Abstract
Background
End‐stage kidney disease (ESKD) is commonly associated with critical limb‐threatening ischemia (CLTI) and frailty. Yet there are no specific tools to predict outcomes of CLTI in ESKD, particularly those that incorporate frailty. We aimed to assess the utility of the medical record–based Hospital Frailty Risk (HFR) score in predicting outcomes of CLTI in ESKD.
Methods and Results
We identified patients with ESKD diagnosed with CLTI from the US Renal Data System from 2015 to 2018. These patients were categorized into 3 frailty risk groups on the basis of their HFR scores: low (<5), intermediate (5–10), high‐risk (>10), and on the basis of whether they underwent revascularization (endovascular revascularization [ER]/surgical revascularization [SR]) or not (no revascularization). Primary outcomes of interest included in‐hospital composite of death or major amputation and in‐hospital death. We included 49 454 eligible patients, with ER/SR cohort including 19.8% (n=9777). A total of 88.4% (ER/SR) and 90.0% (no revascularization) were frail on the HFR scale. We found a nonlinear association between HFR score and in‐hospital adverse outcomes. In both cohorts, intermediate and high‐risk HFR scores were associated with greater risk of in‐hospital death (high‐risk, ER/SR: odds ratio, 2.7 [95% CI, 1.6–4.8]; P<0.0001; no revascularization: odds ratio, 7.8 [95% CI, 5.3–11.6]; P<0.01) and composite of in‐hospital major amputation or death (high‐risk, ER/SR: odds ratio, 2.4 [95% CI, 1.9–3.1]; P<0.0001; no revascularization: odds ratio, 1.7 [95% CI, 1.5–1.9]; P<0.0001).
Conclusions
The HFR score can predict risk of in‐hospital death and composite of death or major amputation in patients with ESKD and CLTI. Further data are needed to determine the utility of the HFR score in this population.
Keywords: critical limb threatening ischemia, end stage kidney disease, frailty, major amputation, peripheral artery disease, revascularization
Subject Categories: Peripheral Vascular Disease
Nonstandard Abbreviations and Acronyms
- CLTI
chronic limb‐threatening ischemia
- ER
endovascular revascularization
- ESKD
end‐stage kidney disease
- HFR
Hospital Frailty Risk
- MACE
major adverse cardiovascular events
- MALE
major adverse limb events
- NR
no revascularization
- SR
surgical revascularization
Clinical Perspective.
What Is New?
This is the first clinical study showing the utility of the Hospital Frailty Risk score in predicting clinical outcomes of patients with critical limb‐threatening ischemia and end‐stage kidney disease.
The Hospital Frailty Risk score had a substantial, nonlinear association with in‐hospital death and composite of death or major amputation, and a weaker association with similar outcomes at 1 year.
What Are the Clinical Implications?
The Hospital Frailty Risk score has the potential to be a standardized frailty assessment tool for the risk stratification of critical limb‐threatening ischemia and end‐stage kidney disease, both of which are high‐risk conditions; it demonstrates the ability to further refine risk stratification within this population, helping physicians and patients make more informed decisions regarding potential outcomes in both the revascularized and non‐revascularized groups.
End‐stage kidney disease (ESKD) is a significant risk factor for critical limb‐threatening ischemia (CLTI), and patients with ESKD frequently have CLTI as a comorbidity. 1 , 2 Further, patients with ESKD have poorer survival and are more likely to undergo amputation for CLTI, compared with patients with normal renal function. 3 , 4
Frailty is a state of reduced physiologic reserve to stressors stemming from multisystemic impairments. 5 Frailty has been demonstrated to be an independent risk factor for poor outcomes in patients with CLTI and ESKD, regardless of their coexistence with each other. 6 , 7 Yet frailty is still not routinely assessed in the care of either ESKD or CLTI. This primarily stems from the cumbersome nature of existing frailty assessment tools in day‐to‐day practice, as well as the lack of specific applicability to these conditions. 8 , 9
While different scoring systems exist to predict post‐revascularization outcomes in CLTI, there are few specific to patients with ESKD, particularly those that incorporate frailty. The Hospital Frailty Risk (HFR) score, which can be easily derived from data in inpatient electronic medical records, is based on diagnoses associated with frailty. 10 It was first validated in the United Kingdom and has since demonstrated its utility in predicting prognosis of patients from the general population undergoing endovascular (ER) revascularization or surgical revascularization (SR) for CLTI in the United States. 8 Thus, our study aimed to assess the HFR score's utility in predicting the prognosis of patients with ESKD and concurrent CLTI.
Methods
The data supporting the findings of this study are available from the corresponding author upon reasonable request.
Data Source
The US Renal Data System is a prospective database covering most patients on hemodialysis in the United States. It derives patient‐specific information from various entities, including the Centers for Medicare and Medicaid Services, the Centers for Disease Control and Prevention, the United Network for Organ Sharing, and different ESKD networks. 11 We obtained deidentified patient‐specific data for individuals with ESKD and CLTI from the US Renal Data System database, spanning from October 2015 to December 2018. We did not require ethical approval or informed consent, as this study was exempted by the University of Kansas Medical Center institutional review board due to the deidentification of patient data and prior approval by the ethical committee.
Study Population
We utilized International Classification of Diseases, Tenth Revision (ICD‐10) codes to query the US Renal Data System, identifying patients diagnosed with CLTI, selecting their initial hospitalization as the index hospitalization (Table S1). We excluded individuals younger than 18 years, non‐US residents, those with primary payer coverage other than Medicare Part A and Part B, and those who underwent renal transplant before or during the index hospitalization. Our analysis also considered the duration of dialysis, requiring a minimum of 3 months (equivalent to 90 days) of continuous dialysis leading up to the index hospitalization. This was to ensure a focus on individuals with a consistent dialysis history. CLTI patients were categorized into revascularization (ER/SR) and no revascularization (NR) groups. Figure 1 depicts the patient selection and exclusion process.
