Factors Associated With Functional Recovery Among Older Survivors of Major Surgery

Abstract

Objective

The objectives of the current study were 2-fold: first, to evaluate the incidence and time to recovery of premorbid function within 6 months of major surgery and second, to identify factors associated with functional recovery among older persons who survive a major surgery with increased disability.

Background

Most older persons would not choose a surgical treatment resulting in persistently increased postsurgical disability, even if survival was assured. Potential predictors of functional recovery after major surgery have, however, not been well-studied among geriatric patients.

Methods

It is a prospective longitudinal study of 754 community-living persons 70 years or older. The analytic sample included 266 person-admissions in which participants survived major surgery with increased disability and were monitored on a monthly basis for 6 months.

Results

Of the 266 person-admissions assessed, 174 (65.4%) recovered to their presurgical level of function, with median time to recovery of 2 months (interquartile range, 1–3), whereas 16 (6.0%) died. Two factors were significantly associated with an increased likelihood of functional recovery: being nonfrail (hazard ratio 1.60; 95% confidence interval 1.03–2.51; P = 0.038) and having elective surgery (hazard ratio 1.72; 95% confidence interval 1.14–2.59; P = 0.009). Three factors were associated with a reduced likelihood of functional recovery: hearing impairment, greater increase in postsurgical disability in the month after hospital discharge, and years of education.

Conclusions

Among older persons, nonfrailty and elective surgery were positively associated with functional recovery, whereas hearing impairment, greater increases in postsurgical disability, and years of education were associated with higher risk of protracted disability.

As our population in the United States ages, an increasing percentage of major operations are being performed on patients 65 years and older. From a healthcare perspective, the standard metric for measuring successful surgical outcomes in this older population is an optimal balance between survival and the minimization of complications. From the patient perspective, however, emphasis is on quality of life, as more than 70% of older persons would not choose a treatment that causes severe functional disability, even if survival was assured.^1,2 To provide optimal, patient-centered perioperative care to older persons, surgeons need to understand how major surgery can affect functional status and lead to protracted disability.

Although interventions to improve functional recovery in nonsurgical patients have been well-studied,^3–5 there are few evidence-based tools specific to geriatric surgical patients. Current guidelines are largely based on medical patients.⁶ The retrospective data that exist may lead to erroneous conclusions about the association of major surgery with functional recovery.^7,8 Evidence from prospective cohort studies has been drawn exclusively from studies of elective surgical patients, within individual subspecialties, and with variable methodologies, resulting in only moderate generalizability.^9–12

To date, no previous study of the geriatric population has evaluated predictors of functional recovery after major surgery that includes both elective and nonelective operations and that spans the spectrum of surgical disciplines. Given the importance of functional outcomes after major surgery, a careful evaluation of potential predictors of functional recovery may clarify and identify targets for intervention aimed at preserving function. This is important, as it has been shown that most older survivors of major surgery are able to achieve some degree of functional recovery, even as their trajectory of recovery can vary considerably.¹³

The objectives of the current study were 2-fold: first, to evaluate the incidence and time to recovery of premorbid function within 6 months of major surgery and second, to identify factors associated with functional recovery among older persons who survive a major surgery with increased disability. To accomplish our objectives, we used data from a unique prospective, longitudinal study that has followed a large cohort of older persons for more than 18 years with monthly assessments of functional status and repeated assessments of patient-related and geriatric factors.¹⁴

METHODS

Study Population

Participants were drawn from the Precipitating Events Project, an ongoing longitudinal study of 754 community-dwelling persons 70 years or older, who were initially nondisabled in 4 basic activities of daily living: bathing, dressing, walking, and transferring.^14,15 Potential participants were members of a large health plan in greater New Haven, Connecticut and were enrolled from March 1998 through October 1999. Exclusion criteria included significant cognitive impairment with no available proxy, life expectancy less than 12 months, plans to leave the area, and inability to speak English. Only 4.6% of persons contacted refused screening, and 75.2% of those eligible consented to participate; those who declined did not differ significantly from those who enrolled.^14,15 The Yale Human Investigation Committee approved the study.

