Table 2.
Quartiles of age-standardised road-traffic mortality per 100 000 population, by exposure
| Quartile 17·6 to 16·7 (n=91) | Quartile 216·7 to 20·2 (n=92) | Quartile 320·2 to 25·5 (n=91) | Quartile 425·5 to 66·7 (n=92) | Wald test p value* | ||
|---|---|---|---|---|---|---|
| Population distribution | ||||||
| Population density in built-up area (inhabitants per km2)† | 6664 (5337 to 9763) | 6281 (5148 to 8230) | 5788 (4744 to 8234) | 6056 (4691 to 7846) | 0·5136 | |
| Annual population growth average (average annual number of people added to the city per annual 100 000 population)† | 1216 (885 to 1608) | 1168 (892 to 1508) | 1212 (876 to 1720) | 1323 (835 to 1804) | 0·7712 | |
| Social environment index | 0·42 (−0·04 to 0·70) | 0·30 (−0·19 to 0·52) | 0·09 (−0·46 to 0·46) | −0·21 (−0·55 to 0·11) | 0·0682 | |
| City gross domestic product (US$)† | 14 729 (10 051 to 20 624) | 14 729 (9848 to 20 046) | 14 272 (8732 to 19 476) | 10 903 (6653 to 16 676) | 0·0186 | |
| Urban landscape | ||||||
| Patch density (patches per 100 hectares) | 0·79 (0·37 to 1·52) | 0·53 (0·28 to 1·07) | 0·44 (0·25 to 0·85) | 0·41 (0·19 to 0·68) | 0·0482 | |
| Area-weighted mean nearest neighbour (isolation, metres) | 91·1 (85·6 to 98·9) | 88·6 (82·6 to 95·2) | 88·9 (82·4 to 94.6) | 88 (82·9 to 94·8) | 0·4342 | |
| Proportion of city built up | 5·2% (1·8 to 9·7) | 4·1% (1·9 to 6·9) | 2·9% (1·3 to 5·2) | 2·4% (1·2 to 4·7) | 0·1033 | |
| Street design | ||||||
| Intersection density (per hectare) | 7·2 (2·8 to 14·0) | 5·9 (2·8 to 9·4) | 3·9 (1·8 to 7·4) | 3·4 (1·7 to 6·4) | 0·0970 | |
| Average streets per node | 2·9 (2·8 to 3·2) | 3 (2·8 to 3·1) | 3 (2·9 to 3·1) | 3·1 (2·9 to 3·2) | 0·3900 | |
| Average street length (metres) | 126 (112 to 154) | 135 (121 to 166) | 135 (118 to 159) | 140 (123 to 180) | 0·2866 | |
| Transportation | ||||||
| Urban travel delay index | 0·17 (0·11 to 0·30) | 0·12 (0·08 to 0·21) | 0·11 (0·08 to 0·17) | 0·1 (0·06 to 0·15) | 0·1738 | |
| Presence of bus rapid transit system or subway | 26 (28·6%) | 14 (15·4%) | 9 (9·8%) | 4 (4·4%) | 0·0141 | |
Data are median (IQR) or n (%).
*
Tests if each quartile of age-standardised road mortality is equal to one another in a linear regression, with each covariate as the outcome and robust standard errors clustered by country.
†
2010 data.