Figure 1. Patient selection flowchart.
ESKD indicates end‐stage kidney disease; ICD‐10, International Classification of Diseases, Tenth Revision; and USRDS, United States Renal Database System.
Definition of Frailty
We determined individual HFR scores by analyzing ICD‐10, Clinical Modification (ICD‐10‐CM) diagnostic codes associated with 107 specific medical conditions, either used as primary or secondary diagnoses (all listed in Table 1; Table S2). The HFR scoring system assigns a weight for each diagnostic code, ranging from 0.1 to 4.4, with a higher score depicting a stronger association with frailty. We excluded chronic kidney disease from our population, as our study focused on ESKD. Individuals were classified into low‐risk (<5), intermediate‐risk (5–10), and high‐risk (>10) categories on the basis of their HFR score, with those in the intermediate and high‐risk categories classified as frail.
Table 1.
List of ICD‐10 Codes, Their Prevalence in Each Group, and the Number of Points That Each Variable Contributes to the Development of the Hospital Frailty Risk Score in Patients Undergoing Revascularization Versus Not Undergoing Revascularization for CLTI
| Comorbidities | Revascularization (n=9777), n (%) | No revascularization (n=39 677), n (%) | Score | |
|---|---|---|---|---|
| G81 | Hemiplegia | 36 (0.4) | 191 (0.5) | 4.4 |
| G30 | Alzheimer disease | 60 (0.6) | 274 (0.7) | 4 |
| I69 | Sequelae of cerebrovascular disease | 506 (5.2) | 2016 (5.1) | 3.7 |
| R29 | Other symptoms and signs involving the nervous and musculoskeletal systems (R29.6, Tendency to fall) | 53 (0.5) | 403 (1.0) | 3.6 |
| N39 | Other disorders of urinary system (includes urinary tract infection and urinary incontinence) | 318 (3.3) | 2017 (5.1) | 3.2 |
| F05 | Delirium, not induced by alcohol and other psychoactive substances | 124 (1.3) | 433 (1.1) | 3.2 |
| W19 | Unspecified fall | 28 (0.3) | 232 (0.6) | 3.2 |
| S00 | Superficial injury of head | <11 | 97 (0.2) | 3.2 |
| R31 | Unspecified hematuria | 26 (0.3) | 156 (0.4) | 3.0 |
| B96 | Other bacterial agents as the cause of diseases classified to other chapters (secondary code) | 918 (9.4) | 3793 (9.6) | 2.9 |
| R41 | Other symptoms and signs involving cognitive functions and awareness | 176 (1.8) | 759 (1.9) | 2.7 |
| R26 | Abnormalities of gait and mobility | 88 (0.9) | 623 (1.6) | 2.6 |
| I67 | Other cerebrovascular diseases | 35 (0.4) | 167 (0.4) | 2.6 |
| R56 | Convulsions | 44 (0.5) | 274 (0.7) | 2.5 |
| R40 | Somnolence | 35 (0.4) | 286 (0.7) | 2.5 |
| T83 | Complications of genitourinary prosthetic devices, implants, and grafts | 13 (0.1) | 112 (0.3) | 2.4 |
| S06 | Intracranial injury | <11 | 91 (0.2) | 2.4 |
| S42 | Fracture of shoulder and upper arm | 22 (0.2) | 76 (0.2) | 2.3 |
| E87 | Other disorders of fluid, electrolyte, and acid–base balance | 4305 (44.0) | 17 771 (44.8) | 2.3 |
| M25 | Other joint disorders, not elsewhere classified | 78 (0.8) | 494 (1.3) | 2.3 |
| E86 | Volume depletion | 170 (1.7) | 1026 (2.6) | 2.3 |
| R54 | Senility | 13 (0.1) | 56 (0.1) | 2.2 |
| Z50 | Care involving use of rehabilitation procedures | <11 | <11 | 2.1 |
| F03 | Unspecified dementia | 411 (4.2) | 1837 (4.6) | 2.1 |
| W18 | Other fall on same level | 12 (0.1) | 109 (0.3) | 2.1 |
| Z75 | Problems related to medical facilities and other health care | <11 | 36 (0.1) | 2.0 |
| F01 | Vascular dementia | 69 (0.7) | 322 (0.8) | 2.0 |
| S80 | Superficial injury of lower leg | 23 (0.2) | 293 (0.7) | 2.0 |
| L03 | Cellulitis | 4738 (48.5) | 19 560 (49.3) | 2.0 |
The score presents the top 29 codes, each of which contributes ≥2 points. All covariates are presented in Table S3. Revascularization includes endovascular revascularization and surgical revascularization. CLTI indicates critical limb‐threatening ischemia and ICD‐10, International Classification of Diseases, Tenth Revision.
Study Outcomes
Our primary outcomes were in‐hospital death and composite outcome of in‐hospital death or major amputation. Secondary outcomes were in‐hospital amputation, in‐hospital morbidity, length of stay, and postdischarge outcomes such as death, composite of death or major amputation, major adverse limb events (MALEs), and major adverse cardiovascular events (MACEs) at 1 year. MALEs comprised postdischarge major amputation, postdischarge death, and reintervention, while MACEs encompassed stroke, postdischarge death, or myocardial infarction. Follow‐up for 1‐year outcomes was until 1 year after discharge, or if coming first, death, Medicare A/B coverage termination, kidney transplant, or study end point on December 31, 2019. We identified all outcomes using relevant ICD‐10‐CM codes (Table S3).
