Functional status was measured using Minimum Data Set for Nursing Homes assessments of self-performance of ADLs. Residents were classified as having functional decline if they had a 2-point or greater increase in their MDS-ADL score, maintenance of functional status was defined as no change or a 1-point change (positive or negative), and improvement in functional status was defined as a 2-point or greater decrease in the MDS-ADL score.
Oresanya L, Zhao S, Gan S, Fries BE, Goodney PP, Covinsky KE, Conte MS, Finlayson E. Functional Outcomes After Lower Extremity Revascularization in Nursing Home ResidentsA National Cohort Study. JAMA Intern Med. 2015;175(6):951-957. doi:10.1001/jamainternmed.2015.0486
Lower extremity revascularization often seeks to allow patients with peripheral arterial disease to maintain the ability to walk, a key aspect of functional independence. Surgical outcomes in patients with high levels of functional dependence are poorly understood.
To determine functional status trajectories, changes in ambulatory status, and survival after lower extremity revascularization in nursing home residents.
Using full Medicare claims data for 2005 to 2009, we identified nursing home residents who underwent lower extremity revascularization. With the Minimum Data Set for Nursing Homes activities of daily living summary score, we examined changes in their ambulatory and functional status after surgery. We identified patient and surgery characteristics associated with a composite measure of clinical and functional failure—death or nonambulatory status 1 year after surgery.
All nursing homes in the United States participating in Medicare or Medicaid.
Nursing home residents who underwent lower extremity revascularization.
Main Outcomes and Measures
Functional status, ambulatory status, and death.
During the study period, 10 784 long-term nursing home residents underwent lower extremity revascularization. Prior to surgery, 75% of the residents were not walking; 40% had experienced functional decline. One year after surgery, 51% of patients had died, 28% were nonambulatory, and 32% had sustained functional decline. Among 1672 residents who were ambulatory before surgery, 63% had died or were nonambulatory at 1 year; among 7188 who were nonambulatory, 89% had died or were nonambulatory. After multivariate adjustment, factors independently associated with death or nonambulatory status were 80 years or older (adjusted hazard ratio [AHR], 1.28; 95% CI, 1.16-1.40), cognitive impairment (AHR, 1.23; 95% CI, 1.18-1.29), congestive heart failure (AHR, 1.16; 95% CI, 1.11-1.22), renal failure (AHR, 1.09; 95% CI, 1.04-1.14), emergent surgery (AHR, 1.29; 95% CI, 1.23-1.35), nonambulatory status before surgery (AHR, 1.88; 95% CI, 1.78-1.99), and decline in activities of daily living before surgery (AHR, 1.23; 95% CI, 1.18-1.28).
Conclusions and Relevance
Of nursing home residents in the United States who undergo lower extremity revascularization, few are alive and ambulatory 1 year after surgery. Most who were still alive had gained little, if any, function.
Lower extremity peripheral arterial disease is a common condition among nursing home residents,1,2 and a substantial number of such individuals are at risk for critical limb ischemia. Lower extremity revascularization is frequently performed with the goal of preserving functional independence through limb preservation. These procedures, however, are frequently associated with operative risk, and their benefit in terms of maintaining the ability to walk is variable.3,4 Because nursing home residents, in general, have high levels of functional dependence unrelated to peripheral arterial disease and higher rates of mortality after most invasive procedures,5- 7 a thorough examination of the outcomes of lower extremity revascularization is needed to inform the use of these procedures.
Although numerous studies8- 10 have shown that functional dependence is associated with postoperative mortality and complications, information about functional outcomes after lower extremity revascularization in such patients is sparse. A review of studies evaluating functional outcomes following lower extremity revascularization identified only 107 nonambulatory patients and 195 functionally dependent patients for whom ambulatory or functional status following lower extremity revascularization was reported.11 Recently, Vogel et al12 reported outcomes following lower extremity revascularization in 702 nursing home residents. Our study, however, focused on a comparison of functional outcomes between open and endovascular revascularization procedures in a selected subset of residents.
Using data from national Medicare claims and the Minimum Data Set for Nursing Homes (MDS), we examined survival, changes in ambulatory status, and functional trajectories after lower extremity revascularization in nursing home residents in the United States between 2005 and 2008, with follow-up through 2009. With regression techniques, we identified factors associated with poor functional outcomes following revascularization.
