Hospital frailty risk score and outcomes of chronic mesenteric ischemia revascularization: Nationwide Inpatient Sample 2005–2020 analysis Authors: Chien-Min Lai, Hui-Pu Liu, Wen-Yen Chang Published: 2024 ABSTRACT Chronic mesenteric ischemia (CMI) typically affects older adults, and often requiring revascularization. Frailty, a state of decreased physiological reserve, may impact the outcomes of CMI treatment. The purpose of was to determine the impact of frailty, assessed by the Hospital Frailty Risk Score, on in-hospital outcomes after CMI revascularization. Data from the Nationwide Inpatient Sample 2005–2020, were analyzed in this retrospective study. Inclusion criteria were adults ≥20 years old who underwent revascularization for CMI. Patients were categorizing patients into low-, intermediate-, and high-frailty risk groups. Logistic regression analyses, adjusted for confounders, was used to assess associations between frailty risk and outcomes. The study included 4100 patients with a mean age of 71.2 years, of whom 71% were female. Patients in the intermediate and high frailty risk groups had significantly increased odds of in-hospital mortality (intermediate risk: adjusted odds ratio [aOR] = 3.33, 95% confidence interval [CI]: 2.07–5.35; high risk: aOR = 3.68, 95% CI: 2.15–6.31), unfavorable discharge (intermediate risk: aOR = 2.25, 95% CI: 1.83–2.77; high risk: aOR = 3.77, 95% CI: 2.91–4.88), prolonged hospital stay (intermediate risk: aOR = 2.40; high risk: aOR = 4.85), transfusions (intermediate risk: aOR = 1.97; high risk: aOR = 4.58), and overall complications (intermediate risk: aOR = 3.84; high risk: aOR = 8.74) compared to low-risk patients. Frailty is associated with worse outcomes of patients with CMI undergoing revascularization. Thus, it is important to incorporate frailty assessment in preoperative risk stratification and management of CMI revascularization. 1. Introduction Chronic mesenteric ischemia (CMI) predominantly affects older adults, and is relatively uncommon compared to other vascular diseases that affect older persons.[1,2] As the arteries supplying blood to the intestines become narrowed or blocked, patients with CMI experience symptoms such as abdominal pain, weight loss, and food fear. If left untreated, CMI can lead to severe complications due to the compromised blood flow to the intestines.[1,2] Prompt intervention is important once the condition is diagnosed, and treatment options for CMI include lifestyle modifications, medications, and revascularization procedures.[3–5] The prognosis of patients with CMI depends on factors such as the degree of arterial blockage, the timeliness of intervention, and the presence of comorbidities. Revascularization plays a crucial role in managing CMI, aiming to restore adequate blood flow to the mesenteric arteries.[6,7] Methods of revascularization include endovascular interventions and open surgical revascularization.[6] The choice of the revascularization approach depends on factors such as the overall health of the patient and the severity and location of the arterial blockages. Timely and appropriate revascularization is essential to mitigate the consequences of CMI and improve in-hospital and long-term outcomes. Frailty is a complex and multifaceted concept that refers to a state of decreased physiological reserve and increased vulnerability to stressors, resulting from a cumulative decline across multiple physiological systems, and is most commonly observed in older persons and the elderly.[8] The Hospital Frailty Risk Score (HFRS) is a tool designed to assess the frailty status of patients admitted to hospitals, and uses clinical data to generate a numerical score that reflects the level of frailty.[9] Frail individuals often exhibit diminished strength, endurance, balance, and overall functional capacity, making them more susceptible to adverse health outcomes, including disability, hospitalization, and mortality.[10–12] The impact of frailty on in-hospital outcomes after revascularization for CMI is an underexplored topic, and is important as both frailty and CMI are more common in older persons and the elderly. Frailty has been previously shown to be associated with a worse recovery process, higher postoperative complications, and lower overall success of revascularization procedures of the extremities and carotid artery.