Statistical Analysis
We reported continuous variables as mean±SD or median with interquartile range, where applicable, and categorical variables as percentages. For each outcome, we constructed categorical models, using the low‐risk group as a reference. To assess the impact of the HFR score, we used logistic regression. To analyze 1‐year outcomes, we used cumulative incidence function curves, considering death as a competing outcome, with the Fine and Gray model for competing‐risk regression. Multivariable models were adjusted for age, sex, and the Elixhauser comorbidity index. Finally, we created restricted cubic spline curves using HFR score as a continuous variable and used logistic regression to assess nonlinear relationships. We evaluated model performance using receiver operating characteristic curves and area under the curve. All P values were 2‐sided, with significance set at P<0.05. We performed all statistical analyses using SAS 9.4 software (SAS Institute, Cary, NC) and STATA version 17 (StataCorp, College Station, TX).
Results
Study Population
From October 2015 to December 2018, 270 887 patients with ESKD underwent hospitalization for peripheral artery disease, 75755 of whom had CLTI. After applying the exclusion criteria, we analyzed data of 49 454 eligible patients, 19.8% (n=9777) of whom underwent ER/SR, with the remainder not undergoing revascularization (Figure 1). The mean±SD age of the ER/SR cohort was 66.2 (11.4) years, while the mean age of the NR cohort was 63.9 (12.4) years. 62.8% (n=6140) of the ER/SR cohort and 61.7% (n=24 460) of the NR cohort were men.
Frailty and Baseline Characteristics
All comorbidities contributing at least 2 points are listed in Table 1, with all the covariates available in Table S3. “Hemiplegia,” “Alzheimer disease,” and “sequelae of cerebrovascular disease” represented diagnostic codes contributing the most point totals toward the HFR score. “Cellulitis” (48.5% and 49.3%) and “other disorders of fluid, electrolyte, and acid–base balance” (44.0% and 44.8%, respectively) were the most diagnosed codes in both ER/SR and NR cohorts.
A total of 88.4% (n=8645) and 90.0% (n=35 712) of the ER/SR and NR cohorts were frail on the HFR scale, with median (interquartile range) frailty scores of 8.1 (6.0–10.5) and 8.6 (6.5–11.0), respectively (Table 2).
Table 2.
Characteristics of Patients Who Underwent Revascularization Versus No Revascularization for CLTI
| Revascularization (n=9777) | No Revascularization (n=39 677) | |
|---|---|---|
| Age, y, mean±SD | 66.2±11.4 | 63.9±12.4 |
| Male, n (%) | 6140 (62.8) | 24 460 (61.7) |
| Elixhauser comorbidity Index, mean±SD | 6.1±1.7 | 6.1±1.7 |
| Hospital Frailty Index, median (IQR) | 8.1 (6.0–10.5) | 8.6 (6.5–11.0) |
| Hospital Frailty Risk categories | ||
| Low risk (<5 on HFR scale), n (%) | 1132 (11.6) | 3965 (10.0) |
| Intermediate risk (5–10), n (%) | 5737 (58.7) | 22 034 (55.5) |
| High risk (>10), n (%) | 2908 (29.7) | 13 678 (34.5) |
| Frail (≥5), n (%) | 8645 (88.4) | 35 712 (90.0) |
Revascularization includes endovascular revascularization and surgical revascularization. CLTI indicates critical limb‐threatening ischemia; and IQR indicates interquartile range.
Primary Outcomes
In the ER/SR cohort, intermediate and high‐risk HFR categories were associated with 2 times and 2.7 times greater risk of in‐hospital death, respectively, and while in the NR cohort, they conferred 4 times and 7.8 times greater risk, respectively, compared with the low‐risk category. The intermediate‐ and high‐risk groups had 50% and 140% greater risk of the composite outcome of in‐hospital death or major amputation in the ER/SR cohort, with 20% and 70% greater risk in the NR cohort, compared with the low‐risk category (Table 3). Figure 2 shows the nonlinear, increasing trends with increasing HFR scores for these primary outcomes.
Table 3.