We linked data from the full Medicare Inpatient Files with the MDS (version 2.0) to identify long-term nursing home residents undergoing lower extremity revascularization. The Medicare Inpatient File contains discharge abstracts for all fee-for-service inpatient hospitalizations for Medicare beneficiaries. The MDS is a mandatory assessment of all nursing home residents who reside in facilities participating in Medicare or Medicaid programs—virtually all US nursing homes. The MDS assessments are completed at admission, readmission, quarterly, and when the resident experiences a change in clinical status. The study was approved by the University of California, San Francisco, Human Research Protection Program.
Residents undergoing lower extremity revascularization were identified in the Medicare Inpatient File by International Classification of Diseases, 9th Edition (ICD-9) procedure codes for open (ICD-9 codes 38.08, 38.18, 39.29) and endovascular (ICD-9 codes 39.50, 39.99) revascularization. Nursing home residents were classified as long-term if they underwent 2 or more consecutive full or quarterly MDS assessments performed over 90 days apart during the 6 months before surgery, indicating a preoperative nursing home length of stay greater than 90 days. To exclude residents who had recently undergone prior revascularization attempts, residents who had undergone a lower extremity revascularization procedure in the 2 years preceding the study period were excluded from analysis. Based on these criteria, we identified all long-term nursing home residents older than 67 years undergoing lower extremity revascularization procedure in the United States between 2005 and 2008, with follow-up to 2009.
The outcomes were survival, functional status, and ambulatory status. Outcomes were assessed at 3, 6, 9, and 12 months after surgery. Death was determined using date of death information from the Medicare Denominator file. We defined ambulatory as a score of 0 (independent), 1 (supervision), or 2 (limited assistance) in the MDS ambulation domain. Conversely, we defined nonambulatory as a score of 3 (extensive assistance), 4 (total dependence), or 8 (activity did not occur).
Functional status was measured using MDS assessments of self-performance of activities of daily living (ADLs). Data on ADL performance in the MDS includes information about mobility in 7 ADLs: mobility while in bed, transferring, ambulation, dressing, eating, toileting, and personal hygiene. The resident’s performance of each of these ADLs is rated using a scoring system of 0 to 4 points, with 0 indicating independence and 4 indicating total dependence. The sum of these 7 scores—the MDS-ADL score—has been validated against standardized measures of functional independence.13 Possible total scores range from 0 to 28, with 0 indicating independence in all ADLs and 28 indicating total dependence in all ADLs. Residents were classified as having functional decline if they had a 2-point or greater increase in their MDS-ADL score. This definition of functional decline is consistent with those in prior studies14 and identifies nursing home residents who have a clear decline in function. For example, a patient who previously required supervision with bathing and now requires extensive assistance with bathing has experienced a 2-point decline. Maintenance of functional status was defined as no change or a 1-point change (positive or negative), and improvement in functional status was defined as a 2-point or more decrease in the MDS-ADL score. We defined 2 distinct adverse clinical outcomes: (1) a combined measure of nonambulatory status or death at 1 year and (2) functional decline among 1-year survivors.
Demographic data (age, sex, and race), the admission acuity (elective, or urgent or emergent), and the nature of the procedure (open or endovascular) were obtained from Medicare files. Comorbid diagnoses were obtained from Medicare Provider Analysis and Review files for the surgical hospitalization and MDS assessments performed prior to surgery. We examined factors commonly associated with adverse outcomes after vascular surgery, including congestive heart failure, diabetes mellitus, cerebrovascular disease, chronic obstructive pulmonary disease, coronary artery disease, renal failure, and cognitive impairment.15,16 Baseline functional and ambulatory status were based on the MDS assessment before the surgical hospitalization. Residents were classified as having preoperative functional decline if they had a 2-point or greater MDS-ADL score increase in the 6 months before their surgical admission. Residents were defined as having cognitive impairment if they met at least 1 of 3 criteria: disease diagnosis of dementia on MDS assessment, ICD-9 diagnosis code for dementia in Medicare claims, or a MDS cognitive performance scale score (CPS) of 3 or greater. The MDS-CPS is a validated scale of cognitive function generated from 5 items in the MDS. Possible scores range from 0 to 6, with 0 indicating intact cognitive function and 6 indicating very severe cognitive impairment.17 A CPS score of 3 indicates moderate cognitive impairment.
