Modified Frailty Index Can Be Used to Predict Adverse Outcomes and Mortality after Lower Extremity Bypass Surgery

Accepted Manuscript Modified Frailty Index Can be used to Predict Adverse Outcomes and Mortality After Lower Extremity Bypass Surgery Tarik Ali, MD, Erik Lehman, MS, Faisal Aziz, MD, FACS PII:

S0890-5096(17)30902-0

DOI:

10.1016/j.avsg.2017.07.007

Reference:

AVSG 3498

To appear in:

Annals of Vascular Surgery

Received Date: 24 March 2017 Revised Date:

10 July 2017

Accepted Date: 11 July 2017

Please cite this article as: Ali T, Lehman E, Aziz F, Modified Frailty Index Can be used to Predict Adverse Outcomes and Mortality After Lower Extremity Bypass Surgery, Annals of Vascular Surgery (2017), doi: 10.1016/j.avsg.2017.07.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

Modified Frailty Index Can be used to Predict Adverse Outcomes and

2

Mortality After Lower Extremity Bypass Surgery

3

Tarik Ali, MD1, Erik Lehman, MS2, Faisal Aziz, MD, FACS1.

RI PT

1

4

5 1

7

Pennsylvania State University College of Medicine.

SC

6

8

2

9

of Medicine.

M AN U

Division of Vascular Surgery, Penn State Heart and Vascular Institute,

Department of Public Health Sciences, Pennsylvania State University, College

11

12

TE D

10

Correspondence to:

14

Faisal Aziz, MD, FACS.

15

Penn State Milton S. Hershey Medical Center, 500 University Drive

16

Mail Code H053, Hershey, PA 17033

17

Tel: 717-531-8898, Fax: 717-531-4151, E-Mail: [email protected]

AC C

EP

13

18

Presented at the Annual Meeting, Vascular and Endovascular Surgery Society,

19

2017

20

1

ACCEPTED MANUSCRIPT 2

21

Modified Frailty Index Can be used to Predict Adverse Outcomes and

22

Mortality After Lower Extremity Bypass Surgery

Objectives

24

Frailty has been increasingly used as a prognostic indicator for various surgical

25

operations. Patients with peripheral arterial disease represent a cohort of

26

population with advanced medical comorbidities. The aim of this study is to

27

correlate the postoperative outcomes after lower extremity bypass surgery with

28

pre operative modified frailty index.

29

Methods

30

Using 2010 American College of Surgeons National Surgical Quality

31

Improvement Program (ACS-NSQIP) database, patients undergoing infrainguinal

32

arterial bypass surgery were identified. Modified frailty index (mFI) with 11

33

variables, based on the Canadian Study of Health and Aging Frailty Index was

34

utilized. Based on mFI score, the patients were divided into four groups: Group

35

1 (mFI score: 0-0.09), Group 2 (mFI score: 0.18-0.27), Group 3 (mFI score: 0.36-

36

0.45), and Group 4 (mFI score: 0.54-0.63). A bivariate and multivariate analysis

37

was done using logistic regression analysis.

38

Results

39

A total of 4,704 patients (64% Males, 36% Females) underwent infrainguinal

40

arterial bypass. Mean age: 67.9 ± 11.7 years. Distribution of patients based on

41

mFI was as follows: Group 1: 14.6%, Group 2: 55.9%, Group 3: 26.9% and

42

Group 4: 2.6%. Increase in mFI was associated with higher mortality rates.

AC C

EP

TE D

M AN U

SC

RI PT

23

2

ACCEPTED MANUSCRIPT 3

Incidence of mortality for Group 1 was 0.6%, for Group 2, it was 1.4%, for Group

44

3, it was 4% and for Group 4, it was 7.4%. Likewise, the incidence of other

45

postoperative complications such as myocardial infarction, stroke, progressive

46

renal failure and graft failure was significantly high among patients with high mFI

47

scores. Following factors were associated with increased risk of mortality: high

48

mFI score, black race, dialysis dependency, post operative renal insufficiency,

49

myocardial infarction and post operative acute renal failure.

50

Conclusions

51

This study demonstrates that the mFI can be used as a valuable tool to identify

52

patients at a higher risk for developing postoperative complications after lower

53

extremity revascularization. For patients with mFI score of 0.54-0.63, the risk of

54

mortality and complications increases significantly. mFI can be used as a useful

55

screening tool to identify patients who are at a high risk for developing

56

complications.

59

60

SC

M AN U

TE D

EP

58

AC C

57

RI PT

43

61

62

63 3

ACCEPTED MANUSCRIPT 4

64

65

Introduction: While surgeons have realized that frailty is an important determinant of surgical outcomes, exact definition of frailty remains unknown.

Review of

67

geriatric literature1,2 fails to identify a consensus statement on definition of frailty.

68

Some authors suggest muscle size measurements3 to be the surrogates of frailty,

69

while others stress the importance of gait, balance and hand grip strength4-6 in

70

determining frailty of an individual.

71

syndrome resulting from multisystem impairments. These impairments include

72

sacropenia, functional decline and neuroendocrine dysregulation. In presence of

73

such impairments, clinical outcomes are considered to be worse. Patients with

74

combination of such impairments are at an even higher risk for development of

75

complications.

76

electively, it is imperative that identification of high-risk patients can potentially

77

prevent postoperative complications.

78

used to determine frailty. Elderly patients are more likely to be frail as compared

79

to the general population. According to the U.S. Census Bureau8, the geriatric

80

population is the fastest growing segment of the U.S. population. It is projected

81

that by year 2030 one in five Americans will be more than 65 years of age8.