In‐Hospital and 1‐Year Outcomes in Patients Who Underwent Revascularization Versus No Revascularization for CLTI on the Basis of HFR Categories
| Revascularization (n=9777) | No revascularization (n=39 677) | |||||
|---|---|---|---|---|---|---|
| HFR categories | Low risk (<5) | Intermediate risk (5–10) | High risk (>10) | Low risk (<5) | Intermediate risk (5–10) | High risk (>10) |
| Sample size, n (%) | 1132 | 5737 | 2908 | 3965 | 22 034 | 13 678 |
| (11.6) | (58.7) | (29.7) | (10.0) | (55.5) | (34.5) | |
| In‐hospital outcomes | ||||||
| Death or amputation | ||||||
| No. (%) | 74 (6.5) | 630 (11.0) | 497 (17.1) | 365 (9.2) | 2646 (12.0) | 2330 (17.0) |
| Odds ratio (95% CI), P value | Reference | 1.5 (1.2–2.0), 0.0012 | 2.4 (1.9–3.1), <0.0001 | Reference | 1.2 (1.1–1.4), 0.0376 | 1.7 (1.5–1.9), <0.0001 |
| Death | ||||||
| No. (%) | 15 (1.3) | 186 (3.2) | 146 (5.0) | 26 (0.7) | 736 (3.3) | 1003 (7.3) |
| Odds ratio (95% CI), P value | Reference | 2.0 (1.2–3.5), 0.0141 | 2.7 (1.6–4.8), <0.0001 | Reference | 4.0 (2.7–5.9), 0.0009 | 7.8 (5.3–11.6), <0.0001 |
| Major amputation | ||||||
| No. (%) | 59 (5.2) | 483 (8.4) | 392 (13.5) | 342 (8.6) | 2014 (9.1) | 1448 (10.6) |
| Odds ratio (95% CI), P value | Reference | 1.5 (1.2–2.0), 0.040 | 2.5 (1.9–3.4), <0.0001 | Reference | 1.1 (0.9–1.2), 0.1530 | 1.2 (1.1–1.4), <0.0001 |
| Morbidity* | ||||||
| No. (%) | 50 (4.4) | 442 (7.7) | 344 (11.8) | 113 (2.9) | 1100 (5.0) | 1074 (7.9) |
| Odds ratio (95% CI), P value | Reference | 1.7 (1.3–2.3), 0.0007 | 2.6 (2.0–3.6), <0.0001 | Reference | 1.7 (1.4–2.1), <0.0001 | 2.7 (2.2–3.3), <0.0001 |
| Length of stay, median (IQR) | ||||||
| Median (IQR) | 7.0 (4.0–11.0) | 9.0 (6.0–14.0) | 13.0 (8.0–19.0) | 5.0 (3.0–8.0) | 6.0 (4.0–10.0) | 8.0 (5.0–13.0) |
| Odds ratio (95% CI), P value | Reference | 2.0 (1.1–3.0), <0.0001 | 6.4 (5.3–7.5), <0.0001 | Reference | 1.7 (1.3–2.0), <0.0001 | 4.2 (3.9–4.6), <0.0001 |
| 1‐y outcomes | ||||||
| Postdischarge amputation/death | ||||||
| No. (%) | 569 (52.1) | 3160 (58.3) | 1698 (62.9) | 1724 (44.9) | 10 395 (50.1) | 7033 (56.9) |
| Hazard ratio (95% CI), P value | Reference | 1.1 (1.0–1.2), 0.1087 | 1.2 (1.1–1.3), 0.0005 | Reference | 1.0 (1.0–1.1), 0.1133 | 1.2 (1.1–1.2), <0.0001 |
| Postdischarge death | ||||||
| No. (%) | 426 (38.1) | 2385 (43.0) | 1358 (49.2) | 1393 (35.4) | 8782 (41.2) | 6315 (49.8) |
| Hazard ratio (95% CI), P value | Reference | 1.0 (0.9–1.2), 0.4085 | 1.2 (1.1–1.3), 0.0022 | Reference | 1.1 (1.0–1.1), 0.0661 | 1.3 (1.2–1.3), <0.0001 |
| MALEs† | ||||||
| No. (%) | 720 (65.7) | 3684 (67.8) | 1911 (70.7) | 1954 (50.7) | 11 416 (54.9) | 7503 (60.6) |
| Hazard ratio (95% CI), P value | Reference | 1.0 (0.9–1.1), 0.6750 | 1.0 (0.9–1.1), 0.4343 | Reference | 1.0 (1.0–1.1), 0.5397 | 1.1 (1.1–1.2), <0.0001 |
| MACEs§ | ||||||
| No. (%) | 548 (50.0) | 2910 (53.0) | 1594 (58.2) | 1747 (40.9) | 10 678 (50.7) | 7273 (58.0) |
| Hazard ratio (95% CI), P value | Reference | 1.0 (0.9–1.1), 0.7398 | 1.1 (1.1–1.2), 0.0327 | Reference | 1.1 (1.1–1.1), 0.0485 | 1.2 (1.1–1.3), <0.0001 |
All models are adjusted for age, sex, Elixhauser comorbidity index. CLTI indicates critical limb‐threatening ischemia; HFR, Hospital Frailty Risk; IQR, interquartile range; MACEs, major adverse cardiac events; and MALEs, major adverse limb events.
Morbidity includes major bleeding requiring blood transfusion, vascular complication, and acute kidney injury requiring dialysis.
MALEs include postdischarge amputation, postdischarge death, and reintervention.
MACEs includes stroke, postdischarge death, and myocardial infarction.
Figure 2. Relationship of HFR score with in‐hospital amputation or mortality (A) and in‐hospital death (B) in patients with ESKD undergoing revascularization or no revascularization for critical limb‐threatening ischemia.
Nonlinear increasing trends for these outcomes against HFR score on restricted cubic spline curves; both intermediate‐ and high‐risk HFR scores correlated with risk of these outcomes. ESKD indicates end‐stage kidney disease; and HFR, Hospital Frailty Risk.
Secondary Outcomes
In the ER/SR cohort, high‐risk HFR score was associated with statistically significant, greater odds of in‐hospital major amputation, morbidity, and length of stay, compared with low‐risk HFR score. Intermediate‐risk HFR score demonstrated a similar, albeit smaller association with in‐hospital major amputation, morbidity, and length of stay. The HFR scores showed similar associations for the NR cohort, except for in‐hospital major amputation in those with intermediate‐risk HFR (Table 3).
In the ER/SR cohort, high‐risk HFR was associated with greater odds of postdischarge death, composite outcome of amputation or death, and MACEs, compared with the low‐risk category. In the NR cohort, high‐risk HFR was associated with greater odds of postdischarge death, amputation or death, MALEs, and MACEs (Table 3). Intermediate‐risk HFR did not demonstrate these associations in either ER/SR or NR cohorts. Figure 3 depicts the relationship between 1‐year amputation/death and 1‐year death, with HFR scores.
Figure 3. Relationship of HFR score with 1‐year amputation or mortality (A) and 1‐year death (B) in patients with ESKD undergoing revascularization or no revascularization for critical limb‐threatening ischemia.
Nonlinear increasing trends for these outcomes against HFR score on restricted cubic spline curves; while high‐risk HFR correlated with greater risk of these outcomes, intermediate‐risk HFR lacked association. ESKD indicates end‐stage kidney disease; and HFR, Hospital Frailty Risk.
