Table 2.
Coefficients (median and 95% credible intervals (CrIs)) from TSIR models including school-term and meteorological covariates, over the period 1906–1920, for the three largest municipalities. Unique variance was estimated from multi-factorial additive linear regression fitted to the estimated (log) scarlet fever transmission rate. Note: coefficients reflect adjustment for other factors in the model.
| covariate | Amsterdam |
Rotterdam |
Den Haag |
|||
|---|---|---|---|---|---|---|
| Coef. | (95% CrI) | Coef. | (95% CrI) | Coef. | (95% CrI) | |
| extreme mean weekly temperature (at lag 14 weeks) | ||||||
| highest quintile | 7.0 × 10−02 | (6.0 × 10−02, 8.0 × 10−02) | 1.8 × 10−01 | (1.7 × 10−01, 1.9 × 10−01) | 4.1 × 10−02 | (2.8 × 10−02, 5.5 × 10−02) |
| lowest quintile | −1.1 × 10−02 | (−2.2 × 10−02, 1.5 × 10−03) | −8.2 × 10−02 | (−9.6 × 10−02, −6.8 × 10−02) | −4.0 × 10−02 | (−5.6 × 10−02, −2.6 × 10−02) |
| quintiles 2–4 | Ref. | Ref. | Ref. | |||
| unique variance (R2) | 0.42% | 0.35% | 0.21% | |||
| extreme mean weekly absolute humidity (at lag 14 weeks) | ||||||
| unique variance (R2) | 0.25% | 0.25% | 0.51% | |||
| cumul. winter prec. | 4.9 × 10−06 | (−3.1 × 10−06, 1.3 × 10−05) | 3.0 × 10−05 | (2.1 × 10−05, 4.0 × 10−05) | 7.8 × 10−05 | (6.7 × 10−05, 8.9 × 10−05) |
| unique variance (R2) | 1.55% | 0.26% | 12.7% | |||
| cumul. spring prec. | 1.9 × 10−05 | (9.3 × 10−06, 2.9 × 10−05) | −1.9 × 10−05 | (−3.1 × 10−05, −7.1 × 10−06) | −6.5 × 10−05 | (−8.0 × 10−05, −5 × 10−05) |
| unique variance (R2) | 1.05% | <0.01% | 4.63% | |||
| school term time | −8.7 × 10−02 | (−1.0 × 10−01, −7.5 × 10−02) | 1.8 × 10−01 | (1.7 × 10−01, 2.0 × 10−01) | 6.5 × 10−02 | (4.5 × 10−02, 8.4 × 10−02) |
| unique variance (R2) | 0.05% | 0.56% | 0.13% | |||