[13–15] Investigating the potential impact of frailty in patients undergoing CMI revascularization will provide valuable insights into tailoring interventions to the specific needs of frail patients. Thus, the purpose of this study was to use a United States (US) nationally representative database to examine the impact of frailty on the outcomes of revascularization for CMI. 2. Methods 2.1. Data source The population-based, retrospective study extracted data from the Nationwide Inpatient Sample (NIS) database from 2005 to 2020. The NIS is the largest publicly available inpatient care database in the US, and is administered by the Healthcare Cost and Utilization Project (HCUP). The NIS database comprises a 20% stratified sample of annual US hospital admissions (https://hcup-us.ahrq.gov/nisoverview.jsp). Patient data include, but are not limited to, primary and secondary diagnoses and procedures, patient demographic information, admission and discharge status, length of hospital stay (LOS), total hospitalization charges, and hospital characteristics. Patient diagnoses and procedures are recorded using the International Classification of Diseases, 9th edition (ICD-9), and 10th edition (ICD-10) coding systems. 2.2. Ethics statement This study complies with the terms of the NIS data-use agreement. The data utilized in this study were obtained through the Online HCUP Central Distributor. Given that this study solely involved the analysis of secondary data, there was no direct involvement of the general public or patients. 2.3. Study population and variables This study included adults ≥20 years old admitted to hospitals with a principal diagnosis of CMI and who underwent endovascular or open surgical revascularization. Patients with missing data, diagnosed as acute mesenteric ischemia, and those that received both endovascular and open surgical revascularization were excluded. The outcomes assessed included all-cause inpatient mortality, unfavorable discharge, prolonged LOS, total hospital costs, transfusion, and common in-hospital complications. Prolonged LOS was defined as ≥75th LOS of the study population. Unfavorable discharge was defined as discharge to a long-term care facility. Diseases and procedures were identified in the database by ICD-9 and ICD-10 codes, as summarized in Table S1, Supplemental Digital Content, http://links.lww.com/MD/O231. 2.4. Assessment of frailty The HFRS was developed by Gilbert et al,[9] and has been validated by analyzing a wide range of ICD diagnostic codes indicative of various conditions commonly associated with frailty. Example conditions include chronic pulmonary disease, heart failure, and volume depletion. The HFRS has been validated among over 1.04 million hospitalized individuals. The selection of the specific ICD codes used for the HFRS is supported by their proven effectiveness in prior studies; thus, supporting the accuracy of the tool for evaluating frailty in older populations. Patients were stratified according to their frailty risk using the HFRS: low frailty risk, a score below 5, termed “non-frail”; medium frailty risk, a score ranging from 5 to 15; and high frailty risk (a score exceeding 15). The original validation study of the HFRS used these cutoffs, and were selected to optimally distinguish individuals based on their health outcomes. This segmentation approach enhances the accuracy and relevance of frailty risk evaluation. The HFRS has also been extensively implemented in various research projects that utilize administrative codes.[16,17] The components and ICD codes to define HFRS are summarized in Tables S2 and S3, Supplemental Digital Content, http://links.lww.com/MD/O231. 2.5. Statistical analysis The weights provided with the NIS were used to generate estimates of the number of admissions each year. Patient descriptive statistics were presented as number (n) and weighted percentage (%), or mean and standard error. Odds ratios (ORs), beta coefficients, and 95% confidence intervals (CIs) for the associations between HFRS and outcomes were analyzed using logistic regression analysis. Multivariable regression was adjusted for age (continuous), insurance status, admission type, coronary artery disease, congestive heart failure, diabetes, cerebrovascular disease, chronic pulmonary disease, chronic kidney disease, type of revascularization, and hospital region. Stratified analysis was performed to determine the association between frailty and the outcomes across different age groups, admission types, and revascularization types. All P-values were 2-sided, and a value of P < .05 was considered statistically significant. All statistical analyses were performed using the statistical software package SAS software version 9.4 (SAS Institute Inc., Cary, NC). 