Table 1 Demographic and clinical characteristics of the five cohorts included in the study.

From: A machine learning model for identifying patients at risk for wild-type transthyretin amyloid cardiomyopathy

Characteristic

Cohort 1: IQVIA (ATTR-CM)

Cohort 2: Optum (ATTR-CM)

Cohort 3: IQVIA (cardiac amyloid)

Cohort 4: Optum (cardiac amyloid)

Cohort 5: NMEDW* (cardiac amyloid)

ATTRwt-CM + HF (n = 1071)

Non-amyloid HF (n = 1071)

ATTRwt-CM + HF (n = 173)

Non-amyloid HF (n = 173)

Cardiac amyloid + HF (n = 7296)

Non-amyloid HF (n = 7296)

Cardiac amyloid + HF (n = 1943)

Non-amyloid HF (n = 1943)

Cardiac amyloid + HF (n = 261)

Non-amyloid HF (n = 39,393)

Age, years

77 ± 7

77 ± 7

77 ± 8

74 ± 11

78 ± 8

77 ± 8

78 ± 9

77 ± 11

74 ± 9

72 ± 11

Male

84%

84%

82%

81%

67%

67%

68%

69%

69%

56%

Ethnicitya

   White

66%

80%

72%

80%

54%

71%

   Black

28%

11%

22%

10%

32%

13%

   Other

6%

9%

6%

10%

14%

16%

Location of final visit

   Midwest

28%

20%

47%

48%

24%

21%

49%

49%

100%

100%

   Northeast

27%

20%

25%

16%

28%

19%

22%

11%

0%

0%

   South

25%

40%

20%

29%

28%

39%

19%

28%

0%

0%

   West

18%

18%

5%

5%

19%

19%

6%

9%

0%

0%

   Other/unknown

1%

2%

3%

2%

1%

1%

3%

4%

0%

0%

Comorbidities

   Hypertension

90%

96%

83%

88%

89%

92%

81%

86%

72%

78%

   Obesity

42%

48%

48%

47%

37%

42%

34%

39%

21%

25%

   Diabetes

42%

62%

35%

52%

45%

55%

39%

50%

38%

36%

   CAD

64%

73%

54%

68%

60%

65%

59%

65%

56%

54%

   CKD

61%

44%

58%

47%

52%

42%

56%

43%

47%

29%

   Atrial fibrillation

72%

52%

65%

52%

64%

50%

61%

50%

56%

41%

Diagnosis history duration, yearsb

9.6 (6.2–10.4)

9.7 (6.3–10.5)

6.2 (2.8–9.8)

6.4 (3.0–9.5)

8.7 (6.0–10.2)

8.7 (6.0–10.2)

5.2 (2.5–8.3)

5.4 (2.7–8.2)

6.2 (1.2–15.7)

4.6 (0.9–10.1)

Number of visitsb

129 (82–210)

131 (87–205)

79 (34–187)

76 (31–168)

117 (63–193)

112 (60–187)

67 (25–141)

66 (24–140)

37 (9–99)

21 (6–60)

  1. T-tests (for age), Wilcoxon rank-sum tests (for diagnosis history duration and number of visits), and χ2-tests (for categorical variables) were used to compare groups. P-values are two-sided. Adjustments were not made for multiple comparisons.
  2. ATTRwt-CM wild-type amyloidogenic transthyretin cardiomyopathy, CAD coronary artery disease, CKD chronic kidney disease, NMEDW Northwestern Memorial Enterprise Data Warehouse.
  3. aEthnicity data were not available in the IQVIA data.
  4. bMedian (25th–75th percentile).
  5. *P < 0.001 for age, sex, ethnicity, CKD, AF, diagnosis history, and number of visits, and P = 0.015 for hypertension for differences between cardiac amyloid and non-amyloid HF in the NMEDW cohort.