Table 5.
Sensitivity and specificity of results when healthcare professionals determined whether patients are at high risk of coronary heart disease.* Values are numbers (percentage, 95% confidence interval)
| Method used to calculate risk | Sensitivity†‡
|
Specificity†‡
|
|||
|---|---|---|---|---|---|
| Practice nurses | General practitioners | Practice nurses | General practitioners | ||
| Subjective risk estimate | 55/83 (66, 56 to 76) | 60/83 (72, 63 to 82) | 70/97 (72, 63 to 81) | 71/97 (73, 64 to 82) | |
| Sheffield table | 45/77 (58, 47 to 69) | 48/75 (64, 53 to 75) | 43/50 (86, 76 to 96) | 50/56 (89, 81 to 97) | |
| New Zealand table | 50/80 (63, 52 to 73) | 60/81 (74, 65 to 84) | 48/68 (63, 52 to 73) | 43/53 (81, 71 to 92) | |
| European table | 51/79 (65, 54 to 75) | 49/80 (61, 51 to 72) | 52/70 (74, 64 to 84) | 50/69 (72, 62 to 83) | |
| British program | 58/73 (79, 70 to 89) | 60/75 (80, 71 to 89) | 23/23 (100) | 20/24 (83, 68 to 98) | |
*
Compared with the reference standard of Framingham equation, with missing values replaced by population normal values.
†
Reference standard for high risk is 10 year risk of coronary heart disease >30%.
‡
Variations in denominators between general practitioners and practice nurses for the same calculation tool reflect variations in number of calculations made.