3. Results 3.1. Patient selection The study population selection process is depicted in Figure 1. There were 4323 hospital discharges for patients ≥20 years old with CMI who underwent revascularization procedures during the period from 2005 to 2020. Patients with missing information on study outcomes (n = 43) were excluded. Additionally, 58 patients who received both endovascular therapy and surgery, 111 patients diagnosed with acute mesenteric ischemia, and 11 patients without weight values in the dataset were further excluded. Finally, 4100 patients were included in the study. This study sample could be extrapolated to a total of 20,048 hospitalized patients in the entire US after weighting (Fig. 1). Flow diagram of patient selection. 3.2. Characteristics of the study population Patient characteristics are summarized in Table 1. The mean patient age was 71.2 years, 71% were female, and a majority (83.3%) were identified as White. The population was stratified into risk groups defined by HFRS, consisting of 750 patients in the high-risk category, 1589 in the intermediate-risk category, and 1761 in the low-risk category. Significant differences were observed in the distribution of age, insurance status, and admission type among the different risk groups. Notably, the low-risk group exhibited the youngest age and a highest percentage of elective admissions (P < .001). Approximately 76% of patients received endovascular therapy, and the remaining were treated with open surgery. Within the low-risk group, 78.3% received endovascular therapy, which was the largest proportion in the 3 groups (P = .006). Examining comorbidities, there were significant differences in the prevalence of several conditions across the 3 risk groups. These differences were statistically significant for the comorbidities of coronary artery disease, congestive heart failure, diabetes, cerebrovascular disease, chronic kidney disease (all, P < .001), and chronic pulmonary disease (P = .002) (Table 1). Table 1Patient characteristics.Total(N = 4100)HFRSP-valueLow risk (< 5) (n = 1761)Intermediate risk (5–15)(n = 1589)High risk (> 15)(n = 750) Demography Age, years71.2 ± 0.269.8 ± 0.372.2 ± 0.272.2 ± 0.4 <.001 20–59625 (15.2)338 (19.1)199 (12.5)88 (11.8) <.001 60–691037 (25.3)462 (26.3)387 (24.3)188 (24.9) 70–791423 (34.7)573 (32.5)570 (35.9)280 (37.3) 80+1015 (24.8)388 (22.0)433 (27.3)194 (26.0)Sex.283 Male1174 (28.6)520 (29.6)435 (27.3)219 (29.1) Female2926 (71.4)1241 (70.4)1154 (72.7)531 (70.9)Race/ethnicity.471 White3022 (83.3)1291 (83.3)1150 (82.7)581 (84.6) Black219 (6.1)90 (5.9)86 (6.2)43 (6.4) Hispanic226 (6.2)99 (6.4)96 (6.9)31 (4.5) Other159 (4.4)69 (4.4)59 (4.2)31 (4.6) Missing47421219864Household income.108 Q11140 (28.2)516 (29.6)417 (26.7)207 (28.0) Q21121 (27.8)481 (27.6)443 (28.4)197 (26.9) Q31033 (25.7)436 (25.2)423 (27.1)174 (23.7) Q4743 (18.4)306 (17.6)277 (17.8)160 (21.3) Missing63222912Insurance status <0.001 Medicare/Medicaid3318 (81.1)1350 (76.7)1330 (84.0)638 (85.0) Private including HMO640 (15.6)345 (19.6)211 (13.3)84 (11.3) Self-pay/no-charge/other136 (3.3)64 (3.7)44 (2.7)28 (3.7) Missing6240Admission type <.001 Elective1388 (33.8)713 (40.4)466 (29.4)209 (27.6) Emergent2698 (66.2)1044 (59.6)1115 (70.6)539 (72.4) Missing14482 Major comorbidities Coronary artery disease1954 (47.7)749 (42.6)784 (49.5)421 (56.0) <.001 Congestive heart failure822 (20.2)283 (16.2)368 (23.2)171 (23.0) <.001 Diabetes1275 (31.2)469 (26.7)528 (33.3)278 (37.0) <.001 Cerebrovascular disease401 (9.8)54 (3.1)131 (8.3)216 (28.6) <.001 Chronic pulmonary disease1404 (34.3)551 (31.4)572 (36.1)281 (37.3) .002 Chronic kidney disease813 (20.0)181 (10.3)419 (26.6)213 (28.6) <.001 Severe Liver disease19 (0.5)5 (0.3)10 (0.6)4 (0.5).253Rheumatic disease158 (3.9)60 (3.4)71 (4.5)27 (3.6).211Any malignancy125 (3.1)48 (2.8)54 (3.4)23 (3.1).552 CCI <.001 0–22140 (52.0)1163 (65.8)724 (45.4)253 (33.6)3–51623 (39.7)500 (28.5)701 (44.2)422 (56.5)6+337 (8.3)98 (5.7)164 (10.4)75 (9.9) Type of revascularization .006 Endovascular therapy3116 (75.9)1380 (78.3)1177 (73.9)559 (74.6)Surgery984 (24.1)381 (21.7)412 (26.1)191 (25.4) Hospital bed size .478Small339 (8.1)157 (8.8)121 (7.5)61 (7.9)Medium912 (22.3)383 (21.8)367 (23.2)162 (21.8)Large2834 (69.5)1215 (69.4)1095 (69.3)524 (70.2)Missing15663 Location/teaching status .179Rural200 (4.8)92 (5.2)79 (4.9)29 (3.8)Urban nonteaching1080 (26.2)463 (25.9)398 (25.0)219 (29.3)Urban teaching2805 (69.0)1200 (68.9)1106 (70.1)499 (66.9)Missing15663 Hospital region <.001 Northeast734 (18.2)351 (20.3)282 (17.9)101 (13.6)Midwest993 (24.2)389 (22.1)405 (25.5)199 (26.4)South1729 (42.0)768 (43.3)638 (40.1)323 (43.0)West644 (15.6)253 (14.3)264 (16.5)127 (16.9)CCI = Charlson Comorbidity Index, HFRS = Hospital Frailty Risk Scores.Continuous variables are presented as mean ± SE. Categorical variables are presented as unweighted counts (weighted percentage).P-value < .05 shown in bold. Patient characteristics. CCI = Charlson Comorbidity Index, HFRS = Hospital Frailty Risk Scores. Continuous variables are presented as mean ± SE. Categorical variables are presented as unweighted counts (weighted percentage). P-value < .05 shown in bold. 3.3. Inpatient outcomes of the study population Inpatient outcomes are summarized in Table 2, and there were significant differences among the 3 frailty risk groups. In-hospital mortality, unfavorable discharge, prolonged LOS, transfusion rates, and complication rates exhibited statistically significant differences, with the high-risk group demonstrating the highest proportions (P < .001). The high-risk group had the highest total hospital costs, followed by the intermediate risk group, and the low-risk group had the lowest costs (P < .001) (Table 2). Table 2Inpatient outcomes.OutcomesTotal (N = 4100)HFRSP-valueLow risk (< 5) (n = 1761)Intermediate risk (5–15)(n = 1589)High risk (> 15)(n = 750)In-hospital mortality149 (3.6)24 (1.3)83 (5.2)42 (5.6) <.001 Unfavorable discharge*615 (15.6)139 (8.0)288 (19.1)188 (26.6) <.001 Prolonged LOS*,†910 (23.1)220 (12.6)417 (27.8)273 (38.7) <.001 Total hospital cost, dollars99,140.3 ± 1677.172,671.6 ± 1314.0113,317.4 ± 2795.7131,089.8 ± 4260.0 <.001 Transfusion541 (13.1)127 (7.1)223 (14.0)191 (25.4) <.001 Complications1786 (43.7)371 (21.1)852 (53.8)563 (75.2) <.001 CVA236 (5.7)6 (0.3)53 (3.4)177 (23.4) <.001 VTE156 (3.8)36 (2.1)74 (4.7)46 (6.2) <.001 Pneumonia161 (3.9)27 (1.5)82 (5.1)52 (6.9) <.001 Sepsis299 (7.3)13 (0.7)126 (7.9)160 (21.5) <.001 Infection512 (12.5)78 (4.5)231 (14.6)203 (27.2) <.001 Respiratory failure361 (8.8)49 (2.7)170 (10.7)142 (18.9) <.001 Mechanical ventilation228 (5.6)38 (2.2)108 (6.9)82 (10.9) <.001 Acute kidney injury550 (13.5)67 (3.9)304 (19.3)179 (24.1) <.001 Shock145 (3.6)25 (1.4)62 (3.9)58 (7.8) <.001 Bleeding470 (11.5)81 (4.6)216 (13.7)173 (23.1) <.001 Wound complication69 (1.7)19 (1.1)23 (1.5)27 (3.6) <.001 Device complication145 (3.5)19 (1.1)75 (4.6)51 (6.8) <.001 Digestive system154 (3.8)40 (2.3)85 (5.4)29 (3.8) <.001 Arterial dissection32 (0.8)15 (0.9)13 (0.8)4 (0.6).688CVA = cerebrovascular accident, HFRS = Hospital Frailty Risk Scores, LOS = length of hospital stay, VTE = venous thromboembolism.Continuous variables are presented as mean ± SE. Categorical variables are presented as unweighted counts (weighted percentage).P-value < .05 shown in bold.* Excluded patients who died in the hospital.† Length of stay > 75th percentile (10 days). Inpatient outcomes. CVA = cerebrovascular accident, HFRS = Hospital Frailty Risk Scores, LOS = length of hospital stay, VTE = venous thromboembolism. Continuous variables are presented as mean ± SE. Categorical variables are presented as unweighted counts (weighted percentage). P-value < .05 shown in bold. Excluded patients who died in the hospital. Length of stay > 75th percentile (10 days). 3.4. Associations between HFRS risk group and outcomes Table 3 presents the associations between frailty risk group and inpatient outcomes. In the multivariable analysis adjusting for relevant confounders revealed that compared to the low-risk group, both the high-risk group (adjusted odds ratio [aOR] = 3.68, 95% CI: 2.15–6.31, P < .001) and the intermediate-risk group (aOR = 3.33, 95% CI: 2.07–5.35, P < .001) had a significantly increased likelihood of in-hospital mortality. In comparison to the low-risk group, the intermediate-risk group (aOR = 2.25, 95% CI: 1.83–2.77, P < .001) and high-risk group (aOR = 3.77, 95% CI: 2.91–4.88, P < .001) had an elevated risk of unfavorable discharge. Similar patterns were identified for prolonged LOS (intermediate risk: aOR = 2.40, 95% CI: 1.98–2.90, P < .001; high risk: aOR = 4.85, 95% CI: 3.85–6.12, P < .001), transfusions (intermediate risk: aOR = 1.97, 95% CI: 1.54–2.52, P < .001; high risk: aOR = 4.58, 95% CI: 3.46–6.05, P < .001), and overall complications (intermediate risk: aOR = 3.84, 95% CI: 3.28–4.49, P < .001; high risk: aOR = 8.74, 95% CI: 7.07–10.80, P < .001) (Table 3). Table 3Associations between HFRS risk groups and in-hospital mortality, unfavorable discharge, prolonged LOS, transfusion, and complications.OutcomesaOR‡ (95% CI)P-value In-hospital mortality HFRS (Intermediate risk vs. Low risk)3.33 (2.07–5.35) <.001 HFRS (High risk vs. Low risk)3.68 (2.15–6.31) <.001 Unfavorable discharge * HFRS (Intermediate risk vs. Low risk)2.25 (1.83–2.77) <.001 HFRS (High risk vs. Low risk)3.77 (2.91–4.88) <.001 Prolonged LOS*,†HFRS (Intermediate risk vs. Low risk)2.40 (1.98–2.90) <.001 HFRS (High risk vs. Low risk)4.85 (3.85–6.12) <.001 Transfusion HFRS (Intermediate risk vs. Low risk)1.97 (1.54–2.52) <.001 HFRS (High risk vs. Low risk)4.58 (3.46–6.05) <.001 Complications HFRS (Intermediate risk vs. Low risk)3.84 (3.28–4.49) <.001 HFRS (High risk vs. Low risk)8.74 (7.07–10.80) <.001 aOR = adjusted odds ratio, CI = confidence, HFRS = Hospital Frailty Risk Scores, LOS = length of hospital stay.P-value < .05 shown in bold.