**Background:** Family violence is a major public health issue, with intimate partner violence being the highest contributing risk factor to disease burden in women aged 25-44 in Australia. Healthcare services are recognized as critical settings for identifying and supporting victim/survivors, yet prior to this initiative, the Royal Melbourne Hospital had no family violence clinical response policy, no internal training, and no standardized screening method. The Strengthening Hospital Responses to Family Violence (SHRFV) initiative, launched by the Victorian state government, provided grant funding for public health services to implement whole-of-hospital transformational change. This study is the first to provide both pre- and post-implementation data evaluating the SHRFV program at a healthcare service.
**Methods:** The study was conducted at a large Tier 1 adult trauma hospital in Melbourne, Australia. All clinical staff with available work email addresses were invited to participate in an online survey (Nursing=1,829; Medical=660; Allied Health=549). The survey was open for six weeks, with up to three reminder emails over four weeks. The survey tool was the 'Assisting Patients/Clients Experiencing Family Violence: Royal Melbourne Hospital Clinician Survey', an 11-question instrument with Likert-type ordinal responses, forced-choice categorical responses, and qualitative free-text sections. Minor modifications at follow-up included adding questions on gender identity, age, specific types of RMH training attended, and estimated total training hours. The tool has demonstrated good internal consistency (Cronbach's alpha 0.77-0.83 in previous studies). Changes in ordinal outcomes were analyzed using Mann-Whitney U tests, and nominal outcomes using chi-square analyses, with a two-tailed alpha of 0.05.
**Key Results:** A total of 526 clinicians completed the follow-up survey (compared to 534 at baseline). Allied health clinicians comprised the largest group at follow-up (46.58%), followed by nursing (40.49%) and medical (12.93%). Overall response rate was 17.1% (allied health 44.63%, nursing 11.65%, medical 10.30%). At follow-up, 75% of clinicians reported some prior family violence training (vs. 65% at baseline), and 61% had training in the last two years (vs. 28% at baseline). Mean training hours at follow-up was 4.31 (SD 7.37, range 0-60). 48.67% endorsed completing at least one specific RMH training type since the project began. Statistically significant improvements were found in: knowledge ratings (moderate or above: 23.96% baseline vs. 55.7% follow-up; U=90,292.00, p<0.001), confidence ratings (moderate or above: 27.71% vs. 49.82%; U=97,115.50, p<0.001), and screening frequency (sometimes/often/always: 31.05% vs. 50.00%; U=105,074.50, p<0.001). Frequency of working with patients experiencing family violence did not change significantly (U=138,744.50, p=0.72). Allied health clinicians had higher knowledge and confidence mean ranks than nursing or medical staff (knowledge: H=21.16, p<0.001; confidence: H=9.50, p=0.009). Medical staff reported working with affected patients more frequently (H=11.03, p=0.004). All three specific clinical skills (knowledge of indicators, asking about family violence, responding to disclosures) showed significant improvement at follow-up. All 13 pre-specified barriers showed reduced endorsement, with five reaching statistical significance: 'I don't know what to do or say' (48.74% to 27.76%; p<0.001), 'concern about offending/affecting rapport' (44.29% to 33.27%; p=0.001), 'lack of supporting policies/procedures' (35.59% to 9.89%; p<0.001), 'concerns about staff safety' (25.34% to 14.26%; p<0.001), and 'little/no access to supervision' (23.79% to 11.41%; p<0.001). The three most endorsed barriers remained the same: perpetrator present (58.41% to 48.67%), patient reluctance to disclose (54.16% to 48.29%), and time limitations (51.54% to 46.58%).
**Clinical Implications:** This study provides evidence that a comprehensive, multi-faceted transformational change project can produce meaningful improvements in healthcare worker family violence knowledge and skills within a large hospital setting. The significant reduction in barriers related to policies and procedures (from 35.59% to 9.89%) reflects the successful implementation of formal clinical infrastructure. However, the persistence of key barriers (perpetrator presence, patient reluctance, time limitations) and the finding that confidence improvements lagged behind knowledge gains suggest that training alone is insufficient. The fact that 25% of clinicians still had no family violence training at follow-up, and that the COVID-19 pandemic disrupted training delivery (4,309 attendances in 2019 vs. 1,089 in 2020), underscores the need for sustained, ongoing investment. The study's limitations include a 17.1% response rate, cross-sectional design preventing tracking of individual change, and the confounding effect of COVID-19 on the follow-up data collection. Future research should include audits of electronic medical record screening tool uptake, patient outcome evaluations, and test-retest validity studies of the survey tool.