82

Elderly patients are considered frail and are deemed to be a high risk for

83

developing postoperative complications. Health care providers are faced with the

84

challenge of providing a reliable and cost effective healthcare to an aging

85

population. Frailty has been associated with poor outcomes after major surgical

86

procedures, including cardiac, colo-rectal, general surgery and urologic

87

operations4-6,9,10. Patients with peripheral arterial disease tend to have severe

RI PT

66

M AN U

SC

Abellan et al7 define frailty as a clinical

TE D

Since the majority of the surgical operations are performed

AC C

EP

Advanced age is just one of the criteria

4

ACCEPTED MANUSCRIPT 5

88

coexisting comorbidities, which place them at a high risk for developing

89

postoperative complications. Traditionally age and coexisting comorbidities have been employed in

91

addition to other tools to predict outcomes for patients undergoing an operation.

92

In recent years there has been a lot of interest in Frailty in the surgical patient

93

and its implication on outcomes. A simplified form of the original Frailty index

94

from the Canadian Study of Health and Aging has been established as the

95

modified frailty index (mFI). In this study, we sought to apply the mFI to below the

96

inguinal ligament vascular bypass surgery in an effort to predict outcomes and

97

validate it as a low cost, noninvasive tool to help clinicians better screen and

98

inform patients.

M AN U

SC

RI PT

90

102

103

104

105

EP

101

AC C

100

TE D

99

106

107

5

ACCEPTED MANUSCRIPT 6

Methods:

109

Dataset: The American College of Surgeons National Surgical Quality

110

Improvement Program (ACS-NSQIP)11 Participant User Files (PUF) were

111

reviewed for the year 2010.

112

information and is compliant with Health Insurance Portability and Accountability

113

Act (HIPPA). Therefore, no institutional review board (IRB) approval or patient

114

consent was required. Methods used to extract data from this database are well

115

described in the literature12-15.

116

responsible for collecting all data points and excluding patients with incomplete

117

data. ACS performs regular audits of the dataset and the outcomes based on

118

NSQIP database are shown to be highly reliable14. This database provides a

119

comprehensive list of pre-operative comorbidities, laboratory values, intra-

120

operative variables and postoperative outcomes. Patients younger than 16 years

121

were not included and patients older than 89 years of age were coded as 90+ to

122

protect patient confidentiality.

123

Patients: Using Current Procedural Terminology (CPT) codes (37220-37235);

124

patients who had undergone infrainguinal arterial bypass surgery during 2010

125

were identified in the NSQIP-PUF file.

126

Modified Frailty Index: Modified Frailty index (mFI) was calculated using 11

127

variables, based on the Canadian Study of Health and Aging Frailty Index (CSHA

128

FI)2,16 and validated in the ACS- NSQIP database. Table I depicts the variables

129

used by CSHA FI and the corresponding variables used in ACS-NSQIP

130

database. The modified Frailty index (mFI) for a patient is calculated by adding

131

one point for the presence of each of these variables divided by 11. This

RI PT

108

SC

This dataset contains patient de-identified

AC C

EP

TE D

M AN U

A trained clinical nurse at each institution is

6

ACCEPTED MANUSCRIPT 7

produces a number between 0 and 1. Based on mFI score, the patients were

133

divided into four groups: Group 1 (mFI score: 0-0.09), Group 2 (mFI score: 0.18-

134

0.27), Group 3 (mFI score: 0.36-0.45), and Group 4 (mFI score: 0.54-0.63).

135

Outcomes: ACS-NSQIP database provides information for any complications

136

within 30 days after surgery.

137

primary outcome and post op myocardial infarction, stroke, renal failure and graft

138

failure were chosen as secondary outcomes.

139

Statistical Analysis:

140

All variables were initially summarized with frequencies and percentages or

141

means, medians, and standard deviations. Binomial logistic regression was used

142

to determine any bivariate associations between the four-group mFI variables

143

and post-operative outcome variables and between other potential covariates

144

and 30-day mortality.

145

direction of any significant associations.

146

independent variables from the bivariate analysis of 30-day mortality along with

147

demographic variables (age, gender, race, and BMI) were then used in a process

148

of stepwise, forward, and backward selection methods to find the group of

149

variables collectively that were most significantly associated with 30-day

150

readmission in a multivariable logistic regression model.

Prior to the

151

multivariable

checked

152

multicollinearity with one another using variance inflation factor (VIF) statistics.

153

With so many variables and a large sample size, a more stringent entry criteria of

154

p<0.05 and a stay criteria of p<0.05 were used for the process of variable

155

selection to be more conservative. The fit of the final model was checked using

RI PT

132

M AN U

SC

For this analysis mortality was chosen as the

The statistically significant (p<0.05)

AC C

EP

TE D

Odds ratios were used to quantify the magnitude and

analysis,

the

independent

variables

were

for

7

ACCEPTED MANUSCRIPT 8

the Hosmer and Lemeshow goodness-of-fit test (p=0.6309).

157

(c=0.887) shows adequate prediction strength of the final model. All analyses

158

were performed using SAS version 9.4 (SAS Institute, Cary, NC).