* Excluded patients who died in the hospital.† Length of stay > 75th percentile (10 days).‡ Adjusted for related variables with a value of P < .05 in Table 1, expect for CCI, including age (continuous), insurance status, admission type, coronary artery disease, congestive heart failure, diabetes, cerebrovascular disease, chronic pulmonary disease, chronic kidney disease, type of revascularization, and hospital region. Associations between HFRS risk groups and in-hospital mortality, unfavorable discharge, prolonged LOS, transfusion, and complications. aOR = adjusted odds ratio, CI = confidence, HFRS = Hospital Frailty Risk Scores, LOS = length of hospital stay. P-value < .05 shown in bold. Excluded patients who died in the hospital. Length of stay > 75th percentile (10 days). Adjusted for related variables with a value of P < .05 in Table 1, expect for CCI, including age (continuous), insurance status, admission type, coronary artery disease, congestive heart failure, diabetes, cerebrovascular disease, chronic pulmonary disease, chronic kidney disease, type of revascularization, and hospital region. Furthermore, patients in the high-risk group (an increase of 47,489.73 USD; 95% CI: 38,387.31–56,592.15, P < .001) and intermediate risk group (an increase of 29,168.01 USD; 95% CI: 23,579.47–34,756.56, P < .001) demonstrated significantly higher total hospital costs compared to those at low risk (Table 4). Table 4Associations between HFRS risk groups and total hospital cost.OutcomesaBeta* (95% CI)P-value Total hospital cost, USD HFRS (Intermediate risk vs. Low risk)29,168.01 (23,579.47–34,756.56) <.001 HFRS (High risk vs. Low risk)47,489.73 (38,387.31–56,592.15) <.001 aBeta = adjusted beta, CI = confidence, HFRS = Hospital Frailty Risk Scores.P-value < .05 shown in bold.* Adjusted for related variables with a value of P < .05 in Table 1, expect for CCI, including age (continuous), insurance status, admission type, coronary artery disease, congestive heart failure, diabetes, cerebrovascular disease, chronic pulmonary disease, chronic kidney disease, type of revascularization, and hospital region. Associations between HFRS risk groups and total hospital cost. aBeta = adjusted beta, CI = confidence, HFRS = Hospital Frailty Risk Scores. P-value < .05 shown in bold. Adjusted for related variables with a value of P < .05 in Table 1, expect for CCI, including age (continuous), insurance status, admission type, coronary artery disease, congestive heart failure, diabetes, cerebrovascular disease, chronic pulmonary disease, chronic kidney disease, type of revascularization, and hospital region. Associations between frailty risk and specific complications are shown in Figure 2. The high- and intermediate-risk groups had significantly higher odds of venous thromboembolism, pneumonia, bleeding, acute kidney injury, shock, infection, and respiratory failure/mechanical ventilation compared to the low risk group (Fig. 2). Associations between HFRS-defined frailty and complications. aAdjusted for variables with a value of P < .05 in Table 1, and expect for CCI including age (continuous), insurance status, admission type, coronary artery disease, congestive heart failure, diabetes, cerebrovascular disease, chronic pulmonary disease, chronic kidney disease, type of revascularization, and hospital region. CCI = Charlson Comorbidity Index; HFRS = Hospital Frailty Risk Score. 