ONLINE APPENDIX Additional information regarding the statistical

ONLINE APPENDIX
Additional information regarding the statistical analysis
1. Noninferiority for MACE of the FFRCT-guided strategy compared with usual care was
assessed using a 3% noninferiority margin and a 1-sided test with a 0.05 type 1 error rate.
All other comparisons were conducted as 2-sided tests with a 0.05 type 1 error rate.
2. We used the Wilcoxon signed rank test to analyze changes in QOL scores from baseline
to 1-year follow-up for the entire cohort and the Wilcoxon rank sum test to compare QOL
changes between groups.
3. Changes in medication use from baseline to 1-year follow-up were compared between
groups using logistic regression fit with generalized estimating equations.
4. Cumulative radiation exposure was compared between groups using the Wilcoxon rank
sum test.
1
Online Figure 1. Case Examples of FFRCT Use in PLATFORM
1. Case 1: 65-year-old man with new onset atypical chest pain and planned noninvasive evaluation.
(A) CTA as the initial test showed 50-70% mid and distal LAD stenoses; calcium score was 387. (B)
FFRCT was 0.86. No angiogram was performed. At 1 year the patient had no further testing, no
clinical events, and was symptom free.
2. Case 2: 65-year-old man with atypical chest pain, positive ETT, and planned invasive coronary
angiography. (A) CTA prior to ICA revealed 50-70% proximal and mid LAD stenoses, and >50%
LCx and RCA stenoses; calcium score was 1509. (B) FFRCT was 0.64 in the distal LAD. (C)
Invasive angiography demonstrated 3-vessel disease with 51% LAD stenosis by quantitative
coronary angiography (QCA) with FFR=0.61. The patient was treated with CABG and was
asymptomatic at 1 year.
3. Case 3: 69-year-old woman with diabetes and typical angina and planned noninvasive evaluation.
(A) CTA as the initial test showed 30-50% LAD stenosis. (B) FFRCT was 0.75 in the distal LAD. (C)
Invasive angiography confirmed 73% distal LAD stenosis by QCA, with FFR=0.69. The patient was
treated with PCI and was pain free at 1 year.
4. Case 4: 70-year-old woman with atypical chest pain and planned noninvasive evaluation. (A) CTA
as the initial test showed calcified 70-90% proximal LAD stenosis. (B) LAD FFRCT = 0.86. (C)
Invasive angiography revealed LAD stenosis >70% visually but 47% by QCA, with FFR=0.82. The
patient was treated with PCI but had continuing chest pain at 1 year.
2
3
Online Table 1. Cost Weights Based on Medicare 2015 Reimbursements (US Dollars)
___________________________________________________________________________________
Resource
Stress testing
Exercise ECG
Stress echo
Stress perfusion imaging
Coronary CTA
Magnetic resonance imaging
Invasive tests
Coronary angiography
FFRinv (1st vessel)
FFRinv (additional vessels)
Intravascular ultrasound
Coronary revascularization
PCI
Drug-eluting stents (per stent)
CABG
PCI complication
CABG complication
Visits
Clinic
Emergency department
Medications (per day)
Aspirin
Clopidogrel
Prasugrel
Statin
Ticagrelor
Cost Weight
($US)
CPT, DRG, or
APC Code
76
349
567
301
581
93015
93351, 93015
78452, 93015
71275
71552
2838
101
81
101
80*, 93454
93571
93572
93978
12,528
736
30,228
6750
14,265
108
119
0.04
0.75
9.68
0.53
8.95
249†
33518, 92928
234†, 33518, 33533
‡
‡
99214
99284
§
§
§
§
§
APC=Ambulatory Payment Classifications; CABG=coronary artery bypass grafting; CPT=Current Procedural Terminology;
CTA=computed tomographic angiography; DRG=Diagnosis-Related Group; ECG=electrocardiogram; FFRINV= fractional flow
reserve determined by invasive coronary angiography; ICA=invasive coronary angiography; PCI=percutaneous coronary
intervention.
*APC code.
†
DRG code.
‡
Difference between DRG codes with and without major complicating conditions.
§
Internet pharmacy costs for typical daily dosage.
4
Online Table 2. Baseline Characteristics of the Study Participants, According to Study Group
Variable
Demographics
Age, mean (SD), y
Female sex, no. (%)
Racial/ethnic minority (self-reported), no.
