SUPPLEMENTAL ONLINE-ONLY MATERIAL

Table E1. Pre-Implant Incremental Risk Factors for Death on HeartMate II Support

Factor Coefficient ± SE P Reliability (%)a

Early Phase

History of valve surgery 1.29 ± 0.52 .01 76

History of CABG 1.33 ± 0.51 .009 76

Lower albuminb 1.79 ± 0.38 <.0001 62

Lower GFRc -1.59 ± 0.64 .01 66

Higher mean arterial blood pressure at implantd

0.16 ± 0.071 .03 52

Constant Phase

History of hypertensione 0.75 ± 0.32 .02 65

Earlier date of implantf -1.76 ± 0.49 .0004 90

a. Percent of times variable appeared in 500 bootstrap models

b. (Albumin/3)2, inverse squared transformation

c. (GFR), logarithmic transformation

d. (Mean arterial blood pressure/25)2, squared transformation

e. Hypertension is defined as a history of drug-treated systemic hypertension, reported in the patient’s

complete history and physical evaluation when they are being considered for transplant or VAD implant.

This variable was deactivated in the UNOS form in 2015.   f. (Date of implant/6), logarithmic transformation

Key: CABG, coronary artery bypass grafting; GFR, glomerular filtration rate; SE, standard error

1

Table E2. Adverse Events during HeartMate II Support

Adverse Event Patients (n) Events (n)

Dialysis 46 46

Stroke (ischemic) 29 31

Intracranial hemorrhage 22 25

Device exchange 16 16

Pump thrombosis

Confirmed 12 12

Suspected 15 15

Hemolysis 30 32

Infection

Blood 78 78

Pump pocket 18 18

Driveline 61 61

Tracheostomy 53 53

Pre-discharge reoperation

For bleeding 50 50

For tamponade 13 13

Gastrointestinal bleed 60 60

2

Table E3. Incremental Risk Factors for Dialysis on HeartMate II Support

Factor Coefficient ± SE P Reliability (%)a

Early Phase

Higher pulmonary systolic pressureb 2.09 ± 0.71 .005 95

History of CABG 0.97 ± 0.34 .005 50

Lower albuminc 1.57 ± 0.42 .0003 65

Higher creatinined 1.06 ± 0.48 .03 73

Higher bilirubin 0.28 ± 0.071 .0002 55

Late Phase

Higher creatinined 9.41 ± 3.34 .005 66

Longer pre-implant hospital stay 0.077 ± 0.029 .01 63

a. Percent of times variable appeared in 500 bootstrap models

b. (Pulmonary artery pressure-systolic), logarithmic transformation

c. (Albumin/3)2, inverse squared transformation

d. (Creatinine), logarithmic transformation

Key: CABG, coronary artery bypass grafting; SE, standard error

3

Table E4. Incremental Risk Factors for Intracranial Hemorrhage on HeartMate II Support

Risk Factor Coefficient ± SE P Reliability (%)a

Early Phase

No significant risk factors

Late Phase

Higher MELD scoreb -4.26 ± 1.92 .03 54

a. Percent of times variable appeared in 500 bootstrap models.

b. (12/MELD), inverse transformation

Key: MELD, Model for End-Stage Liver Disease; SE, standard error

4

FIGURE LEGENDS

Figure E1. Illustrations of strength of risk factors for death during HeartMate II support. A, Survival

stratified by prior coronary artery bypass grafting (CABG). Each symbol represents a death

positioned according to the Kaplan-Meier estimator; vertical bars represent confidence limits

equivalent to ±1 standard error. Depiction stops at last event. B, Survival stratified by prior valve

surgery. Depiction is as in Figure E1A. C, Effect of albumin on 3-month survival. Nomogram of

multivariable Table E1 with the following risk factors: no history of valve surgery or CABG,

glomerular filtration rate of 65 mL-1·min-1·1.73 m2, mean blood pressure 70 mmHg, history of

hypertension, date of implant June 2010. D, Effect of glomerular filtration rate (GFR) on 3-month

survival. Depiction as in Figure E1C, except albumin is 3.5 g·dL-1. E, Effect of mean arterial blood

pressure on 1-year survival. Depiction as in Figure E1C, except albumin is 3.5 g·dL-1. F, Survival

stratified by history of hypertension. Depiction is as in Figure E1A. G, Effect of date of implant on

1-year survival. Depiction is as in Figure E1C except albumin is 3.5 g·dL-1.