3.5. Stratified association between frailty risk and in-hospital mortality, prolonged LOS, complications by age, admission type, revascularization type We further performed stratified analysis based on age, admission type, and revascularization type, and the results are summarized in Table 5. After adjusting for confounding factors, patients in the intermediate- and high-risk groups had significantly higher odds of all outcomes among different subgroups, except for in-hospital mortality in patients age <60 years old (Table 5). Table 5Associations between HFRS-defined frailty and inpatient outcomes stratified by age, admission type, and type of revascularization.OutcomesaOR‡ (95% CI)P-valueAge < 60 years (n = 625)In-hospital mortality HFRS (Intermediate risk vs. Low risk)2.01 (0.66–6.10).217 HFRS (High risk vs. Low risk)1.90 (0.41–8.68).409Prolonged LOS*,† HFRS (Intermediate risk vs. Low risk)2.42 (1.65–3.53) <.001 HFRS (High risk vs. Low risk)3.61 (2.06–6.32) <.001 Complications HFRS (Intermediate risk vs. Low risk)2.99 (2.16–4.13) <.001 HFRS (High risk vs. Low risk)7.89 (4.91–12.67) <.001 Age 60–69 years (n = 1037)In-hospital mortality HFRS (Intermediate risk vs. Low risk)3.83 (1.42–10.37) .008 HFRS (High risk vs. Low risk)4.37 (1.53–12.49) .006 Prolonged LOS*,† HFRS (Intermediate risk vs. Low risk)2.22 (1.54–3.20) <.001 HFRS (High risk vs. Low risk)5.61 (3.59–8.76) <.001 Complications HFRS (Intermediate risk vs. Low risk)4.88 (3.69–6.45) <.001 HFRS (High risk vs. Low risk)15.71 (10.25–24.07) <.001 Age ≥ 70 years (n = 2438)In-hospital mortality HFRS (Intermediate risk vs. Low risk)3.32 (1.88–5.85) <.001 HFRS (High risk vs. Low risk)3.64 (1.91–6.93) <.001 Prolonged LOS*,† HFRS (Intermediate risk vs. Low risk)2.49 (1.94–3.21) <.001 HFRS (High risk vs. Low risk)4.85 (3.62–6.51) <.001 Complications HFRS (Intermediate risk vs. Low risk)3.78 (3.09–4.62) <.001 HFRS (High risk vs. Low risk)7.56 (5.82–9.81) <.001 Admission type (Elective) (n = 1388)In-hospital mortality HFRS (Intermediate risk vs. Low risk)3.24 (1.16–9.01) .025 HFRS (High risk vs. Low risk)7.63 (2.47–23.57) <.001 Prolonged LOS*,† HFRS (Intermediate risk vs. Low risk)2.12 (1.49–3.01) <.001 HFRS (High risk vs. Low risk)4.46 (2.92–6.81) <.001 Complications HFRS (Intermediate risk vs. Low risk)3.98 (3.10–5.11) <.001 HFRS (High risk vs. Low risk)10.83 (7.40–15.85) <.001 Admission type (Emergent) (n = 2698)In-hospital mortality HFRS (Intermediate risk vs. Low risk)3.35 (2.00–5.59) <.001 HFRS (High risk vs. Low risk)3.08 (1.69–5.62) <.001 Prolonged LOS*,† HFRS (Intermediate risk vs. Low risk)2.57 (2.05–3.22) <.001 HFRS (High risk vs. Low risk)5.14 (3.90–6.78) <.001 Complications HFRS (Intermediate risk vs. Low risk)3.78 (3.13–4.55) <.001 HFRS (High risk vs. Low risk)8.04 (6.27–10.31) <.001 Type of revascularization (Endovascular) (n = 3116)In-hospital mortality HFRS (Intermediate risk vs. Low risk)4.45 (2.41–8.20) <.001 HFRS (High risk vs. Low risk)4.53 (2.19–9.34) <.001 Prolonged LOS*,† HFRS (Intermediate risk vs. Low risk)2.55 (2.01–3.23) <.001 HFRS (High risk vs. Low risk)5.32 (4.02–7.05) <.001 Complications HFRS (Intermediate risk vs. Low risk)3.79 (3.16–4.55) <.001 HFRS (High risk vs. Low risk)8.25 (6.50–10.46) <.001 Type of revascularization (Open surgery) (n = 984)In-hospital mortality HFRS (Intermediate risk vs. Low risk)2.30 (1.28–4.12) .005 HFRS (High risk vs. Low risk)3.18 (1.60–6.30) .001 Prolonged LOS*,† HFRS (Intermediate risk vs. Low risk)2.27 (1.70–3.04) <.001 HFRS (High risk vs. Low risk)4.27 (3.02–6.04) <.001 Complications HFRS (Intermediate risk vs. Low risk)3.91 (3.00–5.10) <.001 HFRS (High risk vs. Low risk)11.48 (7.11–18.54) <.001 aOR = adjusted odds ratio, CI = confidence, HFRS = Hospital Frailty Risk Scores, LOS = length of hospital stay.P-value < .05 shown in bold.* Excluded patients who died in the hospital.† Length of stay > 75th percentile (10 days).‡ Adjusted for related variables with a value of P < .05 in Table 1, expect for CCI and stratified variables, including age (continuous), insurance status, admission type, coronary artery disease, congestive heart failure, diabetes, cerebrovascular disease, chronic pulmonary disease, chronic kidney disease, type of revascularization, and hospital region. Associations between HFRS-defined frailty and inpatient outcomes stratified by age, admission type, and type of revascularization. aOR = adjusted odds ratio, CI = confidence, HFRS = Hospital Frailty Risk Scores, LOS = length of hospital stay. P-value < .05 shown in bold. Excluded patients who died in the hospital. Length of stay > 75th percentile (10 days). Adjusted for related variables with a value of P < .05 in Table 1, expect for CCI and stratified variables, including age (continuous), insurance status, admission type, coronary artery disease, congestive heart failure, diabetes, cerebrovascular disease, chronic pulmonary disease, chronic kidney disease, type of revascularization, and hospital region. 