159

160

SC

161

M AN U

162

163

164

169

170

171

EP

168

AC C

167

TE D

165

166

The c-statistic

RI PT

156

172

173

8

ACCEPTED MANUSCRIPT 9

174

Results:

175

Demographics:

176

36% Females), who underwent an infrainguinal arterial bypass in 2010. Mean

177

age was 67.9 ± 11.7 years (Range 19 to >90 years). Race distribution was as

178

follows: 74% White, 16% African American, 10% unknown. Mean BMI was 27.7

179

± 6.4 (BMI<25: 1,637 patients (35.5%), BMI 25-29: 1,593 patients (34.5%) and

180

BMI 30-40: 1,216 patients (26.4%) and BMI ≥ 40: 165 patients (3.6%). Incidence

181

of pre-operative variables was as fellows: hypertension (3981: 84.6%), history of

182

rest pain (2,205: 46.9%), history of transient ischemic attack (302: 6.4%), history

183

of stroke (364 (7.7%), American Society of Anesthesiology (ASA) Score 1 (No

184

disturbance/mild disturbance): 303 (6.4%), ASA Score 2 (Severe Disturbance):

185

3,459 (73.6%) and ASA Score 3 and 4 (Life threatening emergency/ moribund

186

condition): 941 (20%), Active smoking status (1,915 patients: 41%), emergency

187

operation (258 (5.5%) and end-stage renal disease on hemodialysis (336: 7%).

188

Post Operative Complications: The distribution of postoperative complications

189

was as fellows: superficial surgical site infection (SSI): 356 (7.6%), deep wound

190

infection: 110 (2.3%), stroke: 28 (0.6%), graft failure: 206 (4.4%), deep venous

191

thrombosis: 47 (1%), return to operating room: 773 (16.4%), acute renal failure:

192

38 (0.8%) and mortality: 100 (2.1%).

193

Modified Frailty Index: Modified Frailty Index was calculated for all patients and

194

distribution of mFI was as follows:

195

mFI = 0/11 = 0

196

mFI = 1/11 = 0.09 (Number of Patients = 564 (11.9%)

197

mFI = 2/11 = 0.18 (Number of Patients = 1,177 (25.02%)

AC C

EP

TE D

M AN U

SC

RI PT

ACS-NSQIP database identified 4,704 patients (64% Males,

(Number of Patients = 121 (2.6%)

9

ACCEPTED MANUSCRIPT

mFI = 3/11 = 0.27 (Number of Patients = 1,456 (30.95%)

199

mFI = 4/11 = 0.36 (Number of Patients = 914 (19.43%)

200

mFI = 5/11 = 0.45 (Number of Patients = 351 (7.46%)

201

mFI = 6/11 = 0.54 (Number of Patients = 92 (1.96%)

202

mFI = 7/11 = 0.63 (Number of Patients = 29 (0.62%)

203

mFI= 8/11 = 0.72 (Number of Patients = 0, (0%)

204

mFI = 9/11 = 0.81 (Number of Patients = 0, (0%)

205

mFI = 10/11 = 0.90 (Number of Patients = 0, (0%)

206

mFI = 11/11 = 1

207

M AN U

(Number of Patients = 0, (0%)

SC

198

RI PT

10

When grouped together, as described in methods section, the distribution of mFI

209

scores was as follows:

210

Group 1 (mFI score: 0-0.09): 685 Patients (14.56%)

211

Group 2 (mFI score: 0.18-0.27): 2,633 Patients (55.97%)

212

Group 3 (mFI score: 0.36-0.45): 1,265 Patients (26.89%)

213

Group 4 (mFI score: 0.54-0.63): 121 Patients (2.57%)

214

Outcomes: Thirty day mortality was chosen as primary outcome. Increase in

215

mFI was associated with higher mortality rates. Incidence of mortality for Group 1

216

was 0.6%, for Group 2, it was 1.4%, for Group 3, it was 4% and for Group 4, it

217

was 7.4%. The incidence of MI for Group 1 was 1%, for Group 2, it was 2%, for

218

Group 3, it was 3.1% and for Group 4, it was 4.2%. The incidence of stroke

219

among Group 1 patients was 0.6%, for Group 2, it was 0.5% , for Group 3, it was

220

0.6% and for Group 4, it was 1.7%.

221

insufficiency, the incidence among Group 1 patients was 0.6%, among Group 2

AC C

EP

TE D

208

For postoperative progressive renal

10

ACCEPTED MANUSCRIPT 11

patients, it was 0.6%, among Group 3 patients, it was 0.6 and among Group 4

223

patients, it was 0.8%. The incidence of graft failure for Group 1 patients was

224

4.8%, for Group 2 patients, it was 4.3%, for Group 3 patients, it was 4.1% and for

225

Group 4 patients, it was 7.4%. In regards to any occurrence for any of these

226

complications, the incidence among Group 1 patients was 6.9%, for Group 2

227

patients, it was 8.1% , for Group 3 patients, it was 11.1% and for Group 4

228

patients, it was 19% (Table II).

229

The Univariate analysis identified following independent variables to be

230

associated with higher risk of mortality: Age >80 years , higher ASA score (4-5),

231

transfer from nursing home, impaired sensorium, dependent functional status ,

232

diabetes, history of COPD, history of congestive heart failure, history of previous

233

coronary stenting history of angina, history of rest, emergency operation, dialysis

234

dependency, increased length of hospital stay, post operative progressive renal

235

insufficiency, post operative stroke, post operative myocardial infarction, sepsis,

236

return to the operating room, increased length of hospital stay and acute renal

237

failure (Table III).

238

Multivariant analysis identified following factors to be associated with increased

239

risk of mortality: mFI 0.18-0.27 vs. 0-0.09, mFI 0.36-0.45 vs. 0-0.09 , mFI 0.54-

240

0.63 vs. 0-0.09, age 60-70 vs. <60, age 70-80 vs. <60, age >80 vs. <60, black

241

race vs. white race, other races vs. white race, dialysis dependency, post

242

operative progressive renal insufficiency, post operative myocardial infraction

243

and post operative acute renal failure (Table IV).

