Skip to main content

Table 1 Validation of administrative data algorithms to adults identified with diabetes using clinical data from primary care electronic medical records as a reference standard; all adults (prevalence 10.9%), using all administrative data available from 1991 to 2013

From: Identifying diabetes cases from administrative data: a population-based validation study

Algorithm

Sensitivity (%) (95% CI)

Specificity (%)

(95% CI)

PPV (%)

(95% CI)

NPV (%)

(95% CI)

Prevalence

Kappa

Probability+

Probability-

Physician claims only

 1 P

93.6(93.3–94.0)

91.9(91.7–92.0)

58.5(57.9–59.1)

99.2(99.1–99.2)

17.5%

0.68

61.7%

0.96%

 2 P in 1 yr

87.2(86.7–87.7)

98.1(98.0–98.1)

84.6(84.1–85.2)

98.4(98.4–98.5)

11.2%

0.84

86.5%

1.79%

 3 P in 1 yr

79.9(79.3–80.6)

99.1(99.0–99.1)

91.4(91.0–91.9)

97.6(97.5–97.7)

9.5%

0.84

92.5%

2.75%

 2 P in 2 yr

88.4(87.9–88.9)

97.8(97.8–97.9)

83.4(82.9–84.0)

98.6(98.5–98.6)

11.6%

0.84

84.9%

1.63%

 3 P in 2 yr

83.1(82.5–83.7)

98.9(98.8–98.9)

90.1(89.7–90.6)

98.0(97.9–98.0)

10.1%

0.85

91.3%

2.33%

Inclusion of prescription claims

 1 Rx

50.7(50.0–51.5)

99.8(99.8–99.9)

97.3(97.0–97.7)

94.3(94.2–94.4)

5.7%

0.64

97.2%

6.44%

 1 P or 1 Rx

94.4(94.0–94.7)

91.8(91.6–91.9)

58.4(57.8–59.0)

99.3(99.2–99.3)

17.6%

0.68

61.6%

0.84%

 1 P and 1 Rx

50.0(49.3–50.8)

99.9(99.9–99.9)

98.5(98.3–98.8)

94.2(94.1–94.4)

5.5%

0.64

98.6%

6.52%

 (2 P in 1 yr) or (1Rx and 1 P)

88.6(88.1–89.1)

98.0(98.0–98.1)

84.6(84.1–85.1)

98.6(98.5–98.7)

11.4%

0.85

86.1%

1.60%

Inclusion of hospital records

 H

36.7(36.0–37.5)

99.6(99.6–99.6)

91.9(91.2–92.5)

92.8(92.7–92.9)

4.4%

0.49

92.7%

8.14%

 H or 1 P

94.0(93.6–94.3)

91.7(91.5–91.8)

58.0(57.4–58.6)

99.2(99.2–99.2)

17.7%

0.67

61.2%

0.90%

 H or 1 Rx

61.3(60.5–62.0)

99.5(99.4–99.5)

93.3(92.8–93.7)

95.5(95.3–95.6)

7.2%

0.72

94.5%

5.14%

 H or (2 P in 1 yr)

88.4(87.9–88.8)

97.8(97.7–97.9)

83.0(82.4–83.5)

98.6(98.5–98.6)

11.6%

0.84

84.9%

1.63%

 H or (3 P in 1 yr)

82.4(81.8–83.0)

98.8(98.7–98.8)

89.1(88.6–89.6)

97.9(97.8–97.9)

10.1%

0.84

90.5%

2.42%

 H or (2 P in 2 yr) a

89.3(88.9–89.8)

97.6(97.5–97.7)

81.9(81.3–82.4)

98.7(98.6–98.7)

11.9%

0.84

83.8%

1.51%

 H or (3 P in 2 yr)

84.9(84.4–85.5)

98.6(98.5–98.6)

88.0(87.5–88.5)

98.2(98.1–98.2)

10.5%

0.85

89.4%

2.09%

Physician claims, hospital records, and prescription claims

 (H or (2 P in 1 yr)) or 1 Rx

90.0(89.5–90.4)

97.7(97.6–97.8)

82.6(82.0–83.1)

98.8(98.7–98.8)

11.9%

0.84

84.5%

1.41%

 (H or (2 P in 2 yr)) or 1 Rx

90.7(90.3–91.2)

97.5(97.4–97.6)

81.5(80.9–82.0)

98.9(98.8–98.9)

12.1%

0.84

83.5%

1.31%

 (H or (3 P in 2 yr)) or 1 Rx

87.4(86.9–87.9)

98.5(98.4–98.5)

87.5(87.0–88.0)

98.5(98.4–98.5)

10.9%

0.86

89.0%

1.75%

 (H or 1 P) and 1 F

77.2(76.5–77.8)

99.2(99.2–99.3)

92.6(92.1–93.0)

97.3(97.2–97.3)

9.1%

0.82

93.1%

3.10%

 (H or 1 P) and 1 Rx

50.2(49.4–50.9)

99.9(99.9–99.9)

98.5(98.2–98.7)

94.2(94.1–94.4)

5.6%

0.64

98.6%

6.50%

 (H or 1 P) and (1 Rx or 1 F)

84.2(83.6–84.7)

99.2(99.1–99.2)

92.5(92.1–93.0)

98.1(98.0–98.2)

9.9%

0.87

93.6%

2.17%

  1. PPV positive predictive value, NPV negative predictive value; probability+, probability of having diabetes with the algorithm, probability- probability of having disease without the algorithm, H hospital discharge abstracts bearing a diagnosis of diabetes from the Canadian Institute for Health Information Discharge Abstract Database, P physician claims for a diabetes diagnosis (ICD-9 250) from the Ontario Health Insurance Plan Physician Claims Database, Rx prescription for an anti-hyperglycemic medication from the Ontario Drug Benefit Database, F diabetes specific physician fee code from the Ontario Health Insurance Plan Physician Claims Database, yr year
  2. a Current CDSS algorithm
  3. Reference standard: EMR chart – adult (≥20 years old) with diabetes or one of its synonyms were listed in the cumulative patient profile or they had any of the following: haemoglobin A1c greater than 7%, two abnormal blood glucose tests [fasting blood sugar(s) greater than or equal to 7.0 mmol/L, or a random blood sugar(s) greater than or equal to 11.1 mmol/L], or a prescription for an anti-hyperglycemic medication (insulin or an oral hypoglycemic agent). Patients were excluded if they only had a record of gestational diabetes