Data Collection

Comprehensive home-based assessments were completed at baseline and at 18-month intervals for 198 months (except at 126 months), through December 2015. Telephone interviews were completed monthly through December 2016. For participants who had significant cognitive impairment or were unavailable, a proxy informant was interviewed using a rigorous protocol, with demonstrated reliability and validity.¹⁶ Deaths were ascertained by review of obituaries and/or from an informant during a telephone interview, with a completion rate of 100%. A total of 669 (88.7%) participants died after a median follow-up of 100 months, and 43 (5.7%) dropped out of the study after a median follow-up of 27 months. Data were otherwise available for 99.2% of 84,250 monthly interviews.

Potential Patient-related and Geriatric Predictors

Using data from the comprehensive assessments, we considered potential predictors that have been linked to functional recovery in prior studies,^17–20 and additional factors related to hospitalizations for major surgery.

• Five demographic characteristics: age; sex; race/ethnicity; education; and living alone.

• Thirteen health-related characteristics: 9 self-reported physician-diagnosed chronic conditions, including hypertension; myocardial infarction; congestive heart failure; stroke; cancer, excluding minor skin cancers; diabetes; fractures; arthritis; and chronic lung disease; and 4 other conditions, including body mass index, frailty, hearing, and vision. Frailty was assessed using the Fried Frailty Index: a score of 0 to 2 (out of 5) was considered nonfrail.²¹ Hearing was assessed using a handheld AudioScope, with severe impairment defined as 4 out of 4 tones missed, based on 1000- and 2000-Hz measurements for the left and right ears.²² Visual acuity was assessed with a Jaeger card, with the participant wearing their own reading glasses when applicable; vision impairment was defined as inability to read higher than or equal to 6% of the eye chart (moderate visual impairment being 6%–26%, and severe visual impairment being >26%).²³

• Four cognitive-psychosocial parameters: cognitive status, depressive symptoms, functional self-efficacy, and social support. Cognition was assessed using Mini-Mental State Examination score as a continuous variable.²⁴ Depressive symptoms were assessed using Center for Epidemiological Studies Depression scale as a continuous variable.²⁵ Functional self-efficacy, a measure of confidence in performing various activities, was assessed as a continuous variable using the Modified Self-Efficacy Scale.¹⁸ Participants were asked, “how confident/sure are you that you can _______ [ask activity]?” for each of 10 activities, which was scored as: 0=not at all; 1=a little; 2=fairly; 3=very; and 4=completely. Examples of activities include “get dressed or undressed” or “get in and out of a chair.” Social support was assessed with a modified version of the Medical Outcomes Survey Social Support Scale.²⁶

• Two functional markers: physical capacity and disability at 1-month postoperatively. Physical capacity was evaluated with a modified version of the Short Physical Performance Battery²⁷ that included the standard balance maneuvers but substituted 3 timed chair stands (instead of 5) and timed rapid gait (back and forth over a 10-foot course) instead of timed usual gait (over a 4-m course without a turn).¹⁷ It was included as a continuous variable, with scores ranging from 0 to 12 (higher scores mean better physical performance). Increase in disability was calculated as the disability score (on a scale from 0 to 13, with higher scores denoting greater disability) in the month after surgery minus the disability score in the month before surgery (additional details in “Assessment of functional status” section).

• Four variables related to major surgery: an indicator of elective versus nonelective surgery; hospital length of stay in days; calendar year of admission; and type of surgery. Because nearly half of the analytic sample included musculoskeletal operations, type of surgery was defined as musculoskeletal versus nonmusculoskeletal (aditional details in “Ascertainment of major surgery” section).