AC C

EP

TE D

M AN U

SC

RI PT

222

244 245

11

ACCEPTED MANUSCRIPT 12

246

Discussion: Our study shows that variables described in ACS-NSQIP database can be

248

used to calculate a mFI and that a higher mFI score for patients undergoing

249

lower extremity revascularization surgery accurately predicts a higher 30-day

250

mortality and myocardial infarction. For years, vascular surgeons have been

251

relying on a variety of factors and measures to assess a patient’s ability to

252

withstand the stressors of surgery. Routine cardiac testing and pulmonary

253

assessment is regularly performed in anticipation for major surgery, however,

254

rarely are factors such as functional status, congestive heart failure and history of

255

previous cerebral ischemic events are used in such assessments. Our study

256

shows successful use of the mFI scoring system to predict major postoperative

257

complications after lower extremity bypass surgery. It is worth mentioning that

258

frailty is fundamentally different from existence of several comorbidities, as the

259

variables used to calculate the mFI encompass mental and social factors in

260

addition to relying on markers for physical dysfunction. Our study has shown that

261

it can be a reliable and an important low cost, noninvasive tool to aid in the

262

clinical decision making. It shows a noticeable increase in morbidity and mortality

263

with an increasing m-FI. Patients with a mFI score of greater than 0.54 (i.e.

264

presence of 6 or more variables), the 30-day mortality after lower extremity

265

surgical bypass can be as high as 7.4% (OR 13.7, CI 4.14-45.2, p<0.001) and

266

the risk of myocardial infarction is around 4% (OR 4.2, CI 1.3-13.4, p=0.013).

AC C

EP

TE D

M AN U

SC

RI PT

247

267

12

ACCEPTED MANUSCRIPT 13

With a better understanding of the impact of pre-operative factors on the

269

post-operative outcomes, there has been an increasing focus of attention on

270

determining the factors associated with postoperative morbidity and mortality.

271

Historically, surgeons have focused on cardiac causes of morbidity and mortality

272

after the surgical operations. Poldermans et al17 have shown that patients with a

273

positive stress test before a major vascular surgery operations are at a high risk

274

for developing post operative cardiac complications.

275

have identified advanced age, pre-existing cardiac and renal diseases, high ASA

276

scores and the use of general anesthesia to be associated with a significantly

277

high risk of developing post operative complications after lower extremity bypass

278

surgery. The eleven variables used to calculate a mFI score represent physical,

279

mental, social assessment and a combination of these factors can accurately

280

reflect the outcomes after elective operations. Several authors4-6,9,19 have used

281

prospective data to calculate frailty scores for surgical patients.

282

studies10,20,21 have used retrospective methodology to determine frailty scores for

283

surgical patients and predicted the incidence of postoperative complications

284

based on mFI scores. Veanovich et al22 have shown that patients with increased

285

frailty

286

otolaryngology, general, orthopedic, neurosurgical, thoracic, gynecological and

287

urologic procedures. Likewise, Hewitt et al23 have shown that frail patients, who

288

were more than 65 years of age undergoing emergent general surgical

289

procedures had increased incidence of postoperative mortality.

RI PT

268

EP

TE D

M AN U

SC

Likewise, Kehlet et al18

had

AC C

scores

an

increased

mortality

after

undergoing

Some

cardiac,

290

Patients with PAD represent a cohort of patient population, which often

291

has multiple comorbidities making it difficult to access their true physiologic 13

ACCEPTED MANUSCRIPT 14

292

reserve. Outcomes of any intervention for patients with PAD are dependent on

293

anatomic, physiologic and disease severity factors24.

294

surgeons have considered surgical bypasses to have the advantage over

295

endovascular interventions in terms of long term patency and avoiding re-

296

interventions25.

297

surgical bypass should take into account the possibility of all the postoperative

298

complications. As stated above, surgeons in the past have relied heavily on using

299

age and cardiac history as predictors for postoperative outcomes, however

300

applying mFI scores to patients undergoing arterial bypass for PAD provides a

301

better assessment tool to predict postoperative complications. A recent study of

302

patients undergoing lower extremity bypass surgery26 showed that elderly

303

patients (age >80 years) had no increased risk of postoperative complications as

304

compared to those patients who were younger than 80 years of age. Our data

305

clearly shows that age more than 80 years is associated with significantly high

306

risk for mortality after lower extremity bypass surgery.

307

previous major amputation and female gender have been associated with major

308

adverse limb events after lower extremity bypass surgery27. Our analysis shows

309

that the mFI can be used as a useful tool to accurately predict postoperative

310

complications.

311

vascular surgery operations also showed that preoperative mFI could accurately

312

predict postoperative mortality. Interestingly, this analysis showed mFI to be a

313

better predictor of postoperative mortality than other variables for open

314

operations. Our study only consists of open lower extremity bypasses and our

315

results correlate with the findings of Ehlert et al.

RI PT

Traditionally, vascular

Higher ASA scores,

AC C

EP

TE D

M AN U

SC

However, the decision making process before performing a

A recent analysis28 of all the patients undergoing different

The disparity in outcomes 14

ACCEPTED MANUSCRIPT 15

between patients with different mFI scores should be recognized and the

317

informed consent process should include a frank discussion about the risk of

318

complications based on mFI scores. Our data clearly shows that frailty may be

319

present in the younger patient population and relying solely on age to predict

320

postoperative complications may not be a reliable strategy.

