Test accuracy per 1000 patients after each strategy for diagnose (11% prevalence of HIT)
| . | Gestalt . | 4Ts . | Immunoassay (based on IgG ELISA low threshold) . | 95% CI . | Functional assay . |
|---|---|---|---|---|---|
| Sensitivity | 0.969 | 0.921 | 0.98 | 0.95-0.99 | 1.0 |
| Specificity | 0.004 | 0.542 | 0.85 | 0.78-0.91 | 1.0 |
| Strategy | 1. Gestalt | 2. Gestalt, then IA positive, then FA | 3. 4Ts score (high or intermediate) | 4. 4Ts score (high or intermediate), then IA | 5. 4Ts score (high or intermediate), then IA positive, then FA (recommended strategy) |
| True positive. Patients will be appropriately treated and/or have more testing (reduced risk of thrombosis by 55%-70%). | 107 | 105 | 101* | 99 (high, 48; intermediate, 51) | 100 |
| False negative. Patients will be missed and may experience serious consequences of HIT (eg, thrombosis [300 of 1000 more people with HIT if not treated], amputation [60 of 1000 more people with HIT if not treated], death [60 of 1000 more people with HIT if not treated]). | 3 | 5 | 9 | 11 (high, 1; intermediate, 1) | 10 |
| True negative. Patients will appropriately not have more testing and will appropriately not be treated for HIT. | 4 | 890 | 482 | 829 (high, 34; intermediate, 313) | 890 |
| False positive. Patients will continue with unnecessary testing and/or may experience serious consequences of unnecessary treatment of HIT (eg, bleeding ∼8% to 35% over treatment duration) and may be falsely labeled as having HIT over the long term. | 886† | 0 | 408*† | 61 (high, 6; intermediate, 55) | 0 |
| No. of immunoassay tests performed | 509 | ||||
| No. of functional assay tests performed | 160 |
| . | Gestalt . | 4Ts . | Immunoassay (based on IgG ELISA low threshold) . | 95% CI . | Functional assay . |
|---|---|---|---|---|---|
| Sensitivity | 0.969 | 0.921 | 0.98 | 0.95-0.99 | 1.0 |
| Specificity | 0.004 | 0.542 | 0.85 | 0.78-0.91 | 1.0 |
| Strategy | 1. Gestalt | 2. Gestalt, then IA positive, then FA | 3. 4Ts score (high or intermediate) | 4. 4Ts score (high or intermediate), then IA | 5. 4Ts score (high or intermediate), then IA positive, then FA (recommended strategy) |
| True positive. Patients will be appropriately treated and/or have more testing (reduced risk of thrombosis by 55%-70%). | 107 | 105 | 101* | 99 (high, 48; intermediate, 51) | 100 |
| False negative. Patients will be missed and may experience serious consequences of HIT (eg, thrombosis [300 of 1000 more people with HIT if not treated], amputation [60 of 1000 more people with HIT if not treated], death [60 of 1000 more people with HIT if not treated]). | 3 | 5 | 9 | 11 (high, 1; intermediate, 1) | 10 |
| True negative. Patients will appropriately not have more testing and will appropriately not be treated for HIT. | 4 | 890 | 482 | 829 (high, 34; intermediate, 313) | 890 |
| False positive. Patients will continue with unnecessary testing and/or may experience serious consequences of unnecessary treatment of HIT (eg, bleeding ∼8% to 35% over treatment duration) and may be falsely labeled as having HIT over the long term. | 886† | 0 | 408*† | 61 (high, 6; intermediate, 55) | 0 |
| No. of immunoassay tests performed | 509 | ||||
| No. of functional assay tests performed | 160 |
IA, immunoassay; IgG, immunoglobulin G; FA, functional assay.
Eighty-nine of these patients would have a high 4Ts score, and 420 would have an intermediate 4Ts score.
When gestalt or 4Ts score is followed by IA and FA (when IA is positive), these patients would receive non-heparin anticoagulants unnecessarily for varying periods of time depending on timing of the follow-up tests.