Assessment of Functional Status

The monthly assessment of functional status included 13 activities. Participants were asked, “At the present time, do you need help from another person to ________ [ask activity]?” for each of 4 basic activities (bathing, dressing, walking, transferring), 5 instrumental activities (shopping, housework, meal preparation, taking medications, managing finances), and 3 mobility activities (walk ¼ mile, climb flight of stairs, lift or carry 10 pounds). Disability was operationalized as the need for personal assistance or inability to perform the task. Participants were also asked about a fourth mobility activity: “Have you driven a car during the past month?” To maintain consistency with the other activities, participants who responded “no” were classified as being “disabled” in driving.¹⁹ To address the small amount (0.9%) of missing disability data, we used multiple imputation with 100 random draws per missing observation.²⁸ Details regarding the assessment of disability are provided elsewhere.^16,19,29

Ascertainment of Major Surgery

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 any major surgery. Major surgery was defined as any procedure in an operating room requiring the use of general anesthesia for a nonpercutaneous, nonendoscopic, invasive operation (either open or laparoscopic). This definition has been previously implemented by our group in the Annals of Surgery¹³ and is consistent with other definitions of high-risk major surgery in older persons.³⁰ Please see supplementary online material, Appendix, http://links.lww.com/SLA/B589, for full list of operations included.

We categorized each major surgery into 1 of 6 categories:

1. Neurologic (for brain and spine surgeries)

2. Cardiothoracic

3. Abdominal-gastrointestinal

4. Musculoskeletal

5. Vascular (for endovascular surgeries, noncoronary bypass grafts, and amputations)

6. Other (including thyroidectomies, major breast operations, extensive lymph node excisions, burn debridements, and skin grafts).

Major surgeries identified from CMS records were categorized as elective or nonelective by a CMS indicator variable; nonelective surgeries included both urgent and emergent operations.³¹ Major surgeries identified by self-report and chart review were categorized as elective or nonelective based on the history in the chart: any admission for major surgery originating from the emergency department was categorized as nonelective, as were unscheduled operations due to a time-sensitive condition.

Assembly of Analytic Sample

Major surgeries were included through December 2015. Participants could contribute more than one major surgery over the course of the study, but only the first major surgery associated with a hospital admission within an 18-month interval was eligible, because most of the potential predictors were measured during the comprehensive assessments. To permit an evaluation of recovery, major surgical admissions were only included when the first postsurgical count of disability was greater than the presurgical count of disability (additional details in “Statistical Analysis” section).

The assembly of the analytic sample is depicted in Figure 1. A total of 369 person-admissions occurring first within an 18-month interval were identified: 297 (80.5%) via Medicare claims and 72 (19.5%) via self-reported admissions for surgery. Of the 369, 103 were subsequently excluded due to not meeting our inclusion/exclusion criteria, including 84 which did not have increased disability at the first postsurgical assessment. The final sample included 266 person-admissions contributed by 216 participants.

FIGURE 1.

Assembly of analytic sample from the parent cohort. Of the 754 Precipitating Events Project (PEP) participants, 569 major surgeries were identified from 3070 CMS Hospital stays and 148 self-reported admissions for surgery. Of the 569 major surgeries, 369 represented the first hospitalization within an 18-month interval. Of these, 15 surgeries were excluded because the participants had withdrawn from the study and 4 were excluded as they occurred within 6 months of prior major operation, leaving 350 major surgery admissions. Of these, 84 were excluded because they did not result in increased disability in the month after surgery. The final sample included 266 major surgeries, contributed by 216 participants. CMS, Centers for Medicare and Medicaid Services. FFS, fee for service.

Statistical Analysis

The unit of analysis was the first person-admission to a hospital for major surgery during an 18-month interval resulting in increased disability. Functional recovery was defined as the return, within 6 months of the first postsurgery interview, to a total disability count less than or equal to that from the month immediately preceding the surgical admission. The 6-month follow-up period was chosen in accordance with previous publications which have documented that most participants recover function within that time frame.^17,20,32

To build the multivariable model for recovery of preadmission function, 20 potential predictors, chosen a priori based on clinical relevance, were tested for their bivariate associations with recovery (Table 1). The bivariate associations were calculated from a proportional hazards model employing generalized estimating equations with a compound symmetry correlation structure to account for the multiple events contributed by some participants.³³ All multivariable models included 4 variables: age, sex, race, and increase in disability count. Because the number of recovery events justified the inclusion of 17 explanatory variables,³⁴ backwards selection was used to select the other 13 explanatory variables that were retained in the final multivariable model.