RI PT

316

Patients presenting with severe PAD have some degree of frailty at

322

baseline. Most surgeons would recommend surgical bypasses for PAD only for

323

the select group of patients who have either critical limb ischemia or severe,

324

disabling claudication. While the purpose of revascularization in patients with

325

critical limb ischemia is limb salvage, the goal of surgical bypass in claudicants is

326

improvement in quality of life. The decision to proceed with surgical bypass

327

should be made after careful evaluation of all risks and benefits. Many factors

328

play an important role in determining the balance between risks and benefits of

329

an operation. This data serves to add one more data point i.e. mFI, in helping

330

vascular surgeons determining the risk/benefit ratio.

331

surgeons would still recommend revascularization for patients with severe tissue

332

loss, our analysis clearly shows that it may not be advisable to offer an open

333

revascularization procedure on a claudicant with high mFI. Our data should be

334

interpreted with caution and operative decisions should be made after careful

335

consideration of all risk factors and tailoring the outcomes to individual patients.

While most vascular

AC C

EP

TE D

M AN U

SC

321

336

Our analysis is limited by the inherent nature of any retrospective study.

337

The NSQIP database does not contain the variables as gait speed or grip

338

strength, which can provide additional information while calculating frailty. The

15

ACCEPTED MANUSCRIPT 16

database is also limited to 30-day outcomes making it difficult to evaluate long-

340

term progress of these patients. The findings are limited to only those hospitals,

341

which were enrolled in the NSQIP database and may not be applicable to all

342

patients. Despite these limitations, The NSQIP provides us with a large and

343

diverse sample size and the statistics performed on such a large sample do

344

provide results with statistical significance. It also provides us access to a

345

diversity of surgeons and institutions where patients are cared.

SC

RI PT

339

To summarize, this study demonstrates that the m-FI does identify

347

patients at a higher risk for morbidity and mortality. It shows that for a m-FI 0.54-

348

0.63 the risk of 30-day mortality and post-operative complications increases

349

significantly. When used appropriately, the mFI can be a valuable, low cost,

350

noninvasive tool in accessing and stratifying patients preoperatively.

354

355

356

TE D

353

EP

352

AC C

351

M AN U

346

357

358

359 16

ACCEPTED MANUSCRIPT 17

360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393

References

394 395 396 397 398 399 400 401 402 403 404 405

. 12. Khuri SF, Daley J, Henderson W, Hur K, Gibbs JO, Barbour G, et al. Risk adjustment of the postoperative mortality rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997;185:315-27. 13. Bilimoria KY, Cohen ME, Ingraham AM, Bentrem DJ, Richards K, Hall BL, et al. Effect of postdischarge morbidity and mortality on comparisons of hospital surgical quality. Ann Surg 2010;252:183-90. 14. Shiloach M, Frencher SK, Jr., Steeger JE, Rowell KS, Bartzokis K, Tomeh MG, et al. Toward robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg 2010;210:6-16.

AC C

EP

TE D

M AN U

SC

RI PT

1. Morley JE, Vellas B, van Kan GA, Anker SD, Bauer JM, Bernabei R, et al. Frailty consensus: a call to action. J Am Med Dir Assoc 2013;14:392-7. 2. Rockwood K, Andrew M, Mitnitski A. A comparison of two approaches to measuring frailty in elderly people. J Gerontol A Biol Sci Med Sci 2007;62:73843. 3. Sheetz KH, Zhao L, Holcombe SA, Wang SC, Reddy RM, Lin J, et al. Decreased core muscle size is associated with worse patient survival following esophagectomy for cancer. Dis Esophagus 2013;26:716-22. 4. Makary MA, Segev DL, Pronovost PJ, Syin D, Bandeen-Roche K, Patel P, et al. Frailty as a predictor of surgical outcomes in older patients. J Am Coll Surg 2010;210:901-8. 5. Robinson TN, Wu DS, Pointer L, Dunn CL, Cleveland JC, Jr., Moss M. Simple frailty score predicts postoperative complications across surgical specialties. Am J Surg 2013;206:544-50. 6. Revenig LM, Canter DJ, Taylor MD, Tai C, Sweeney JF, Sarmiento JM, et al. Too frail for surgery? Initial results of a large multidisciplinary prospective study examining preoperative variables predictive of poor surgical outcomes. J Am Coll Surg 2013;217:665-70 e1. 7. Abellan van Kan G, Rolland Y, Houles M, Gillette-Guyonnet S, Soto M, Vellas B. The assessment of frailty in older adults. Clin Geriatr Med 2010;26:27586. 8. Colby S OJ. Projections of the Size and Composition of the U.S. Population: 2014 to 2060, Current Population Reports. US Census Bureau 2014:P25-1143. 9. Lee DH, Buth KJ, Martin BJ, Yip AM, Hirsch GM. Frail patients are at increased risk for mortality and prolonged institutional care after cardiac surgery. Circulation 2010;121:973-8. 10. Adams P, Ghanem T, Stachler R, Hall F, Velanovich V, Rubinfeld I. Frailty as a predictor of morbidity and mortality in inpatient head and neck surgery. JAMA Otolaryngol Head Neck Surg 2013;139:783-9. 11. Improvement ACoSNSQ, at: PA. https://www.facs.org/qualityprograms/acs-nsqip. Accessed January 10, 2017

