Variation of case fatality rates with different methods of identifying cases
| GP | Hospital | Paired GP/hospital | Pooled GP/hospital | p Value | |
| Myocardial infarction | |||||
| N | 145 | 171 | 105 | 209 | |
| 30-day case fatality rate (%) | 4 | 20 | 6 | 17 | 0.001 |
| Ischaemic heart disease | |||||
| N | 362 | 529 | 270 | 585 | |
| 30-day case fatality rate (%) | 2 | 9 | 3 | 8 | 0.002 |
| Cerebrovascular disease | |||||
| N | 302 | 330 | 153 | 424 | |
| 30-day case fatality rate (%) | 6 | 16 | 5 | 10 | 0.001 |
The 30-day case fatality rates for myocardial infarction, ischaemic heart disease and cerebrovascular disease, identified using GP, hospital, paired GP/hospital and pooled GP/hospital data. The significance of the differences between coding methods is adjusted for confounding factors using logistic regression (see text for details).
GP, general practitioner.