.84, 1.23)	1.21 (1.03, 1.44)
Neighborhood disadvantage	1.15 (0.99, 1.30)	1.09 (0.97, 1.22)
Health related
Multimorbidityb	1.17 (1.04, 1.32)	0.97 (0.89, 1.05)
Frailty	1.19 (1.04, 1.37)	1.36 (1.23, 1.49)
Functional
One or more disabilities	2.06 (1.76, 2.37)	1.48 (1.31, 1.68)
Cognitive impairmentb	1.48 (1.19, 1.86)	1.09 (0.99, 1.20)
Sensory
Visual impairment	1.03 (0.87, 1.20)	1.03 (0.94, 1.11)
Hearing impairment	1.13 (0.98, 1.30)	0.98 (0.90, 1.07)
Psychological
Depressive symptoms	0.98 (0.84, 1.14)	1.06 (0.95, 1.18)
Low functional self-efficacy	1.42 (1.24, 1.64)	1.31 (1.19, 1.44)
Behavioral
Current or former smoker	1.10 (0.98, 1.25)	1.08 (1.00, 1.19)
Obesity	1.11 (0.97, 1.26)	1.08 (0.98, 1.20)

^aThe 197 observations for elective major surgery were contributed by 158 participants, while the 130 observations for elective major surgery were contributed by 116 participants. The multivariable models were adjusted for type of surgery, length of hospital stay, number of months to major surgery from start of 18-month interval, and number of the specific 18-month interval.

^bFormal tests of interaction between elective and non-elective surgery were statistically significant.

As shown in Figure 2, the burden of disability in the six months after major surgery increased with each additional vulnerability factor that was statistically significant in the overall model (i.e., “All”), with mean values increasing from 1.6 disabilities for 0–1 vulnerability factors to 6.6 disabilities for 7 or more vulnerability factors. The increase was greatest, however, between 3 (2.7 disabilities) and 4 (5.1 disabilities) vulnerability factors, with corresponding credible intervals that did not overlap. Comparable findings were observed for elective and non-elective surgeries with only one exception. For elective surgeries, the number of disabilities was 5.4 for 5 vulnerability factors and 5.0 for 6 vulnerability factors, although the corresponding credible intervals overlapped. Tests of linear association between the number of vulnerability factors and burden of disability were highly significant, as denoted by adjusted rate ratios of 1.26 (credible interval: 1.24–1.29) for all major surgeries, 1.32 (1.28–1.36) for elective surgeries, and 1.20 (1.17–1.23) for non-elective surgeries.

Figure 2.

Burden of Disability over 6 Months after Hospital Discharge for Major Surgery According to the Number of Vulnerability Factors. The 10 statistically significant factors from the multivariable model in Table 2 were included. Values represent the adjusted least square means for the number of disabilities, accompanied by credible intervals. The multivariable models were adjusted for type of surgery, non-elective surgery, length of hospital stay, number of months to major surgery from start of 18-month interval, and number of the specific 18-month interval.

DISCUSSION

In this prospective longitudinal study of community-living older persons, we identified ten indicators of vulnerability that were independently associated with the burden of disability over 6 months after hospital discharge for major surgery and demonstrated that the burden of disability increases as the number of vulnerability factors increases. At least one vulnerability factor was identified from each of the seven domains with the exception of Sensory, suggesting that disability after major surgery is a multifactorial process and that interventions to promote functional independence after major surgery will likely require a multifaceted, multidisciplinary approach.

One of the most feared complications of major surgery is the loss of independence. Prior research has shown that major surgery in older persons is a potent precipitant of functional decline and disability.^{32, 33} In the setting of major surgery, the level of disability in daily activities increases in about three-quarter of the cases,³² and the likelihood of developing severe disability, defined as the need for personal assistance with three or more essential activities, is increased more than 600-fold among previously nondisabled older persons.³³

The factors identified in the current study can be used to identify older persons who are particularly susceptible to poor functional outcomes after major surgery and a subset of these factors may be amenable to intervention, including frailty, low functional self-efficacy, smoking, and obesity. The burden of disability after major surgery was especially high among persons who had 4 or more of the vulnerability factors, representing more than 60% (210/327) of the overall sample. These findings were observed for elective and non-elective surgeries, although the course of disability was worse for the latter group. The modest differences in the specific vulnerability factors identified in the elective and non-elective surgery groups must be interpreted carefully given the large number of statistical comparisons without correction for multiplicity.

Our findings may inform co-management models for geriatric surgery. The Society for Perioperative Assessment and Quality Improvement, which includes a multidisciplinary group of surgeons, anesthesiologists, nurses and other clinicians, has recently provided recommendations for launching a geriatric surgery center, with a primary focus on elective surgeries.³⁴ A key element includes preoperative screening for frailty and cognitive impairment, two factors that were independently associated with the burden of disability after major elective surgery in the current study. The presence of either of these vulnerability factors would trigger referral to a geriatrician for a more comprehensive geriatric assessment, focusing on many of the other vulnerability factors evaluated in the current study.