17

ACCEPTED MANUSCRIPT 18

15. Khuri SF, Daley J, Henderson W, Barbour G, Lowry P, Irvin G, et al. The National Veterans Administration Surgical Risk Study: risk adjustment for the comparative assessment of the quality of surgical care. J Am Coll Surg 1995;180:519-31. 16. Mitnitski AB, Mogilner AJ, Rockwood K. Accumulation of deficits as a proxy measure of aging. ScientificWorldJournal 2001;1:323-36. 17. Poldermans D, Fioretti PM, Forster T, Thomson IR, Boersma E, el-Said EM, et al. Dobutamine stress echocardiography for assessment of perioperative cardiac risk in patients undergoing major vascular surgery. Circulation 1993;87:1506-12. 18. Kehlet M, Jensen LP, Schroeder TV. Risk Factors for Complications after Peripheral Vascular Surgery in 3,202 Patient Procedures. Ann Vasc Surg 2016;36:13-21. 19. Kim SW, Han HS, Jung HW, Kim KI, Hwang DW, Kang SB, et al. Multidimensional frailty score for the prediction of postoperative mortality risk. JAMA Surg 2014;149:633-40. 20. Etzioni DA, Liu JH, Maggard MA, Ko CY. The aging population and its impact on the surgery workforce. Ann Surg 2003;238:170-7. 21. Karam J, Tsiouris A, Shepard A, Velanovich V, Rubinfeld I. Simplified frailty index to predict adverse outcomes and mortality in vascular surgery patients. Ann Vasc Surg 2013;27:904-8. 22. Velanovich V, Antoine H, Swartz A, Peters D, Rubinfeld I. Accumulating deficits model of frailty and postoperative mortality and morbidity: its application to a national database. J Surg Res 2013;183:104-10. 23. Hewitt J, Moug SJ, Middleton M, Chakrabarti M, Stechman MJ, McCarthy K, et al. Prevalence of frailty and its association with mortality in general surgery. Am J Surg 2015;209:254-9. 24. Lo RC, Darling J, Bensley RP, Giles KA, Dahlberg SE, Hamdan AD, et al. Outcomes following infrapopliteal angioplasty for critical limb ischemia. J Vasc Surg 2013;57:1455-63; discussion 63-4. 25. Feinglass J, Pearce WH, Martin GJ, Gibbs J, Cowper D, Sorensen M, et al. Postoperative and late survival outcomes after major amputation: findings from the Department of Veterans Affairs National Surgical Quality Improvement Program. Surgery 2001;130:21-9. 26. Shirasu T, Hoshina K, Nishiyama A, Akagi D, Miyahara T, Yamamoto K, et al. Favorable outcomes of very elderly patients with critical limb ischemia who undergo distal bypass surgery. J Vasc Surg 2016;63:377-84. 27. Brothers TE, Zhang J, Mauldin PD, Tonnessen BH, Robison JG, Vallabhaneni R, et al. Predicting outcomes for infrapopliteal limb-threatening ischemia using the Society for Vascular Surgery Vascular Quality Initiative. J Vasc Surg 2016;63:114-24 e5.

447 448 449

28. Ehlert BA, Najafian A, Orion KC, Malas MB, Black JH 3rd, Abularrage CJ. Validation of a modified frailty index to predict mortality in vascular surgery patients. J Vasc Surg 2016 Jun;63(6):1595-1601.

AC C

EP

TE D

M AN U

SC

RI PT

406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446

450

18

ACCEPTED MANUSCRIPT

Table I: Variables Used to Calculate mFI by the Canadian Study of Health and Aging (CSHA) and the Corresponding Variables Identified in the American College of Surgeons National Surgical Quality Improvement Program (ACSNSQIP).

Impaired Sensorium

Pre-Operative Functional Status: - Independent - Partially dependent - Totally dependent

M AN U

History relevant to cognitive impairment or loss Problems with bathing Problems with dressing Problems with cooking Problems with personal grooming Problems with going out alone Cardiac Disease

Congestive Heart Failure

TE D

Myocardial Infarction

RI PT

ACS-NSQIP

Clouding or Delirium

Arterial hypertension History of diabetes mellitus

EP

Chronic/Acute lung disease

Cereberovascular problems History of Stroke

AC C

Pre-Operative Comorbidities

CSHA

SC

Category of Variables Functional and Cognitive Impairment

Decreased peripheral pulses

- Previous percutaneous coronary intervention or cardiac surgery - History of angina <1 month before surgery Congestive heart failure <1 month before surgery History of myocardial infarction <6 months before surgery Hypertension requiring medication Diabetes Mellitus - Insulin dependent - Non-insulin dependent History of severe COPD Current pneumonia History of transient ischemic attack Cerebrovascular accident or stroke with neurologic deficit - History of revascularization or amputation for peripheral vascular disease - Rest pain or gangrene