TABLE 1.

Characteristics and Bivariate Associations of Potential Predictors of Functional Recovery Within 6 Months of First Post-Major Surgery Assessment in Older Adults (N = 266 Person-admissions)

Potential Predictors*	Operational Details*	n (%) or Mean (SD)	HR (95% CI)	P †
Demographic
Age	In years, at time of major surgery	82.2 (6.1)	0.97 (0.94–0.99)	0.008
Female sex		181 (68%)	0.77 (0.56–1.05)	0.097
Non-Hispanic white	Self-identified as white and non-Hispanic	241 (91%)	1.17 (0.69–2.00)	0.559
Education count	In years	12.3 (2.9)	1.00 (0.95–1.05)	0.099
Lives alone		120 (45%)	0.83 (0.62–1.12)	0.233
Health related
Number chronic conditions	Nine self-reported, physician-diagnosed conditions	2.2 (1.3)	0.92 (0.82–1.04)	0.171
BMI, per unit increase	Self-reported height and weight, kg/m2	26.5 (5.6)	0.99 (0.97–1.02)	0.527
Nonfrail	Nonfrail score 0–2 (out of 5) based on Fried Frailty Index.	187 (67%)	1.96 (1.25–3.03)	0.003
Hearing impairment, severe	Four tones missing out of 4; assessed with AudioScope	35 (13%)	0.90 (0.81–0.99)	0.025
Vision impairment	Impairment of ≥6%, based on reading Jaeger eye chart (with use of corrective lenses, if applicable)	162 (61%)	1.45 (1.07–1.98)	0.018
Cognitive-psychosocial
Cognitive status	Scored on Folstein Mini-Mental Status Examination	26.5 (3.2)	1.05 (0.99–1.11)	0.056
Depressive symptoms	Scored on Center for Epidemiological Studies Depression scale	10.5 (9.3)	0.98 (0.97–1.00)	0.084
Functional self-efficacy	Scored on Modified Self-Efficacy Scale	27.5 (8.6)	1.04 (1.02–1.06)	<0.001
Social support	Scored on a modified version of the Medical Outcomes Survey Social Support Scale	22.2 (5.5)	1.01 (0.99–1.04)	0.369
Physical ability
Physical capacity	Scored on a modified version of the Short Physical Performance Battery; range from 0–12.	6.0 (2.9)	1.12 (1.06–1.18)	<0.001
Increase in disability‡	The amount disability increased above baseline, defined as the difference between the first postsurgery disability count and the last presurgery disability count	5.8 (3.2)	0.89 (0.84–0.94)	<0.001
Surgery characteristics
Elective operation	Scheduled operation (as opposed to nonelective urgent/emergent operation)	150 (56%)	2.32 (1.69–3.17)	<0.001
Length of stay	In days, hospital length of stay for major surgery	7.1 (7.2)	0.98 (0.96–1.01)	0.172
Admission year	Year in which hospitalization for major surgery occurred	2004 (4)	0.98 (0.95–1.02)	0.323
Type of surgery, MSK	Yes for musculoskeletal surgery	131 (49%)	0.65 (0.48–0.87)	0.005

^*All potential predictors are presented. Please see Methods section in the text for full details of the variables.

^†P < 0.05 is considered statistically significant from proportional hazards model with compound symmetric correlation structure.

^‡Increase in disability was a continuous variable, defined as the difference between the first postsurgery disability count and the last presurgery disability count. Disability was assessed monthly, in each of 13 functional activities, with a possible score of 0 to 13. Please see text for details.

BMI indicates body mass index; CI, confidence interval; MSK, musculoskeletal operations; n, number; %, percentage; SD, standard deviation.

