Table 3.
Conventional and spatial regression models to explain adjusted mean patient ages of Kawasaki disease and paediatric infectious diseases
| Variable* | Kawasaki disease | Exanthema subitum | Herpangina | HFMD | Chickenpox | PCF | Erythema infectiosum | GAS |
|---|---|---|---|---|---|---|---|---|
| Conventional regression | ||||||||
| Total fertility rate | −0·29 | −0·13 | −1·9 | −1·7 | −0·92 | −1·2 | −0·69 | n.s. |
| (P < 0·0001) | (P = 0·0207) | (P < 0·0001) | (P < 0·0001) | (P = 0·0053) | (P = 0·0019) | (P = 0·0159) | ||
| Health insurance | n.s. | −1·3 | −3·1 | −4·2 | −2·2 | −8·0 | −6·2 | −11 |
| (P < 0·0001) | (P = 0·0228) | (P = 0·0020) | (P = 0·0151) | (P < 0·0001) | (P < 0·0001) | (P = 0·0001) | ||
| Varicella | n.s. | n.s. | n.s. | n.s. | 1·1 | n.s. | n.s. | n.s. |
| Vaccination | (P = 0·0077) | |||||||
| R 2 | 0·34 | 0·45 | 0·56 | 0·58 | 0·62 | 0·46 | 0·44 | 0·29 |
| Spatial regression | ||||||||
| Total fertility rate | −0·26 | −0·16 | −0·18 | −0·16 | −0·66 | −1·0 | −0·62 | n.s. |
| (P < 0·0001) | (P = 0·0022) | (P < 0·0001) | (P < 0·0001) | (P = 0·0281) | (P = 0·0030) | (P = 0·0195) | ||
| Health insurance | n.s. | −0·15 | n.s. | −2·7 | n.s. | −6·3 | −5·7 | −11 |
| (P < 0·0001) | (P = 0·0253) | (P = 0·0002) | (P < 0·0001) | (P = 0·0001) | ||||
| Varicella | n.s. | n.s. | n.s. | n.s. | 1·4 | n.s. | n.s. | n.s. |
| Vaccination | (P < 0·0001) | |||||||
| R 2 | 0·41 | 0·50 | 0·58 | 0·66 | 0·62 | 0·53 | 0·45 | 0·29 |
HFMD, Hand, foot and mouth disease; PCF, pharyngoconjunctival fever; GAS, group A streptococcus; n.s., not significant.
*
Although five variables (total fertility rate, pupils per class, health insurance, waste process, and chickenpox vaccination) were incorporated into the multivariate analysis, only total fertility rate, health insurance, and chickenpox vaccination remained as statistically significant contributor(s).