Literature used to inform relationship between surrogate outcomes and cases of maltreatment
| Intermediate outcome | Relationship | Relationship used in sensitivity analysis | Assumption | Abuse outcome | Reference | Interventions utilising this relationship |
|---|---|---|---|---|---|---|
| Injury in those less than age 1 | 5.17% of those with an injury up until age 1 also had child protection contact | 2.585% to 10.34%† | Each contact represents a new child | Child protection substantiation | Pratt 66 | 7, 27b |
| Hospitalisation for injury* | OR of 9.6 for child maltreatment allegation with every one admission for injury per year (indicates a 9.6 times greater likelihood of a maltreatment report for children hospitalised for injury) | OR 95% CI 7.86 to 11.83 | Baseline allegation rate of 0.0465 South Australia 2005/0667 | Child protection report | O'Donnell et al 12 | 23b, 20, 24b |
| Hospitalisation all causes | OR of 1.49 for child maltreatment allegation with every one admission per year (indicates a 1.49 greater likelihood of a maltreatment report for children hospitalised for all causes) | OR 95% CI 1.44 to 1.53 | Baseline allegation rate 0.0465 South Australia 2005/0667 | Child protection report | O'Donnell et al 12 | 15, 10, 26 |
↵* For this outcome hospitalisations and emergency department attendance are treated the same.
↵† No confidence limits are available as this is an actual rate. Therefore, the figures used in sensitivity analysis are double and half the rate reported in the original study. Wide limits have been used due to the smaller sample size used in this Australian cohort (11 000 children).