HRs for El-Escorial criteria after inclusion in final model
| El-Escorial category | HR | 95% CI | Wald test |
|---|---|---|---|
| Suspected | 0.86 | 0.68 to 1.08 | 0.199 |
| Possible | 1 | – | – |
| Probable—laboratory supported | 1.38 | 1.31 to 1.44 | <0.001 |
| Probable | 1.49 | 1.39 to 1.60 | <0.001 |
| Definite | 2.09 | 1.99 to 2.21 | <0.001 |
| Sex/site of onset/C9orf72 | |||
| Spinal onset excluding male C9orf72 expanded cases | 1 | – | – |
| All bulbar onset cases | 1.33 | 1.21 to 1.45 | <0.001 |
| Male C9orf72 expanded cases only | 1.57 | 1.26 to 1.96 | <0.001 |
HRs for each of the El-Escorial criteria from a Royston-Parmar model on the hazard scale with 3 degrees of freedom and including age of onset (time varying), diagnostic delay, site of onset, country and sex-site-C9 as a hybridised variable with a variance–covariance matrix allowed to pool by country. Note that Belgium omitted as the El-Escorial criteria was not available, the UK data included only Probable and Definite category ALS, while Ireland does not use the ‘probable laboratory supported’. Nevertheless, HRs are in line with expectations showing a gradually increasing hazard with increasing category severity.