4. Discussion This study used data from the US NIS to examine the influence of frailty, as determined by the HFRS, on the in-hospital outcomes of patients undergoing revascularization procedures for CMI. The results showed that in comparison to patients categorized as low-risk for frailty, those in the intermediate-risk group exhibited significantly increased risks of adverse outcomes, including in-hospital mortality (3.3-times), unfavorable discharge (2.3-times), prolonged LOS (2.4-times), the need for transfusions (2.0-times), and a higher incidence of overall complications (3.8-times). The risks were even more pronounced in the high-risk group; patients had a 3.7- to 8.7-times higher likelihood of experiencing these adverse outcomes. Additionally, both the high- and intermediate-risk groups incurred higher total hospital costs than the low-risk group. In terms of specific complications, patients in the high- and intermediate-risk groups were significantly more likely to have a venous thromboembolism, pneumonia, bleeding, acute kidney injury, shock, infection, and respiratory failure requiring mechanical ventilation, compared to the low-risk group. This pattern of increased risk was consistent across various patient subgroups. Collectively, these findings indicate HFRS-assessed frailty is a strong predictor for poor outcomes in patients with CMI undergoing revascularization. CMI is more common in older and elderly persons,[5,18] and older individuals typically have comorbid conditions and are in poorer health than younger persons.[19] Because the symptoms of CMI are nonspecific, and there is no “gold-standard” test for diagnosis, treatment is often delayed leading to poorer outcomes.[5,18] This is especially true for older persons and the elderly. Endovascular therapy is the treatment of choice for CMI, and is associated with less postoperative morbidity than open surgery, not all patients are candidates for endovascular treatment and open surgery in the elderly is associated with a higher risk of postoperative complications.[5,18] A comparison of outcomes of open surgical and endovascular revascularization for CMI found that the 2 methods had comparable perioperative outcomes, endovascular therapy had a higher technical failure rate while open surgery had high early and late technical success rates.[20] Notably, our results showed that the effect of frailty with respect to poorer outcomes was greater in patients undergoing open surgery than endovascular therapy. The greater impact of frailty on complications in surgical compared to endovascular revascularization for CMI may be attributed to the invasiveness of open surgery with a longer recovery periods and higher risks of complications. This highlights the need for a careful assessment of a patient’s health and frailty status when choosing the most suitable revascularization approach, aiming to balance the benefits of the procedure against the potential risks for frail patients. There are a number of tools that have been developed to assess frailty, and while they all have adequate predictive value for determining frailty, some are better for specific circumstances than others.[21,22] Yin et al[22] evaluated preoperative frailty with the 5-item FRAIL questionnaire, the 54-item frailty index (FI), and the 9-item Clinical Frailty Scale (CFS). The incidences of frailty as assessed using FRAIL, FI, and CFS were 44%, 32%, and 37%, respectively. Notably, the 90-day mortality and prolonged LOS were significantly higher for frail patients, compared to non-frail patients, regardless of which tool was used. However, the CFS and FI predicted 90-da mortality better than FRAIL, and the CFS predicted prolonged LOS better than FRAIL. A recent study compared the CFS and HFRS for predicting long-term survival of critically ill patients hospitalized in the intensive care unit.[21] The study included 7001 patients, and the results showed that both tools independently predicted 1-year survival; however, the CFS had slightly better predictive value than the HFRS (AUC 0.66 vs. 0.63, P < .0001). We choose to use the HFRS in our study because it is based on administrative claim codes, which makes it very suitable for retrospective studies and those using claims-based databases. The current study is unique because it examined the impact of frailty on the outcomes of treatment for CMI, a topic few studies in the literature have addressed. However, many studies have examined the impact of frailty on other types of surgery. For example, Suzuki et al[23] studied the outcomes of patients ≥80 years old undergoing colectomy. Patients were categorized as high-risk and low-risk of frailty using the modified Frailty Index 11 (mFI-11). Frail patients were significantly more likely to have postoperative