Table 4. Predictors of 1- and 5-year hospital care costs after index TIA or stroke.
| 1-year costs (log transformed) |
5-year costs (log transformed) |
|||
|---|---|---|---|---|
| Coef. | p>|z| | Coef. | p>|z| | |
| Gender | ||||
| Female | Reference case | Reference case | ||
| Male | 0.14 | 0.274 | -0.01 | 0.924 |
| Age, years | ||||
| <65 | -0.14 | 0.531 | -0.51 | 0.003 |
| 65-74 | -0.10 | 0.466 | -0.18 | 0.146 |
| ≥75 | Reference case | Reference case | ||
| Previous history of: | ||||
| TIA | -0.04 | 0.815 | -0.05 | 0.747 |
| Stroke | 0.13 | 0.356 | 0.32 | 0.011 |
| Peripheral vascular disease | 0.35 | 0.066 | 0.26 | 0.137 |
| Myocardial infarction | 0.31 | 0.060 | -0.04 | 0.775 |
| Angina | -0.09 | 0.545 | -0.02 | 0.867 |
| Hypertension | -0.16 | 0.167 | -0.17 | 0.085 |
| Atrial fibrillation | 0.28 | 0.030 | 0.09 | 0.476 |
| Disability | 0.20 | 0.228 | -0.02 | 0.908 |
| Event type | ||||
| TIA | Reference case | Reference case | ||
| Stroke | 0.39 | 0.007 | 0.16 | 0.195 |
| NIHSS score | 0.20 | <0.001 | 0.13 | <0.001 |
| NIHSS score ^ 2* | -0.006 | <0.001 | -0.005 | <0.001 |
| Subsequent events: | ||||
| Stroke | 0.69 | <0.001 | 0.43 | <0.001 |
| Coronary | 1.16 | <0.001 | 0.54 | <0.001 |
| Peripheral vascular | 0.92 | 0.103 | 0.67 | 0.081 |
| Pre-morbid place of residence | ||||
| Own home | Reference case | Reference case | ||
| Home friend/family | 0.01 | 0.976 | -0.33 | 0.346 |
| Warden accommodation | 0.26 | 0.342 | 0.55 | 0.009 |
| Care home | -0.97 | 0.072 | -1.57 | 0.002 |
| Lived alone before event | 0.38 | 0.036 | 0.06 | 0.718 |
| Pre-morbid marital status | ||||
| Married | Reference case | Reference case | ||
| Widowed | -0.03 | 0.882 | 0.06 | 0.757 |
| Single | 0.51 | 0.066 | 0.55 | 0.041 |
| Separated | 0.13 | 0.637 | 0.02 | 0.946 |
| Living with partner | -0.17 | 0.736 | 0.06 | 0.822 |
| Pre-morbid employment status | ||||
| Full-time | -0.12 | 0.611 | -0.04 | 0.820 |
| Part-time | -0.04 | 0.872 | 0.15 | 0.472 |
| Caring for home | 0.03 | 0.935 | -0.01 | 0.982 |
| Unemployed | -0.95 | 0.376 | -0.44 | 0.300 |
| Unable-work | 0.65 | 0.088 | 0.16 | 0.632 |
| Retired | Reference case | Reference case | ||
| Socio-economic status | ||||
| Professional | -0.13 | 0.626 | 0.22 | 0.362 |
| Managerial | -0.03 | 0.847 | 0.04 | 0.788 |
| Skilled non-manual | -0.06 | 0.749 | -0.12 | 0.478 |
| Skilled manual | Reference case | Reference case | ||
| Partly skilled | -0.17 | 0.411 | -0.04 | 0.828 |
| Non-skilled | 0.03 | 0.881 | 0.29 | 0.082 |
| Age left education | -0.02 | 0.150 | -0.04 | 0.046 |
| Index of multiple deprivation | 0.001 | 0.951 | 0.003 | 0.637 |
| Constant | 0.72 | <0.001 | 9.02 | <0.001 |
| No. | 756 | No. | 756 | |
| p>F | <0.001 | p>F | <0.001 | |
| Adj R2 | 0.241 | Adj R2 | 0.175 | |
*The non-linear association between event severity and costs was further tested by including an interaction term between severity and case-fatality (defined as mortality within 30 days of event) into the regression. We found this to be statistically significant, although the non-linear relationship between cost and severity still held.