In the ER/SR group, the concordance statistic for in‐hospital death without the HFR score was 0.697, compared with 0.708 after including the HFR score. Similarly, in the NR group, the concordance statistic improved from 0.711 to 0.738 after including HFR score. The concordance statistic for in‐hospital death or amputation improved from 0.608 to 0.630 in the ER/SR cohort, and from 0.578 to 0.593 in the NR cohort (Table S4).
Discussion
Our study demonstrates that the HFR score, which can be easily calculated from routinely available data, can predict the in‐hospital outcomes in patients with ESKD and CLTI irrespective of whether they undergo revascularization. Patients ESKD and CLTI are a high‐risk group with significant comorbidities, and thus frailty is a significant issue. We found a significant, nonlinear association of higher HFR scores with the primary study outcomes of in‐hospital death and the composite outcome of in‐hospital death or major amputation in both cohorts. High‐risk HFR score was significantly associated with greater odds of secondary outcomes of in‐hospital major amputation, morbidity, and length of stay, in both cohorts. This had a similar but weaker association with postdischarge death, composite of amputation or death, and MACEs in both cohorts and with MALEs in the NR cohort, compared with the low‐risk category.
Despite advances in medical treatment, clinical outcomes of CLTI remain abysmal, with reported 2‐year survival rates of only ≈50% overall, 7 and ≈40% in patients with ESKD. 12 The diagnosis of peripheral artery disease in patients with ESKD is often delayed, as classical claudication symptoms are often absent, and non–peripheral artery disease leg symptoms are frequent, owing to comorbid diabetic neuropathy, pruritus, arthritis, and restless leg syndrome. Other textbook‐described physical signs are neither specific nor sensitive, and in chronic kidney disease/ESKD, the ankle–brachial index can be falsely elevated. 13 Thus, many patients with ESKD are first diagnosed at the stage of CLTI. 14 Further, patients with impaired kidney function undergo fewer revascularization procedures for peripheral artery disease than people with normal renal function, despite comparable severity. 15 Thus, the availability of a convenient scoring tool to predict clinical outcomes in this medically complex group of patients could potentially aid in clinical decision making.
Frailty is common in ESKD, with studies reporting a prevalence of up to 80%, depending on the tool used. 16 , 17 In our sample, we found a much larger prevalence, with almost 90% of both ER/SR and NR cohorts categorized as frail on the basis of the HFR score. Commonly used frailty assessment tools in practice include Fried's Frailty Scale, which has been the most validated in chronic kidney disease/ESKD, the Clinical Frailty Scale, and the Frailty Index, among others. These bedside tools have the limitations of being time consuming, requiring subjective assessment, and being tough for physicians not accustomed to their regular usage. 9 , 10 , 18 Further, when applied to CLTI, they may overestimate frailty due to the limited mobility of these patients secondary to rest pain or foot ulceration. 8 These challenges have led to scanty usage of frailty assessment in patients with ESKD and CLTI, despite the prognostic information they may provide. 8 , 9 Similarly, several scoring tools have been developed to predict survival, both overall and amputation free, in patients undergoing ER/SR for CLTI. 8 , 19 , 20 , 21 While some did account for frailty, 7 , 22 , 23 their assessment and definition of frailty were often inadequate, with poorly generalizable results. 8 Further, most of them were not specific to ESKD.
The HFR score addresses most of these limitations. It uses ICD diagnostic codes, which are already available in electronic medical records. Thus, it adds no additional effort for clinicians, and there is no possibility of interobserver variability. Further, it considers 107 comorbidities, ensuring comprehensiveness. Finally, it has been validated in multiple populations and for multiple conditions and shown to correlate well with preexisting frailty scales. 10
Recently, Abualhin et al 12 analyzed the outcomes of 180 patients with ESKD undergoing ER/SR for CLTI. They found that clinical success of revascularization (patient survival, with a vital limb not requiring reintervention) negatively correlated with advanced age, the presence of coronary artery disease, and Texas University Wound Classification Stage D (infected wound), deriving a predictive score from these 3 variables. Similarly, Farchioni et al 24 analyzed the outcomes of 80 patients with ESKD undergoing ER for CLTI and derived a predictive score for major amputation in these patients. They found that the severity of the arterial lesion, according to the Trans‐Atlantic Inter‐Society Consensus classification, and extension of trophic lesions, according to the wound, ischemia, and foot classification, correlated with the probability of major amputation. Older studies have found that different negative predictive variables affect survival and limb salvage: advanced age (>80 years), peritoneal dialysis and greater number of years on dialysis, coronary artery disease, chronic obstructive pulmonary disease, extensive tissue loss, and being dependent for ambulation. 25 , 26 , 27 However, the generalizability of these prior findings may be questioned given that these studies were retrospective, observational, single‐center studies that considered a small number of comorbidities.