complications, longer LOS, and discharge to a location other than home than patients at low-risk of frailty; however, comprehensive preoperative management reduced the incidence of postoperative complications and other adverse outcomes in frail patients. Gu et al[24] studied the value of the FI for predicting postoperative morbidity in elderly patients undergoing gastrointestinal surgery. The study included 246 patients with median age of 72 years, of whom 19% were classified as frail. Frailty was associated with adverse perioperative outcomes, and was a more reliable predictor of 30-day mortality than age and American Society of Anesthesiologists grade. Some studies have examined the impact of frailty on outcomes of revascularization for different conditions. Gonzalez et al[25] studied the impact of frailty, assessed with the mFI, on male veterans undergoing first time revascularization, open or endovascular. Regardless of the method of revascularization, frailty independently predicted major amputation (aHR = 2.16) and mortality (aHR = 2.62). Similarly, Al-Damluji et al[26] reported that frailty was predictive of mortality in patients undergoing revascularization (endovascular and open) for lower extremity peripheral artery disease. In our study, higher frailty risk was associated with higher in-hospital mortality in patients over 60 years but not in those <60 years old. Younger individuals, even those classified as frail, might still have stronger recovery abilities allowing them to withstand the rigors of revascularization procedures more effectively. This suggests that the intersection of frailty and aging significantly influences mortality risks, emphasizing the need for tailored approaches in managing older, frail patients undergoing such interventions. Not surprisingly, our results showed that frailty was associated with higher hospital costs, a finding that is consistent with the results of other studies. For example, Hadaya et al[27] reported that frailty was associated with worse outcomes and higher costs in patients undergoing elective colectomy. 4.1. Strengths and limitations Using the NIS dataset offers a detailed analysis of how frailty affects CMI revascularization outcomes, significantly contributing to the field with its broad and nationally representative data. This study used the validated and widely accepted HFRS for precise patient stratification, taking into account a wide range of demographic, clinical, and hospital-level confounders. The analysis also spans a wide array of in-hospital outcomes, presenting a thorough understanding of the effects of frailty on outcomes. However, its retrospective nature might limit the accuracy of data, and the focus on short-term, in-hospital outcomes leaves long-term effects unexplored. The reliance on ICD codes for identifying conditions could introduce bias, and there may be unmeasured confounders affecting the results. Despite their importance, certain preoperative and intraoperative factors, including medications, laboratory data, and operative duration, were not evaluated in this study. Additionally, its findings, based on US data, may not fully translate to non-US healthcare settings, suggesting a need for cautious application of these insights globally. 4.2. Conclusion This study demonstrates that in the older US population, frailty, as assessed by the HFRS, is a crucial predictor of poor in-hospital outcomes in patients undergoing revascularization for CMI. Specifically, patients with higher HFRS scores have significantly increased risks of adverse outcomes, such as in-hospital mortality, prolonged LOS, and various complications, leading to higher hospital costs. These findings highlight the importance of considering frailty in preoperative risk stratification and patient management in CMI revascularization, and that healthcare providers should integrate frailty assessment into preoperative evaluations to identify patients at higher risk of adverse outcomes. A proactive approach to frailty may allow the development of tailored perioperative care plans, including enhanced monitoring, targeted interventions to mitigate frailty-related risks, and informed decision-making regarding the surgical options and postoperative care. Author contributions Data curation: Chien-Min Lai, Hui-Pu Liu, Wen-Yen Chang. Formal analysis: Wen-Yen Chang. Investigation: Wen-Yen Chang. Methodology: Chien-Min Lai. Project administration: Hui-Pu Liu. Supervision: Hui-Pu Liu. Validation: Chien-Min Lai. Visualization: Chien-Min Lai. Writing – original draft: Hui-Pu Liu. Supplementary Material