# Outcomes following a behaviour change intervention within hospitals to improve birth registrations and hospital utilisation for Aboriginal and/or Torres Strait Islander infants: a quasi-experimental and cohort study

## TL;DR

A behavior change intervention for hospital staff slightly reduced emergency visits for Aboriginal infants but had limited impact on birth registrations.

## Contribution

This study evaluates a targeted behavior change intervention to improve outcomes for Aboriginal infants in hospitals.

## Key findings

- A 38% reduction in emergency presentations for Aboriginal infants within 6 months after staff training.
- Little improvement in birth registrations or hospital admissions for Aboriginal infants.
- Low birth weight and late antenatal visits were linked to lower odds of birth registration.

## Abstract

The primary objective was to determine whether a behaviour change intervention delivered to hospital staff would (1) improve the proportion of Aboriginal and/or Torres Strait Islander (Aboriginal) babies being registered and (2) reduce hospital admissions and emergency presentations for babies <6 months old. The secondary objective was an observational analysis to determine factors that might influence the proportion of registered Aboriginal births in Western Australia (WA).

Quasi-experimental design and cohort study.

Five tertiary birthing hospitals in WA.

The intervention was delivered to health service providers who were in the five tertiary birthing hospitals. Outcome data were collected on Aboriginal babies born between 1 January 2016 and 30 June 2018 who were delivered within these hospitals. Babies in the control group (n=226) were born 6 months before the intervention and intervention babies (n=232) were born 6 months following the intervention. For the secondary objective, there were 4573 babies included in the analysis.

A behaviour change intervention delivered to hospital staff in five hospitals.

The primary outcomes were the proportion of babies who were registered and whether a baby had been admitted to hospital or an emergency department by 3 and 6 months old. The secondary outcome was to determine factors that might influence the proportion of registered Aboriginal births in WA (cohort study).

There was evidence of a 38% reduction in emergency presentations within 6 months for babies born to hospitals 6 months following the staff training (OR 0.62, 95% CI 0.42 to 0.91), and little evidence of improvements in birth registrations, hospital admissions within 3 or 6 months of birth or emergency department presentations within 3 months of birth. Of the 4573 babies included in the cohort study, 3769 (82.4%) babies had their births registered and 804 (17.6%) babies did not. Factors that were associated with not having a birth registered included low birth weight babies with a 34% decrease in odds of having a registered birth compared with those with a normal birth weight (adjusted OR (aOR) 0.66, 95% CI 0.51 to 0.86). Timing of first antenatal visit was associated with reduced odds of having a birth registered if this occurred in the second (aOR 0.77, 95% CI 0.64 to 0.93) or third trimester (aOR 0.59, 95% CI 0.45 to 0.77) compared with the first trimester.

Our study identifies the complexities surrounding birth registrations and improved hospital utilisation for Aboriginal babies, the importance of targeted interventions and ongoing efforts needed to address this issue comprehensively.

ACTRN12615000976583.