We compared descriptive variables by using χ2 analysis for dichotomous categorical variables and analysis of variance for categorical and continuous variables. The composite measure of death or nonambulatory status was defined as a “time to event” outcome. If the resident died, the death date was the event date. If the resident survived 1 year, nonambulatory status on the last assessment was the event and date of assessment was the event date. We used the Kaplan-Meier method to estimate cumulative death or nonambulatory status at 1 year and Cox proportional hazards models to estimate the risk associated with individual resident characteristics and being nonambulatory or dead at 1 year after surgery. We fitted log Poisson regression models to estimate the relative risks of functional decline among 1-year survivors associated with individual resident and procedure characteristics.
Functional trajectories before and after surgery were explored using the mixed-effects spline models to incorporate multiple measurements for each patient. Specifically, restricted cubic spline models with 5 knots placed at 3 months and 1 month before surgery and 1, 4, and 12 months after surgery were used. Models included fixed and random effects for the coefficients of the spline, implying that each person’s measurements were scattered around a patient-specific smooth curve; these patient-specific smooth curves were then departures from the average smooth trajectory for the population. Fixed effects for age, sex, race, and comorbidity were also included, allowing the population trajectory to shift based on the baseline characteristics. To account for residents who were alive but did not have ADL assessments (missing assessments) in months 9 to 12, we used multiple imputations to predict MDS-ADL scores at 12 months.
For all analyses, P = .05 on 2-sided significance test was considered statistically significant. The University of California at San Francisco committee on human research approved this study.
Between 2005 and 2008, 10 784 nursing home residents underwent lower extremity revascularization. Their mean age was 82 years, 37% were men, and 80% were white. Comorbid disease was common; 60% had cognitive impairment, 57% had congestive heart failure, and 29% had renal failure (Table 1). Quiz Ref IDRevascularization was performed electively in 67% of cases, and over one-half of all procedures were performed using an endovascular approach. Prior to surgery, 75% of patients were nonambulatory. The mean baseline MDS-ADL score was 14.5, indicating a high level of preoperative functional dependence. A substantial proportion of residents (40%) had experienced functional decline in the 6 months prior to surgery.
Overall 1-year mortality was 51% to 43% among ambulatory residents and 53% among nonambulatory residents (Figure 1). Quiz Ref IDOne year after surgery, 82% of residents had either died or were nonambulatory. Among residents who were ambulatory before surgery, 63% had died or were nonambulatory. Among residents who were nonambulatory prior to surgery, 89% had died or were nonambulatory. In multivariate analysis, preoperative ambulatory status had the strongest association with this measure of clinical failure (hazard ratio [HR], 1.88; 95% CI, 1.78-1.99) (Table 2). Quiz Ref IDWhen compared with open surgery, an endovascular approach was associated with a lower likelihood of clinical failure (HR, 0.88; 95% CI, 0.85-0.92). Among nursing home residents who were alive 1 year after surgery, 34% who were ambulatory at baseline became nonambulatory, and 24% who were nonambulatory at baseline became ambulatory.
The mean MDS-ADL impairment score for the cohort increased from 14.5 points at baseline to 16.7 points at 1 year following revascularization; the 2.2-point change represents a clinically significant decline in function. In the mixed-effects model, the greatest functional decline occurred within the first 3 months following surgery, during which the mean MDS-ADL score increased to 17.8. By 6 months, the mean MDS-ADL score decreased slightly to 17.2, followed by a small gradual functional improvement among patients who were alive 1 year after revascularization.
The proportion of nursing home residents whose function declined from preoperative levels was greatest during the first 3 months after surgery: 40% of the entire cohort and 56% of surviving residents (Figure 2). Three months after surgery, 29% of residents had died and 32% had improved or maintained baseline function. Six months after surgery, 39% of residents had died and 51% of the survivors had functional decline. At 1 year, 50% of residents had died, 36% of survivors had improved or maintained ADL function, and 64% of survivors had experienced decline. After multivariate adjustment, nonambulatory status before surgery had the strongest association with functional decline among survivors (adjusted relative risk, 3.12; 95% CI, 2.84-3.43) (Table 3).
We found that a substantial number of nursing home residents in the United States undergo lower extremity revascularization, and many gain little, if any, function. The mortality rate, however, is high, with half of residents dying within a year of surgery. One year after surgery, 13% of the initial cohort was ambulatory and 18% had maintained or improved on their baseline functional status. Quiz Ref IDResidents who were ambulatory prior to revascularization had better outcomes than those who were nonambulatory; revascularization rarely allowed a nonambulatory resident to become ambulatory.