For each explanatory variable in the model, the proportional hazards assumption was checked with cumulative sums of martingale residuals.³⁵ After final selection of the explanatory variables, the multivariable model was re-run with the Fine and Gray approach to evaluate the hazards of subdistribution, which address the competing risk of death.³⁶ The result for each explanatory variable is reported as a hazard ratio (HR), which can be thought of as the average rate of the event per unit of time (day) among those with the variable, referenced against the average rate of those without that variable.

All analyses were performed using SAS version 9.4 (SAS Institute, Inc, Cary, NC), where P less than 0.05 (2-tailed) denoted statistical significance.

RESULTS

Among the 266 major surgeries contributed by participants who survived to their first posthospitalization functional assessment with increased disability, 181 (68.0%) were among women, 241 (90.6%) were among non-Hispanic white, and 6 were among participants admitted from a nursing home. In the month preceding a person-admission for major surgery, the median number of disabilities was 2 [interquartile range (IQR), 1–4]. Other patient characteristics that were evaluated as potential predictors of functional recovery are described in Table 1.

The most common type of major surgery was musculoskeletal (n = 131; 49.3%), followed by abdominal-gastrointestinal (36; 13.5%), other-type surgeries (35; 13.2%), vascular (31; 11.7%), cardiothoracic (18; 6.8%), and neurologic (15; 5.6%). A majority of the operations were elective (n = 150; 56%). Please see supplementary online material, Appendix, http://links.lww.com/SLA/B589, for full list of operations included.

Of the 266 patient-operations assessed, 174 (65.4%) recovered to their presurgery level of function within 6 months. Among those recovering, median time to recovery was 2 months (IQR, 1–3). For the 245 who survived through 6 months after their first postsurgical interview, the preoperative median number of disabilities was 3 (IQR, 1–7). Death occurred in 16 (6.0%) patients during the 6-month follow-up period, with a median time to death of 74.5 days (IQR, 46.5–122.5).

The bivariate associations of the 20 potential predictors of recovery are presented in Table 1. Five variables were significantly associated with an increased hazard of functional recovery: nonfrailty, vision impairment, functional self-efficacy, physical capacity, and elective operation. Four variables were significantly associated with a decreased hazard of recovery: age, severe hearing impairment, greater increase in postsurgical disability in month after hospital discharge, and type of surgery.

The multivariable associations with time to recovery for the potential predictors in the final model are presented in Table 2. Two factors were significantly associated with an increased likelihood of functional recovery: having an elective operation (HR 1.72) and being nonfrail (HR 1.60). Three factors were associated with a reduced likelihood of functional recovery: greater increases in postsurgical disability (HR 0.87), hearing impairment (HR 0.86), and years of education (HR 0.94).

TABLE 2.

Multivariable Associations of Potential Predictors of Functional Recovery Within 6 Months of First Post-Major Surgery Assessment in Older Adults (N = 266 Person-admissions)

Explanatory Variables*	Hazard of Subdistribution Model
	HR (95% CI)	P†
Demographic
Age, y	1.01 (0.97–1.05)	0.598
Female sex	0.84 (0.60–1.18)	0.314
Non-Hispanic white	0.64 (0.34–1.23)	0.182
Education count, y	0.94 (0.90–0.99)	0.029
Lives alone	1.06 (0.79–1.44)	0.688
Health related
Number chronic conditions	0.96 (0.83–1.10)	0.547
BMI, per unit increase	1.00 (0.96–1.03)	0.778
Nonfrail	1.60 (1.03–2.51)	0.038
Hearing impairment, severe	0.86 (0.76–0.98)	0.019
Cognitive-psychosocial
Functional self-efficacy	1.02 (0.99–1.05)	0.251
Social support	0.99 (0.97–1.02)	0.644
Physical ability
Physical capacity	1.01 (0.94–1.09)	0.721
Increase in disability‡	0.87 (0.82–0.92)	<0.001
Surgery characteristics
Elective operation	1.72 (1.14–2.59)	0.009
Length of stay, days	0.99 (0.96–1.02)	0.563
Admission year	1.04 (0.99–1.09)	0.098
Type of surgery, MSK	0.94 (0.67–1.33)	0.732

^*All explanatory variables included in the final multivariable model are presented. Please see Methods section in the text for full description of all explanatory variables. Note that the following variables were significant in the bivariate analysis (Table 1) but were not retained in the final multivariable model using the backwards selection process described in the statistical analysis section: age; vision impairment; functional self-efficacy; physical capacity; type of surgery.