ACCEPTED MANUSCRIPT

Table II

AC C

EP

TE D

M AN U

SC

RI PT

Modified Frailty N (%) with Outcome (N=4704) Index Outcome OR (95% CI)* P-value* 30-day mortality 0.0 – 0.09 4 (0.6) Reference <0.001 0.18 – 0.27 37 (1.4) 2.4 (0.86, 6.83) 0.36 – 0.45 50 (4.0) 7 (2.52, 19.48) 0.54 – 0.63 9 (7.4) 13.7 (4.14, 45.18) Post-operative myocardial infarction 0.0 – 0.09 7 (1.0) Reference 0.013 0.18 – 0.27 53 (2.0) 2 (0.90, 4.39) 0.36 – 0.45 39 (3.1) 3.1 (1.37, 6.92) 0.54 – 0.63 5 (4.1) 4.2 (1.30, 13.37) Post-operative stroke (CVA) 0.0 – 0.09 4 (0.6) Reference 0.516 0.18 – 0.27 14 (0.5) 0.9 (0.29,2.77) 0.36 – 0.45 8 (0.6) 1.1 (0.33,3.611) 0.54 – 0.63 2 (1.7) 2.9 (0.52,15.79) Post-operative progressive renal 0.0 – 0.09 4 (0.6) Reference 0.985 failure 0.18 – 0.27 16 (0.6) 1 (0.35,3.124) 0.36 – 0.45 7 (0.6) 1 (0.28,3.25) 0.54 – 0.63 1 (0.8) 1.4 (0.16,12.80) Graft/prosthesis/flap failure 0.0 – 0.09 33 (4.8) Reference 0.354 0.18 – 0.27 112 (4.3) 0.9 (0.59,1.31) 0.36 – 0.45 52 (4.1) 0.9 (0.54,1.32) 0.54 – 0.63 9 (7.4) 1.6 (0.74,3.41) Any occurrence (of the above 0.0 – 0.09 47 (6.9) Reference < 0.001 outcomes) 0.18 – 0.27 213 (8.1) 1.2 (0.86,1.66) 0.36 – 0.45 140 (11.1) 1.7 (1.2,2.38) 0.54 – 0.63 23 (19.0) 3.2 (1.85,5.48) * All odds ratios and p-values are from binomial logistic regression modeling the outcome variables. Odds ratios with 95% confidence limits not including 1 are considered significant. The unit of reference for the odds ratios is mFI 0.0-0.09

ACCEPTED MANUSCRIPT

P-value*

Reference 2.3 (0.95, 5.40) 5.5 (2.44, 12.30) 9.2 (4.05, 20.69)

<0.001

1.2 (0.78, 1.77) Reference

0.431

RI PT

OR (95% CI)*

AC C

EP

TE D

M AN U

SC

Table III: Univariable Analysis for Factors Affecting Mortality No 30-day 30-day Mortality Mortality Variable (N=4604) (N=100) Age (years) 1227 (99.4) 7 (0.6) ≤60 >60-70 1471 (98.7) 19 (1.3) 1217 (97.0) 38 (3.0) >70-80 >80 689 (95.0) 36 (5.0) Sex Female 1626 (97.7) 39 (2.3) Male 2947 (98.0) 60 (2.0) Race Non-Hispanic Black 713 (97.1) 21 (2.9) Non-Hispanic American Indian or 12 (100) 0 (0) Alaska native 38 (95.0) 2 (5.0) Non-Hispanic Asian 70 (2.0) 3400 (98.0) Non-Hispanic White BMI (kg/m^2) <25 1592 (97.3) 45 (2.8) 25-<30 1567 (98.3) 26 (1.6) 30-<40 1192 (98.3) 21 (1.7) >40 163 (98.8) 2 (1.2) ASA class No Disturbance-Mild Disturbance (1-2) 303 (100) 0 (0) Severe Disturbance (3) 3415 (98.7) 44 (1.3) Life Threatening-Moribund (4-5) 885 (94.0) 56 (6.0) Transfer status From acute care hospital inpatient 155 (96.9) 5 (3.1) Not transferred (admitted from home) 4246 (98.2) 78 (1.8) Nursing home - Chronic care 139 (89.7) 16 (10.3) Intermediate care Outside emergency department 33 (97.1) 1 (2.9) Transfer from other 28 (100) 0 (0) Smoker Yes 1895 (99.0) 20(1.0) No 2709 (97.1) 80 (2.9) Alcohol use Yes 281 (97.9) 6 (2.1) No 4323 (97.9) 94 (2.1) Impaired sensorium Yes 28 (82.4) 6 (17.7) No 4576 (97.7) 94 (2.0)

1.4 (0.87,2.35) <0.001(<0.001,>999.9) 2.6 (0.61,10.81) Reference

0.015

Reference 0.6 (0.36,0.96) 0.6 (0.37,1.05) 0.4 (0.10,1.81)

0.092

Reference 5.6 (1.25,Infinity) 27.4 (6.184,Infinity)

<0.001

Reference 1.8 (0.70,4.40) 6.3 (3.57,11.01) 1.7(0.22,12.21) <0.001(<0.001,>999.9)

<0.001

0.4 (0.22,0.59) Reference

<0.001

1.0 (0.43,2.26) Reference

0.966

10.4 (4.22,25.79) Reference

<0.001

ACCEPTED MANUSCRIPT

39 (6.3) 61 (98.5)

4.4 (2.90,6.67) Reference

<0.001

1990 (97.2) 2614 (98.4)

58 (2.8) 42 (1.6)

1.8 (1.21,2.71) Reference

0.004

612 (96.4) 3992 (98.1)

23 (3.6) 77 (1.9)

11 (100) 4593 (97.9)

0 (0) 100 (2.1)

1.06 (93.0) 4498 (98.0)

8 (7.0) 92 (2.0)

3.7 (1.75,7.80) Reference

<0.001

938 (97.8) 3.666 (98.0)

21 (2.2) 79 (2.1)

1.0 (0.64,1.69) Reference

0.876

1074 (96.9) 3530 (98.2)

34 (3.1) 66 (1.8)

1.7 (1.11,2.58) Reference

0.014

84 (93.3) 4520 (98.0)

6 (6.7) 94 (2.0)

3.4 (1.46,8.06) Reference

<0.005

3893 (97.8) 711 (98.3)

88 (2.2) 12 (1.7)