In comparison, we analyzed a large, multicentric database, and determined outcomes by incorporating the effect of >100 comorbidities. We found that an increased HFR score conferred an incremental increase in the risk of in‐hospital adverse outcomes. High‐risk HFR score conferred 140% and 70% greater risk of the composite outcome of in‐hospital death or major amputation in the ER/SR and NR cohorts, respectively, compared with the low‐risk score. Similarly, it conferred 4 times and 7.8 times greater risk of in‐hospital death in the ER/SR and NR cohorts, compared with the low‐risk score. High‐risk HFR score was also associated with greater morbidity and length of stay in both cohorts, reflecting increased economic costs for these patients. For postdischarge outcomes, we found similar but weaker associations. High‐risk HFR was associated with greater odds of postdischarge death, composite of amputation or death, and MACEs in both cohorts, with an additional association with MALEs in the NR cohort. These findings add to the literature showing the association between frailty and worse cardiovascular outcomes. 28
Prior meta‐analyses have shown that patients with ESKD have higher risk of all‐cause death and major amputation than patients without ESKD. Further, even with similar primary vessel patency, patients with ESKD had worse revascularization success, regardless of the mode, compared with patients without ESKD. 29 While we had similar findings overall, we also demonstrated a stepwise increase in the odds of poor outcomes with increasing HFR score. This likely reflects the reality of frailty, which rather than a dichotomous “frail or not frail” classification, represents the accumulation of deficits over a spectrum. 5 , 8 Further, the HFR score correlated with in‐hospital outcomes more strongly than postdischarge outcomes. This likely reflects that frailty is dynamic and can either improve or deteriorate with time. 30 Thus, the HFR score may provide greater insight for physicians to estimate clinical outcomes of patients with ESKD and CLTI, particularly at the time of admission. This would also improve shared decision making with patients, regardless of whether the final therapeutic decision involves revascularization or medical therapy alone. Further, this score could be used to identify patients who would require more intensive posttreatment rehabilitation, or even pretreatment “prehabilitation,” to ameliorate adverse events. 31
We also found that Alzheimer disease contributed one of the greatest point totals to the HFR frailty score in our sample. Impaired kidney function is associated with earlier onset and greater severity of cognitive impairment that worsens with declining renal function. 32 , 33 , 34 This involves a complex interplay of vascular risk factors, advanced age, and uremia, among others. 35 However, patients with dementia are commonly excluded from clinical trials, 36 and thus the optimum treatment modality for CLTI in these patients is unclear. 37 In those with significant cognitive impairment, more frequent follow‐up and close involvement with family are warranted, to ensure patient‐centric clinical decision making in the setting of CLTI. Exercise training and targeted cognitive training has been shown to improve postoperative outcomes by reducing cognitive frailty, providing a potential prehabilitation avenue for patients undergoing ER/SR. 38 , 39 Overall, our analysis demonstrates that the HFR score can predict in‐hospital outcomes and prognosis of CLTI in patients with ESKD, potentially representing a convenient tool for clinicians.
Limitations
Our study has several limitations. As with other administrative databases, the US Renal Data System has the potential to under‐/overcode for comorbidities. It also does not reflect the severity of comorbidities; for example, dementia, which was a significant contributory variable in our study. We also did not have data on the anatomic characteristics of CLTI, such as lesion length, complexity, and specific details of the intervention received in the ER/SR cohort. Patients with impaired kidney function manifest vascular calcification earlier and have more diffuse, multilevel stenoses than patients with normal kidney function, influencing clinical decision making. 14 , 29 The HFR score does not include physiologic measures of frailty that are used in bedside frailty assessments, nor does it account for laboratory data. Further studies are needed to compare the predictive role of HFR score with bedside frailty assessment tools in patients with ESKD and CLTI. Further, we could not assess certain outcomes like wound healing. Finally, while we compared outcomes across frailty categories, within these categories, we did not compare outcomes of nonrevascularized and revascularized patients. Our objective was primarily to study predictive value of the HFR score in predicting outcomes of ESKD with CLTI, regardless of the therapeutic avenue taken. Now that its value has been shown, and given the complexity involved in clinical decision making for revascularization, a comparison of NR and ER/SR cohorts in ESKD with CLTI warrants a separate, follow‐up study, assessing patients within the same HFR category.
Conclusions
This study of a large sample from a multi‐institutional database provides valuable information on the outcomes of patients with ESKD and CLTI. The HFR score can predict risk of in‐hospital death or amputation in patients with ESKD and CLTI. Further data are needed to guide clinical decision making and determine the utility of the HFR score in this population.
Sources of Funding
None.
Disclosures
None.
Supporting information
Tables S1–S4
Acknowledgments
The data reported here have been supplied by the US Renal Data System The interpretation and reporting of these data are the responsibility of the author(s) and in no way should be seen as an official policy or interpretation of the US government.
This manuscript was sent to Yen‐Hung Lin, MD, PhD, Associate Editor, for review by expert referees, editorial decision, and final disposition.
Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.124.036963
For Sources of Funding and Disclosures, see page 9.