5

6

7

8

Figure E2. Survival stratified by device exchange. Time zero for curves for “Second HeartMate II”

is time of pump exchange. Depiction is as in Figure 1.

9

10

Appendix E1. Variables Considered in the Analyses

Pre–Mechanical Circulatory Support (MCS)

Demographics: Age (y), sex, body surface area (m2), body mass index (kg∙m-2),

height (cm), weight (kg), weight-to-height ratio (kg∙m-1), race

Blood group: A, B, AB, O

Hemodynamics: Pulmonary artery pressure (systolic, mean, diastolic; mmHg),

cardiac index (L∙min-1·m-2), heart rate, systemic blood pressure

(systolic, mean, diastolic, mmHg)

Medical condition: Dilated cardiomyopathy, history of coronary artery disease,

previous myocardial infarction, previous cardiac operation,

coronary artery bypass grafting, implantable cardioverter-

defibrillator, medical condition at time of MCS (intensive care unit

[ICU], in-hospital but not ICU, not hospitalized, home on

ventricular assist device), ventricular tachycardia or fibrillation,

New York Heart Association functional class, hypertension

Noncardiac comorbidity: Albumin (g∙dL-1), blood urea nitrogen (mg∙dL-1), creatinine

(mg∙dL-1), creatinine clearance, glomerular filtration rate

(mL-1·min-1·1.73 m2), MELD score, hematocrit (%), hemoglobin

(g·dL-1), International Normalized Ratio, platelets, sodium, total

bilirubin (mg∙dL-1), smoking, diabetes, alcohol use

11

Echocardiogram: Ejection fraction (%), ventricular function, insufficiency

Hospital length of stay: Admission to implant (days)

INTERMACS: Level 1-6

Strategy: Destination therapy, bridge to transplant/decision

Temporary support: Before HeartMate II: Intra-aortic balloon pump, intravenous

inotropes, ventilator, extracorporeal membrane oxygenation

(ECMO)

Panel reactive antibodies (PRA): Pre-HeartMate II PRA

Experience: Interval from 1/1/2004 to date of MCS

During MCS

Mode of support: ECMO, number of device exchanges

Adverse events: Stroke, intracranial bleed, thrombosis, hemolysis, driveline

infection, pump pocket infection, blood stream infection, dialysis,

tracheostomy, reoperation for bleeding and tamponade

Pre-Transplant

Time: Interval between HeartMate II and transplant

Demographics: Age (y), body surface area (m2), body mass index

12

(kg∙m-2), height (cm), weight (kg), weight-to-height ratio

(kg∙m-1), race

Hemodynamics: Pulmonary artery pressure (systolic, mean, diastolic; mmHg),

cardiac index (L∙min-1·m-2), heart rate, systemic blood pressure

(systolic, mean, diastolic, mmHg), wedge pressure (mmHg),

central venous pressure (mmHg)

Noncardiac comorbidity: Albumin (g∙dL-1), blood urea nitrogen (mg∙dL-1),

creatinine (mg∙dL-1), creatinine clearance, glomerular filtration rate

(mL-1·min-1·1.73 m2), MELD score, hematocrit (%), hemoglobin

(g·dL-1), International Normalized Ratio, platelets, sodium, total

bilirubin (mg∙dL-1), smoking, diabetes, alcohol use

Echocardiogram: Ejection fraction (%), ventricular function, insufficiency

Hospital length of stay: Admission to transplant (days)

Experience: Interval from 1/1/2004 to date of transplant

13