^†P < 0.05 is considered statistically significant from proportional hazards model with compound symmetric correlation structure.

^‡Increase in disability was a continuous variable, defined as the difference between the first postsurgery disability count and the presurgery disability count. Disability was assessed monthly, in each of 13 functional activities, with a possible score of 0 to 13. Please see text for details.

BMI indicates body mass index; CI, confidence interval; HR, hazard ratio; MSK, musculoskeletal operations; y, years.

DISCUSSION

In this prospective longitudinal study of older adults who survived major surgery with increased disability, we found that approximately 2 out of 3 recovered to their presurgery level of function within 6 months of their first postsurgical interview, with a median time to recovery of 2 months. In a multivariable analysis that evaluated a large array of demographic, health-related, cognitive-psychosocial, physical, and surgical factors, 2 factors—being nonfrail and elective operations—were associated with greater likelihood of functional recovery, whereas 3 factors—hearing impairment, years of education, and greater increases in postsurgery disability—were associated with a lower likelihood of recovery.

The reasons why each of these factors may adversely or positively affect recovery after major surgery are likely multifactorial. Frailty is a state of decreased physiologic reserve and increased vulnerability to both internal and external stressors.²¹ As a measure, frailty is multidimensional, and is clinically unique from both comorbidities and disability.^21,37 Frailty leaves older patients at risk for poor health outcomes both during hospitalizations and after surgery, including complications, prolonged lengths of stay, readmissions, and death.^38,39 One would expect, therefore, that being nonfrail is associated with functional recovery, whereas frailty leads to protracted disability and dependency after the stress of major surgery. In the present study, nonfrailty was associated with a 60% increased likelihood of functional recovery relative to frailty. As such, efforts to minimize frailty—including physical and occupational therapy and optimizing nutrition—may enhance the chance for postsurgical functional recovery.

The primary mechanism through which sensory impairments such as hearing loss lead to postsurgery disability is delirium.^40,41 Pre-existing hearing impairments in older hospitalized patients are predictive of delirium,⁴² which is recognized to be a common, morbid, and costly condition in hospitalized geriatric patients.^43,44 One way to potentially address the hearing-impairment-delirium pathway is with the use of hearing aids, which may reduce the incidence of delirium.⁴⁵ Further investigation of hearing aids or other types of sound amplifying devices in the geriatric population is needed. The association between greater increases in postsurgery disability (at 1 month after discharge) and decreased likelihood of functional recovery makes sense. Disability was operationalized as trouble independently performing a range of 13 tasks; it therefore follows that the higher the postoperative disability count, the more likely it is that a patient’s recovery of function after a major surgery would be hindered.

The association of elective surgery with improved functional recovery is attributable to having the time and ability to optimize each patient preoperatively, and to cancel or reschedule the operation if necessary. On the contrary, nonelective operations put patients at risk for severe physiologic insults,⁴⁶ further reducing physiologic reserve and priming older patients for disability and dependency. In the present study, elective operations had a 72% increased likelihood of functional recovery relative to nonelective operations. In contrast to the other 4 significant predictors, the negative association between more years of education and recovery is counterintuitive. Because of its lack of bivariate, that is, unadjusted, association and the small magnitude of its adjusted association, we believe that its clinical relevance is limited.