References
- 1. Conte MS, Bradbury AW, Kolh P, White JV, Dick F, Fitridge R, Mills JL, Ricco JB, Suresh KR, Murad MH, et al. Global vascular guidelines on the management of chronic limb‐threatening ischemia. J Vasc Surg. 2019;69:3S–125S.e40. doi: 10.1016/j.jvs.2019.02.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Johansen KL, Garimella PS, Hicks CW, Kalra PA, Kelly DM, Martens S, Matsushita K, Sarafidis P, Sood MM, Herzog CA, et al. Central and peripheral arterial diseases in chronic kidney disease: conclusions from a kidney disease: improving global outcomes (KDIGO) controversies conference. Kidney Int. 2021;100:35–48. doi: 10.1016/j.kint.2021.04.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Dawson DB, Telles‐Garcia NA, Atkins JL, Mina GS, Abreo AP, Virk CS, Dominic PS. End‐stage renal disease patients undergoing angioplasty and bypass for critical limb ischemia have worse outcomes compared to non‐ESRD patients: systematic review and meta‐analysis. Catheter Cardiovasc Interv. 2021;98:297–307. doi: 10.1002/ccd.29688 [DOI] [PubMed] [Google Scholar]
- 4. Meyer A, Fiessler C, Stavroulakis K, Torsello G, Bisdas T, Lang W; CRITISCH collaborators . Outcomes of dialysis patients with critical limb ischemia after revascularization compared with patients with normal renal function. J Vasc Surg. 2018;68:822–829.e1. doi: 10.1016/j.jvs.2017.12.048 [DOI] [PubMed] [Google Scholar]
- 5. Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. Lancet. 2013;381:752–762. doi: 10.1016/S0140-6736(12)62167-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Kallenberg MH, Kleinveld HA, Dekker FW, van Munster BC, Rabelink TJ, van Buren M, Mooijaart SP. Functional and cognitive impairment, frailty, and adverse health outcomes in older patients reaching ESRD‐A systematic review. Clin J Am Soc Nephrol. 2016;11:1624–1639. doi: 10.2215/CJN.13611215 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Takeji Y, Yamaji K, Tomoi Y, Okazaki J, Tanaka K, Nagae A, Jinnouchi H, Hiramori S, Soga Y, Ando K. Impact of frailty on clinical outcomes in patients with critical limb ischemia. Circ Cardiovasc Interv. 2018;11:e006778. doi: 10.1161/CIRCINTERVENTIONS.118.006778 [DOI] [PubMed] [Google Scholar]
- 8. Majmundar M, Patel KN, Doshi R, Mehta H, Vindhyal MR, Hance KA, Ali A, Gupta K. Prognostic value of hospital frailty risk score and clinical outcomes in patients undergoing revascularization for critical limb‐threatening ischemia. J Am Heart Assoc. 2023;12:e030294. doi: 10.1161/JAHA.123.030294 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Nagaraju SP, Shenoy SV, Gupta A. Frailty in end stage renal disease: current perspectives. Nefrologia (Engl Ed). 2022;42:531–539. doi: 10.1016/j.nefroe.2021.05.008 [DOI] [PubMed] [Google Scholar]
- 10. Gilbert T, Neuburger J, Kraindler J, Keeble E, Smith P, Ariti C, Arora S, Street A, Parker S, Roberts HC, et al. Development and validation of a hospital frailty risk score focusing on older people in acute care settings using electronic hospital records: an observational study. Lancet. 2018;391:1775–1782. doi: 10.1016/S0140-6736(18)30668-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. U.S. Renal Data System . USRDS Annual Data Report: Epidemiology of Kidney Disease in the United States. National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD: National Institutes of Health; 2023:2023. [Google Scholar]
- 12. Abualhin M, Gargiulo M, Bianchini Massoni C, Mauro R, Morselli‐Labate AM, Freyrie A, Faggioli G, Stella A. A prognostic score for clinical success after revascularization of critical limb ischemia in hemodialysis patients. J Vasc Surg. 2019;70:901–912. doi: 10.1016/j.jvs.2018.11.034 [DOI] [PubMed] [Google Scholar]
- 13. Garimella PS, Hirsch AT. Peripheral artery disease and chronic kidney disease: clinical synergy to improve outcomes. Adv Chronic Kidney Dis. 2014;21:460–471. doi: 10.1053/j.ackd.2014.07.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Hsieh MY, Chuang SY, Lee CK, Luo CM, Cheng CH, Liao MT, Lin PL, Yang TF, Wu CC. Risks and outcomes of critical limb ischemia in hemodialysis patients: a prospective cohort study. Clin Kidney J. 2022;16:585–595. doi: 10.1093/ckj/sfac263 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Arinze NV, Gregory A, Francis JM, Farber A, Chitalia VC. Unique aspects of peripheral artery disease in patients with chronic kidney disease. Vasc Med. 2019;24:251–260. doi: 10.1177/1358863X18824654 [DOI] [PubMed] [Google Scholar]
- 16. Zhang Q, Ma Y, Lin F, Zhao J, Xiong J. Frailty and mortality among patients with chronic kidney disease and end‐stage renal disease: a systematic review and meta‐analysis. Int Urol Nephrol. 2020;52:363–370. doi: 10.1007/s11255-019-02369-x [DOI] [PubMed] [Google Scholar]
- 17. Kojima G. Prevalence of frailty in end‐stage renal disease: a systematic review and meta‐analysis. Int Urol Nephrol. 2017;49:1989–1997. doi: 10.1007/s11255-017-1547-5 [DOI] [PubMed] [Google Scholar]
- 18. Chin M, Kendzerska T, Inoue J, Aw M, Mardiros L, Pease C, Andrew MK, Pakhale S, Forster AJ, Mulpuru S. Comparing the hospital frailty risk score and the clinical frailty scale among older adults with chronic obstructive pulmonary disease exacerbation. JAMA Netw Open. 2023;6:e2253692. doi: 10.1001/jamanetworkopen.2022.53692 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Moxey PW, Brownrigg J, Kumar SS, Crate G, Holt PJ, Thompson MM, Jones KG, Hinchliffe RJ. The BASIL survival prediction model in patients with peripheral arterial disease undergoing revascularization in a university hospital setting and comparison with the FINNVASC and modified PREVENT scores. J Vasc Surg. 2013;57:1–7. doi: 10.1016/j.jvs.2012.04.074 [DOI] [PubMed] [Google Scholar]