1.3 (0.73,2.46) Reference

0.347

2608 (98.2) 1996 (97.4)

47 (1.8) 53 (2.6)

0.7 (0.456,1.009) Reference

0.056

2140 (97.1) 2464 (98.6)

65 (3.0) 35 (1.4)

2.1 (1.41,3.24) Reference

<0.001

293 (97.0) 4311 (97.9)

9 (3.0) 91 (2.1)

1.5 (0.73,2.92) Reference

0.290

355 (97.5) 4249 (97.9) 2.69 ± 2.63

9 (2.5) 91 (2.1) 2.87 ± 2.46

1.2 (0.59,2.37) Reference 1.0 (0.95, 1.11)

0.633 0.492

242 (93.8) 4362 (98.1)

16 (6.2) 84 (1.9)

3.4 (1.98,5.95) Reference

<0.001

RI PT

582 (93.7) 4022 (98.5)

2.0 (1.21,3.13) Reference

0.006

<0.001(<0.001,>999.9) Reference

0.983

SC

M AN U

TE D

AC C

EP

Dependent Functional status Yes No Diabetes Yes No History of COPD Yes No History of pneumonia Yes No History of CHF Yes No History of PCI Yes No History of PCS Yes No History of angina Yes No Hypertension Yes No History of PVD Yes No History of rest pain Yes No History of TIA Yes No History of CVA Yes No PGY level (1-unit increase) Emergency operation Yes No

ACCEPTED MANUSCRIPT

312.04 ± 129.43

Dialysis Yes No

309 (92.0) 4295 (98.3)

Hospital length of stay (7-day increase)

8.08 ± 8.54

27 (8.0) 73 (1.7) 12.96 ± 10.59 2 (0.6) 98 (2.3)

106 (96.4) 4498 (97.9)

4 (3.6) 96 (2.1)

10 (100) 4594 (97.9)

0 (0) 100 (2.1)

EP

AC C

5.1 (3.26,8.11) Reference 1.3 (1.16, 1.39)

M AN U

0.121

<0.001 <0.001

0.3 (0.06,0.10) Reference

0.050

1.8 (0.064,4.90) Reference

0.273

<0.001 (<0.001,999.9) Reference

0.984

SC

354 (99.4) 4250 (97.8)

23 (82.1) 4581 (98.0)

5 (17.9) 95 (82.1)

10.5 (3.90,28.16) Reference

<0.001

25 (89.3) 4579 (97.9)

3 (10.7) 97 (2.1)

10.5 (3.90,28.16) Reference

0.005

86 (82.7) 4518 (98.2)

18 (17.3) 82 (1.8)

11.5 (6.63,20.05) Reference

<0.001

201 (97.6) 4403 (97.9)

5 (2.4) 95 (2.1)

1.2 (0.46,2.87) Reference

0.759

47 (100) 4557 (97.8)

0 (0) 100 (2.2)

<0.001 (<0.001,999.9) Reference

0.985

123 (91.8) 4481 (98.0)

11 (8.2) 89 (2.0)

4.5 (2.35,8.64) Reference

<0.001

748 (96.8) 3856 (98.1)

25 (3.2) 75 (1.9)

1.7 (1.09,2.72) Reference

0.021

6.21 ± 6.58

9.04 ± 7.80

1.3 (1.12, 1.40)

154 (96.2) 4450 (97.9) 13.97 ±

6 (3.8) 94 (2.1) 8.83 ±

1.9 (0.80,4.28) Reference 0.68 (0.47, 0.99)

TE D

Superficial incisional SSI Yes No Deep incisional SSI Yes No Post-op pulmonary embolism Yes No Post-op progressive renal insufficiency Yes No Post-op stroke Yes No Post-op MI Yes No Graft/prosthesis/flap failure Yes No DVT requiring therapy Yes No Sepsis Yes No Return to OR Yes No Days from operation to discharge (7-day increase) Steroid use Yes No Pre-op albumin serum level

1.1 (0.97, 1.32)

RI PT

293.75 ± 116.62

Anesthesia time (100-minute increase)

<0.001

0.154 0.041

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

19.49 16.08 Acute renal failure requiring dialysis w/in 30 days of operation 31.1 (15.54,62.11) Yes 24 (63.2) 14 (36.8) <0.001 No 4580 (98.2) 86 (1.8) Reference * All odds ratios and p-values are from binomial logistic regression modeling 30-day mortality, exact logistic regression used as needed. Odds ratios with 95% confidence limits not including 1 are considered significant.

ACCEPTED MANUSCRIPT

Table IV: Multivariant Analysis for Factors Affecting 30-day Mortality

Odd’s Ratio (CI)

mFI 0.18-0.27 vs. 0-0.09

4.4 (0.9-21.9)

mFI 0.36-0.45 vs. 0-0.09

7 (1.4-34.5)

mFI 0.54-0.63 vs. 0-0.09

7.9 (1.3-48.01)

Age 60-70 vs. <60

1.7 (0.64-4.3)

SC

RI PT

Variables

Age 70-80 vs. <60

3.1 (1.3-7.4) 3.1 (1.2-7.9)

Black Race Vs. White Race Other Race Vs. White Race Dialysis Dependency

M AN U

Age >80 vs. <60

1.4 (0.8-2.6)

3.1 (0.98-9.8) 2.7 (1.5-4.8)

TE D

Post operative progressive renal insufficiency

7.7 (2.1-28.9)

7.5 (3.7-15.1)

Post operative Acute Renal Failure requiring Dialysis

13.8 (5.4-35.3)

AC C

EP

Post operative Myocardial infarction