Cushing's pretest score
Estimates the pretest probability of canine hyperadrenocorticism from signalment, clinical signs, and routine bloodwork. Useful as a gate before confirmatory testing. A very low score makes Cushing's a poor next move; a very high score keeps it plausible despite an equivocal test.
Enter clinical signs and labs to compute pretest probability.
How the score works
The tool was derived from a multivariable binary logistic regression model developed in 398 Cushing's cases and 541 noncases from the UK VetCompass primary-care database. Each predictor's regression coefficient was divided by a common factor (the smallest significant optimism-adjusted coefficient, for "USG not recorded" = 0.38) and rounded to the nearest integer to produce the point values in Table 5.
The predicted likelihood for each total score is calculated from the inverse logit of the linear predictor: LP = β + B × (points total), where β = −0.42 (optimism-adjusted constant) and B = 0.38 (common factor). The model demonstrated good discrimination (AUROC = 0.78, optimism-adjusted 0.76) and calibration (C-slope = 0.86).
Important: the tool was developed in dogs already suspected of Cushing's by their attending veterinarian, it is a decision aid at the point of first suspicion, not a screening tool for the general population. External validation has not yet been performed.
Score → likelihood table
Table 6, Schofield et al. 2020.
| Score | Predicted likelihood | Score | Predicted likelihood |
|---|---|---|---|
| −13 | 0% | −1 | 27% |
| −12 | 1% | 0 | 35% |
| −11 | 1% | 1 | 44% |
| −10 | 1% | 2 | 53% |
| −9 | 2% | 3 | 63% |
| −8 | 3% | 4 | 71% |
| −7 | 4% | 5 | 78% |
| −6 | 5% | 6 | 84% |
| −5 | 8% | 7 | 88% |
| −4 | 11% | 8 | 92% |
| −3 | 15% | 9 | 94% |
| −2 | 20% | 10 | 96% |
Sources
- Schofield I, Brodbelt DC, Niessen SJM, Church DB, Geddes RF, Kennedy N, O'Neill DG. Development and internal validation of a prediction tool to aid the diagnosis of Cushing's syndrome in dogs attending primary-care practice. J Vet Intern Med. 2020;34:2306–2318. doi:10.1111/jvim.15851
- Open access article. Scoring values implemented directly from Tables 5 and 6 of the paper. This tool has not been externally validated. Performance may differ in populations with different Cushing's prevalence than the VetCompass UK primary-care cohort.