- 20. Schanzer A, Mega J, Meadows J, Samson RH, Bandyk DF, Conte MS. Risk stratification in critical limb ischemia: derivation and validation of a model to predict amputation‐free survival using multicenter surgical outcomes data. J Vasc Surg. 2008;48:1464–1471. doi: 10.1016/j.jvs.2008.07.062 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Biancari F, Salenius JP, Heikkinen M, Luther M, Ylönen K, Lepäntalo M. Risk‐scoring method for prediction of 30‐day postoperative outcome after infrainguinal surgical revascularization for critical lower‐limb ischemia: a Finnvasc registry study. World J Surg. 2007;31:217–227. doi: 10.1007/s00268-006-0242-y [DOI] [PubMed] [Google Scholar]
- 22. Morisaki K, Yamaoka T, Iwasa K, Ohmine T. Influence of frailty on treatment outcomes after revascularization in patients with critical limb ischemia. J Vasc Surg. 2017;66:1758–1764. doi: 10.1016/j.jvs.2017.04.048 [DOI] [PubMed] [Google Scholar]
- 23. Houghton JSM, Nickinson ATO, Morton AJ, Nduwayo S, Pepper CJ, Rayt HS, Gray LJ, Conroy SP, Haunton VJ, Sayers RD. Frailty factors and outcomes in vascular surgery patients: a systematic review and meta‐analysis. Ann Surg. 2020;272:266–276. doi: 10.1097/SLA.0000000000003642 [DOI] [PubMed] [Google Scholar]
- 24. Farchioni L, Gennai S, Giuliani E, Cuccì A, Lauricella A, Leone N, Silingardi R. A prognostic risk score for major amputation in dialysis patients with chronic limb‐threatening ischemia after endovascular revascularization. Int Angiol. 2021;40:206–212. doi: 10.23736/S0392-9590.21.04523-5 [DOI] [PubMed] [Google Scholar]
- 25. Ramdev P, Rayan SS, Sheahan M, Hamdan AD, Logerfo FW, Akbari CM, Campbell DR, Pomposelli FB Jr. A decade experience with infrainguinal revascularization in a dialysis‐dependent patient population. J Vasc Surg. 2002;36:969–974. doi: 10.1067/mva.2002.128297 [DOI] [PubMed] [Google Scholar]
- 26. Jones DW, Dansey K, Hamdan AD. Lower extremity revascularization in end‐stage renal disease. Vasc Endovasc Surg. 2016;50:582–585. doi: 10.1177/1538574416674843 [DOI] [PubMed] [Google Scholar]
- 27. Korn P, Hoenig SJ, Skillman JJ, Kent KC. Is lower extremity revascularization worthwhile in patients with end‐stage renal disease? Surgery. 2000;128:472–479. doi: 10.1067/msy.2000.108049 [DOI] [PubMed] [Google Scholar]
- 28. Liu X, Tou NX, Gao Q, Gwee X, Wee SL, Ng TP. Frailty and risk of cardiovascular disease and mortality. PLoS One. 2022;17:e0272527. doi: 10.1371/journal.pone.0272527 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Anantha‐Narayanan M, Sheikh AB, Nagpal S, Jelani QU, Smolderen KG, Regan C, Ionescu C, Ochoa Chaar CI, Schneider M, Llanos‐Chea F, et al. Systematic review and meta‐analysis of outcomes of lower extremity peripheral arterial interventions in patients with and without chronic kidney disease or end‐stage renal disease. J Vasc Surg. 2021;73:331–340.e4. doi: 10.1016/j.jvs.2020.08.032 [DOI] [PubMed] [Google Scholar]
- 30. Romero‐Ortuno R, Hartley P, Knight SP, Kenny RA, O'Halloran AM. Frailty index transitions over eight years were frequent in the Irish longitudinal study on ageing. HRB Open Res. 2021;4:63. doi: 10.12688/hrbopenres.13286.1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Norris CM, Close JCT. Prehabilitation for the frailty syndrome: improving outcomes for our Most vulnerable patients. Anesth Analg. 2020;130:1524–1533. doi: 10.1213/ANE.0000000000004785 [DOI] [PubMed] [Google Scholar]
- 32. Banerjee A, Fowkes FG, Rothwell PM. Associations between peripheral artery disease and ischemic stroke: implications for primary and secondary prevention. Stroke. 2010;41:2102–2107. doi: 10.1161/STROKEAHA.110.582627 [DOI] [PubMed] [Google Scholar]
- 33. Harhay MN, Xie D, Zhang X, Hsu CY, Vittinghoff E, Go AS, Sozio SM, Blumenthal J, Seliger S, Chen J, et al. Cognitive impairment in non‐dialysis‐dependent CKD and the transition to dialysis: findings from the chronic renal insufficiency cohort (CRIC) study. Am J Kidney Dis. 2018;72:499–508. doi: 10.1053/j.ajkd.2018.02.361 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Berger I, Wu S, Masson P, Kelly PJ, Duthie FA, Whiteley W, Parker D, Gillespie D, Webster AC. Cognition in chronic kidney disease: a systematic review and meta‐analysis. BMC Med. 2016;14:206. doi: 10.1186/s12916-016-0745-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Drew DA, Weiner DE, Sarnak MJ. Cognitive impairment in CKD: pathophysiology, management, and prevention. Am J Kidney Dis. 2019;74:782–790. doi: 10.1053/j.ajkd.2019.05.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Prusaczyk B, Cherney SM, Carpenter CR, DuBois JM. Informed consent to research with cognitively impaired adults: transdisciplinary challenges and opportunities. Clin Gerontol. 2017;40:63–73. doi: 10.1080/07317115.2016.1201714 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Cutmore C, Aitken S. Managing chronic limb‐threatening Ischaemia in patients at the extremes of older age requires a patient‐focused approach. World J Surg. 2022;46:2832–2833. doi: 10.1007/s00268-022-06701-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Li H, Su W, Dang H, Han K, Lu H, Yue S, Zhang H. Exercise training for mild cognitive impairment adults older than 60: a systematic review and meta‐analysis. J Alzheimers Dis. 2022;88:1263–1278. doi: 10.3233/JAD-220243 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Fink A, Humeidan M, Bentov I. Cognitive Prehabilitation: how can we counter neurocognitive frailty? Eur Urol Focus. 2024;10:16–19. doi: 10.1016/j.euf.2023.11.005 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Tables S1–S4