Outcomes in nursing home residents are substantially worse than those in the general population of patients undergoing lower extremity revascularization. The Vascular Study Group of Northern New England (VSGNNE) registry10 prospectively collects data on all lower extremity revascularization procedures performed within the region. A study10 evaluating ambulatory function 1 year after lower extremity bypass showed that 83% of patients were alive and walking 1 year after open lower extremity bypass. In our cohort, only 17% of patients were alive and walking 1 year after undergoing lower extremity revascularization.
Several groups have been reported series in which patients—even the most elderly—had excellent outcomes following lower extremity revascularization. For example, Pomposelli et al18 and Ballotta et al19 reported excellent outcomes after revascularization in octogenarians and nonagenarians. The 1-year mortality rates in these studies were 5% and 9%, respectively, and the rates of maintenance of ambulatory function were 78% and 80%. These studies, however, examined outcomes in a group of selected older adults at a single institution. The study by Vogel et al12 examined a highly selected subset of nursing home residents: their cohort excluded residents who underwent nonelective procedures, had a hospital stay of more than 30 days, died in the hospital, or had more than 20 hospital stays during the study period. Our study examined a population-based cohort of nursing home residents who underwent revascularization across a variety of hospitals.
Our findings are consistent with those of previous studies that found that individuals who are ambulatory prior to lower extremity revascularization have better outcomes that those who are nonambulatory, and that revascularization only rarely allows a patient who is nonambulatory to become ambulatory postoperatively.20,21 Even among the nursing home residents who were ambulatory before revascularization, however, we identified a high rate of mortality and functional decline. It is unclear whether the poor outcomes were a result of revascularization per se or the insufficient physiologic reserve of many of the patients who underwent the procedure.
Our study has several limitations. We lacked detailed information about the indication for surgery. Although it might be presumed that most nursing home patients undergoing lower extremity revascularization would have critical limb ischemia, not claudication, we have no information about the severity of ischemia and the prevalence and duration of nonhealing wounds or gangrene. Quiz Ref IDIn addition, we were not able to identify patients with critical limb ischemia who did not undergo surgery and who were treated with either palliative amputation or simply pain control and localized wound care. As dismal as the outcomes may seem, we were unable to compare the changes in the ability to walk and functional decline in patients undergoing revascularization with patients treated with either amputation or palliation. Although it is possible that the stress of surgery and hospitalization in patients who are already frail may worsen their functional dependence, it is also possible that nursing home residents with critical limb ischemia who do not undergo revascularization may experience greater functional decline.
Treatment decisions for elderly, frail patients with critical limb ischemia are complex; shared decision-making is needed. Our findings can inform conversations between physicians, patients, and families about the risks and expected outcomes of surgery and whether the surgery is likely to be worthwhile. Our findings also highlight the importance of carefully considering a prognosis independent of vascular disease and assessing the goals of care. Ambulatory function, although clearly an important goal, may not be the primary objective of treatment and may be impossible to attain. Nonambulatory patients with refractory ischemic rest pain, wounds that do not heal despite months of nursing care, or worsening gangrene seek palliation for the relief of symptoms. Among the treatment options, palliative care, primary amputation, and lower extremity revascularization are associated with different risks, benefits, and expected outcomes. Thus, our findings should be interpreted cautiously; successful relief of pain, healing of wounds, and avoidance of major amputation may benefit some of the patients who underwent lower extremity revascularization.
Corresponding Author: Emily Finlayson, MD, MS, Philip R. Lee Institute for Health Policy Studies, 3333 California St, San Francisco, CA 94118 (email@example.com).
Accepted for Publication: February 4, 2015.
Published Online: April 6, 2015. doi:10.1001/jamainternmed.2015.0486.
Author Contributions: Dr Finlayson had full access to all of 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: Oresanya, Fries, Conte, Finlayson.
Acquisition, analysis, or interpretation of data: Oresanya, Zhao, Gan, Fries, Goodney, Covinsky, Finlayson.
Drafting of the manuscript: Oresanya.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Oresanya, Zhao, Gan, Finlayson.
Administrative, technical, or material support: Fries, Goodney.
Study supervision: Finlayson.
Conflict of Interest Disclosures: None reported.
Funding/Support: Dr Finlayson was supported by a National Institute on Aging/ Paul B. Beeson Clinical Scientist Development Award in Aging (K08AG028965) and the University of California, San Francisco, Claude D. Pepper Older Americans Independence Center.
Role of the Funder/Sponsor: The funding sources 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.