Performance of multinomial regression and a rule-based algorithm using ICD-10 codes for determining diabetes status and type using SEARCH cohort status as gold standard
| . | Multinomial regression* . | Rule-based algorithm . |
|---|---|---|
| Diabetes (n = 5,308) | ||
| Se | 0.964 | 0.991 |
| Sp | 0.982 | 0.966 |
| PPV | 0.987 | 0.969 |
| NPV | 0.918 | 0.983 |
| Type 1 diabetes (n = 4,732) | ||
| Se | 0.953 | 0.978 |
| Sp | 0.963 | 0.968 |
| PPV | 0.978 | 0.980 |
| NPV | 0.957 | 0.967 |
| Type 2 diabetes (n = 400) | ||
| Se | 0.573 | 0.899 |
| Sp | 0.992 | 0.975 |
| PPV | 0.778 | 0.642 |
| NPV | 0.977 | 0.995 |
| Other diabetes type, e.g., medication-induced, monogenic (n = 176) | ||
| Se | 0.381 | 0.496 |
| Sp | 0.981 | 0.996 |
| PPV | 0.512 | 0.698 |
| NPV | 0.986 | 0.988 |
| κ statistic | 0.870 | 0.910 |
| Accuracy | 0.936 | 0.955 |
| . | Multinomial regression* . | Rule-based algorithm . |
|---|---|---|
| Diabetes (n = 5,308) | ||
| Se | 0.964 | 0.991 |
| Sp | 0.982 | 0.966 |
| PPV | 0.987 | 0.969 |
| NPV | 0.918 | 0.983 |
| Type 1 diabetes (n = 4,732) | ||
| Se | 0.953 | 0.978 |
| Sp | 0.963 | 0.968 |
| PPV | 0.978 | 0.980 |
| NPV | 0.957 | 0.967 |
| Type 2 diabetes (n = 400) | ||
| Se | 0.573 | 0.899 |
| Sp | 0.992 | 0.975 |
| PPV | 0.778 | 0.642 |
| NPV | 0.977 | 0.995 |
| Other diabetes type, e.g., medication-induced, monogenic (n = 176) | ||
| Se | 0.381 | 0.496 |
| Sp | 0.981 | 0.996 |
| PPV | 0.512 | 0.698 |
| NPV | 0.986 | 0.988 |
| κ statistic | 0.870 | 0.910 |
| Accuracy | 0.936 | 0.955 |
Accuracy = number correctly classified / N. Positive (LR+) and negative (LR−) likelihood ratios may be calculated with the following formulas: LR+ = Se / (1 − Sp), LR− = (1 − Se) / Sp.
Variables in the final multinomial regression model included the following: most common diabetes type–specific code, maximum HbA1c, proportion of type 2 diabetes codes, any elevated outpatient glucose, any metformin, any antidiabetes medicine, age, proportion of type 1 diabetes codes, multiple elevations in outpatient random glucose, obesity, any diabetic ketoacidosis, ethnicity, any contraceptive medication, count of type 1 diabetes codes, proportion of other diabetes codes, and polycystic ovarian syndrome.