In parallel, the American College of Surgeons has launched the Geriatric Surgery Verification (GSV) Program, which has developed a core group of 30 standards to address the most important aspects of geriatric surgical care within hospitals, with the goal of improving outcomes among patients 75 years or older.³⁵ Recent evidence indicates that implementation of the GSV Program standards, including assessment of geriatric-specific vulnerabilities and interdisciplinary care, reduces length of stay in older persons undergoing major elective and non-elective surgeries,³⁶ although data on longer-term outcomes are not currently available.

The results from the current study provide strong epidemiologic evidence to support recommendations from the Society for Perioperative Assessment and Quality Improvement and the GSV Program. Nonetheless, clinical trials will be needed to determine whether geriatric co-management or comparable interventions improves longer-term functional outcomes after major elective and non-elective surgery. Our results can be used to inform the design of such trials by helping to identify older persons who are particularly susceptible to poor functional outcomes after major surgery and to prioritize target areas for multifaceted interventions to improve these outcomes after major surgery.

A unique feature of our study is the assembly of a large sample of major surgeries from an ongoing longitudinal study, thereby reducing sampling bias³⁷ and permitting assessment of a comprehensive set of premorbid indicators of geriatric vulnerability, including among persons having non-elective operations. Additional strengths include assessment of a comprehensive set of essential, instrumental and mobility activities from monthly interviews, which allowed us to more precisely quantify the burden of disability after hospital discharge for major surgery; reassessment of the vulnerability factors at 18-month intervals; and complete ascertainment of a diverse mix of major surgeries, both elective and nonelective, using a standard definition²⁸ and several different sources of information, including self-report, medical records and CMS files. In contrast, most prior studies have focused on a limited set of elective surgeries.³⁸

Our findings should be interpreted in the context of several potential limitations. First, detailed information was not available on the index hospitalization for major surgery, including acuity of illness, surgical complexity, surgeon experience, and postoperative complications. To partially address this issue, the multivariable models included length of hospital stay. Second, the vulnerability factors identified in the current study could operate differently for different surgeries, even those within specific surgery types, i.e., musculoskeletal (rotator cuff repair vs. closed reduction of femur fracture with internal fixation), vascular (carotid endarterectomy vs. amputation below knee), etc. Because of relatively small samples sizes, surgery-specific analyses are not feasible for prospective longitudinal studies such as ours that include detailed information on vulnerability factors and functional outcomes. Third, we did not consider factors that occurred after the major surgery, such as rehabilitation or receipt of adjuvant chemotherapy or radiation for cancer. Fourth, because this was an observational study, the reported associations cannot be construed as causal. Even if the associations were causal, whether the burden of disability after major surgery could be reduced through currently available interventions is uncertain. Fifth, because our participants were members of a single health plan in South Central Connecticut, our findings may not be generalizable to older persons in other settings. However, the demographic characteristics of our cohort reflect those of older persons in New Haven County, Connecticut, and are similar to the characteristics of the U.S. population, except in terms of race or ethnicity.²⁵ Furthermore, generalizability depends not only on the choice of the study sample but also on the stability of the sample over time.³⁹ One of the great strengths of our study is the low attrition rate. The generalizability of our findings is also enhanced by our high participation rate, which was greater than 75%.

In summary, the burden of disability after hospital discharge for major surgery increases progressively as the number of geriatric vulnerability factors increases. These factors can be used to identify older persons who are particularly susceptible to poor functional outcomes after major elective and non-elective surgery and a subset of these factors, including frailty, low functional self-efficacy, smoking, and obesity, may be amenable to intervention. These findings should assist geriatricians and other geriatric providers to work more effectively with surgical teams, including surgeons, anesthesiologists and nurses, to improve longer-term functional outcomes after major elective and non-elective surgery, particularly in vulnerable older persons; and they can also inform the design of clinical trials to rigorously evaluate whether geriatric co-management or comparable interventions improves functional outcomes after major surgery.

Supplementary Material

tS1

Supplementary Table S1. Surgical Characteristics According to Major Elective and Non-elective Surgery

tS2

Supplementary Table S2. Complete List of 327 Major Surgeries by Operation Type

Key Points.

Ten indicators of vulnerability were independently associated with disability burden over 6 months after major surgery.

These indicators can be used to identify older persons who are particularly susceptible to poor functional outcomes, and a subset, including frailty, low functional self-efficacy, smoking and obesity, may be amenable to intervention.

Why does this matter

By informing co-management models of geriatric surgery, our findings should assist geriatricians to work more effectively with surgical teams to improve long-term functional outcomes, particularly in vulnerable older persons.