Table 4.
Models for binary logistic regression of factors associated with initial ED visit (N = 16 543)
| ED visit | ||||
| Bivariate | Model 1 | Model 2 | Model 3 | |
| OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Individual factors | ||||
| Demographics and health | ||||
| Age | 1.08 (1.07 to 1.09)* | 1.09 (1.08 to 1.09)* | 1.06 (1.05 to 1.06)* | 1.06 (1.05 to 1.07)* |
| Gender | ||||
| Female | Ref | |||
| Male | 1.04 (0.97 to 1.11) | |||
| No of chronic diseases | 1.22 (1.20 to 1.25)* | 1.24 (1.21 to 1.26)* | 1.12 (1.10 to 1.15)* | 1.12 (1.10 to 1.15)* |
| Home help and healthcare use | ||||
| Home help | ||||
| No | Ref | Ref | Ref | |
| Instrumental services | 1.74 (1.51 to 2.01)* | 1.31 (1.13 to 1.52)* | 1.26 (1.08 to 1.46)* | |
| Personal care | 2.00 (1.71 to 2.34)* | 1.49 (1.26 to 1.76)* | 1.55 (1.31 to 1.83)* | |
| Instrumental and personal care | 3.38 (3.10 to 3.68)* | 2.49 (2.27 to 2.74)* | 2.48 (2.26 to 2.73)* | |
| No of primary care visits | 1.05 (1.04 to 1.05)* | 1.03 (1.03 to 1.04)* | 1.03 (1.03 to 1.04)* | |
| No of specialist care visits | 1.02 (1.01 to 1.02)* | 1.00 (1.00 to 1.01) | 1.00 (1.00 to 1.01) | |
| No of hospital admissions | 1.37 (1.32 to 1.42)* | 1.01 (0.96 to 1.05) | 1.00 (0.96 to 1.05) | |
| Contextual factors | ||||
| Annual social care cost per person 80+ | 1.00 (1.00 to 1.00) | |||
| Response, trust and safety | 1.00 (0.99 to 1.00) | |||
| Influence and adequate time | 1.00 (1.00 to 1.01) | |||
| Median days in residential care | 0.99 (0.99 to 0.99)* | 1.00 (0.99 to 1.00) | ||
| Proportion of 80+ in the population | 0.94 (0.91 to 0.98)* | 1.09 (1.05 to 1.13)* | ||
| Distance to the nearest ED | 0.99 (0.99 to 0.99)* | 0.99 (0.98 to 0.99)* | ||
| Deviance (−2LL) | 20 624.749 | 19 883.865 | 19 701.645 | |
| Cox and Snell R2CS | 0.062 | 0.103 | 0.113 | |
| Nagelkerke R2N | 0.085 | 0.142 | 0.155 | |
Note: The negative bivariate association for proportion of 80+ in the population with initial ED visit becomes a positive association and increases in magnitude in the multivariable model, indicating negative suppression. Through trial variable removal from the model,30 it was determined that distance to the ED and median days in residential care both acted as suppressor variables for proportion 80+ in the population.
*Significant on values estimated from 95% CI.
ED, emergency department.