(%)
Cardiac risk factors
Hypertension, no. (%)
Diabetes, no. (%)
Dyslipidemia, no. (%)
Current or past tobacco use, no. (%)
Pre-test probability of obstructive CAD
(SD)*, %
Relevant medications, no. (%)
Aspirin
Statin
Anginal type, no. (%)
Typical angina
Atypical angina
Non-cardiac chest pain
Prior noninvasive testing, no. (%)
Planned Noninvasive Test
(N=204)
FFRCTUsual care
guided
strategy
strategy
(N=100)
(N=104)
P Value
Planned Invasive Test
(N=380)
FFRCTUsual care
guided
strategy
strategy
P
(N=187)
(N=193)
Value
57.9 (10.7)
34 (34.0)
59.5 (9.3)
44 (42.3)
0.25
0.22
63.4 (10.9)
79 (42.2)
60.7 (10.2)
74 (38.3)
0.02
0.44
5 (5.0)
0 (0.0)
0.06
2 (1.1)
1 (0.5)
0.60
38 (38.0)
8 (8.0)
22 (22.0)
52 (52.0)
57 (54.8)
6 (5.8)
28 (26.9)
59 (56.7)
0.02
0.52
0.49
0.50
111 (59.4)
36 (19.3)
76 (40.6)
103 (55.1)
111 (57.5)
30 (15.5)
77 (39.9)
101 (52.3)
0.72
0.33
0.81
0.59
44.5 (15.3)
45.3 (16.8)
0.89
51.7 (16.7)
49.4 (17.2)
0.26
29 (29.0)
24 (24.0)
45 (43.3)
29 (27.9)
0.039
0.58
0.02
115 (61.5)
83 (44.4)
90 (46.6)
77 (39.9)
<0.01
0.37
0.09
8 (8.0)
91 (91.0)
1 (1.0)
NA
18 (17.3)
80 (76.9)
6 (5.8)
NA
52 (27.8)
122 (65.2)
13 (7.0)
92 (49.2)
45 (23.3)
142 (73.6)
5 (2.6)
101 (52.3)
0.54
CAD=coronary artery disease. FFRCT=fractional flow reserve estimated using computed tomography.
*Mean pre-test probability of obstructive CAD±SD calculated by updated Diamond and Forrester score.1
5
Online Table 3. Performance of FFRCT vs FFR in PLATFORM
Value
Confidence Limits
N
Diagnostic accuracy
84.0%
70.9%, 92.8%
50
Sensitivity
77.8%
52.4%, 93.6%
18
Specificity
87.5%
71.0%, 96.5%
32
PPV
77.8%
52.4%, 93.6%
18
NPV
87.5%
71.0%, 96.5%
32
Among the 50 vessels in 29 patients for which both FFRCT and invasively measured FFR were available, comparison using FFR and
FFRCT cutoffs of 0.80 showed: True Negative=28; True Positive=14; False Negative=4; False Positive=4, for a total of 50 vessels.
The resulting performance as shown in the table is similar to that reported in the NXT trial. 2
6
Online Table 4. Mean 12-Month Costs in Dollars According to Angina Classification
Planned Noninvasive Test
Usual Care Strategy
(N=8)
Planned Invasive Test
FFRCT-guided Strategy
(N=18)
P-value
Usual Care Strategy
(N=52)
FFRCT-guided Strategy
(N=45)
P value
3215.58 (6147.52)
0.420
12303.88 (13200.35)
11206.47 (14618.15)
0.047
3113.53 (6552.59)
0.761
11685.53 (12725.22)
7310.19 (10781.40)
<0.001
Typical angina
3292.20 (4620.60)
Atypical angina
2543.34 (6929.64)
Data are presented as mean (SD). Although this table compares point estimates for costs, it is important to note that the interaction
P value for angina type and strategy for cost difference is 0.16, so any trends are not significant.
7
References
1. Genders TS, Steyerberg EW, Alkadhi H, et al. A clinical prediction rule for the diagnosis of
coronary artery disease: validation, updating, and extension. Eur Heart J 2011;32:1316–30.
2. Nørgaard BL, Leipsic J, Gaur S, et al. Diagnostic performance of noninvasive fractional flow
reserve derived from coronary computed tomography angiography in suspected coronary artery
disease: the NXT trial (Analysis of Coronary Blood Flow Using CT Angiography: Next Steps). J
Am Coll Cardiol 2014;63:1145–55.
8