Our study represents the first major prospective analysis of functional recovery in older patients after major surgery. A handful of other prospective cohort studies based on elective surgery have attempted to identify independent preoperative risk factors of postoperative functional recovery or decline in older patients.^9–12 A detailed review of these studies, however, reveals a major methodological weakness that undermines their conclusions: they all included patients who did not become more disabled after surgery and hence could not truly “recover.” To accurately study postsurgical functional recovery, patients should only be included if they exhibit higher levels of postsurgical disability.

A major clinical implication of this work is that a geriatric-centered preoperative assessment can help to more accurately inform the postsurgical recovery of older patients. The preoperative assessment should include a standard evaluation of health and an assessment for frailty, functional status, physical capacity, cognitive ability, decision-making capacity, hearing impairment, depression, and social support.^6,47 Although only some of these factors are associated with functional recovery in the current study, the preoperative assessment should not be limited to a focus on postoperative functional recovery, as there are multiple types of postsurgical recovery (including cognitive recovery and physical recovery) which are important for the broader postoperative well-being of older patients. This preoperative assessment would not be time consuming or resource intensive, because there are brief tools that can be used to screen for deficits in each of these areas. It is important for medical and surgical teams to know that the current study does not specifically address the question of whether or not an older adult should have an operation. Rather, the findings herein allow clinicians to more accurately counsel older adults about the specific risk of postoperative functional decline—which to many geriatric patients is a fate worse than death—and to potentially recommend delaying an operation to allow more time for preoperative optimization. This geriatric-centered preoperative assessment will add clarity to preoperative counseling and expectation setting, and ultimately benefit the patient.

The ability of older adults to perform daily activities, such as bathing, dressing, and transferring, are fundamental to independent living. The onset and worsening of postoperative disability, defined as the need for personal assistance to perform these daily activities, marks a serious decline in functional health.¹³ Studies in geriatric community-living persons have shown that increased disability confers a greater risk of institutionalization and death, and leads to substantial financial costs and reductions in quality of life.^48,49 One can therefore infer that a reduction in functional recovery after surgery may lead to a similar increased risk of morbidity and mortality. Preoperative knowledge of the major risk factors for slow recovery may facilitate efforts to reduce postsurgery disability and proactively manage these risks. In the context of what surgery is meant to accomplish—the restoration of health and function—such an evaluation is both logical and necessary.

A key strength of our study is the ability, based on availability of monthly data, to ascertain functional decline shortly after major surgery, so that we could rigorously evaluate functional recovery. This was possible because of the prospective, longitudinal design, with monthly assessments of functional status for more than 18 years. The availability of monthly data on 13 basic, instrumental, and mobility activities allowed us to precisely identify the occurrence and time to recovery during the follow-up period. In addition, the comprehensive home-based assessments provided a large array of potential patient-related and geriatric predictors of functional recovery, even in patients with no prior hospitalizations in the 18-month interval. This is a distinct strength, and is why nonelective surgeries could be rigorously evaluated in our study. Finally, because major surgeries were identified within the context of an ongoing longitudinal study with few exclusion criteria, our results are more generalizable than those of previous prospective studies that enrolled an inception cohort of older patients who underwent a specific subtype of major surgery.

Our study has several limitations that warrant discussion. First, our sample size was relatively small, which limited our power to detect statistically significant effects for factors that have a low prevalence, such as depression and impaired cognition. Second, information about surgical complexity of the different major surgical operations or the sickness level of individual patients was not included in our regression models. Third, we do not have information about postoperative, in-hospital engagement with physical or occupational therapy. Fourth, because our study participants were members of a single health plan in a small urban area, our results may not be generalizable to geriatric patients in other settings.

In summary, this prospective longitudinal study indicates that 4 clinically relevant factors impact postoperative functional recovery after major surgery in older adults. Nonfrailty and elective surgery were associated with recovery to presurgical function, whereas hearing impairment and greater increases in postsurgical disability were associated with protracted disability and dependency. Perioperative assessment of these factors will facilitate efforts by health care providers to enhance the likelihood of functional recovery after major surgery, and will allow patients to be more informed of their postsurgical prognosis.