Translating staff experience into organisational improvement: the HEADS-UP stepped wedge, cluster controlled, non-randomised trial.

Abstract

Frontline insights into care delivery correlate with patients' clinical outcomes. These outcomes might be improved through near-real time identification and mitigation of staff concerns. We evaluated the effects of a prospective frontline surveillance system on patient and team outcomes. Prospective, stepped wedge, non-randomised, cluster controlled trial; prespecified per protocol analysis for high-fidelity intervention delivery. Seven interdisciplinary medical ward teams from two hospitals in the UK. Prospective clinical team surveillance (PCTS): structured daily interdisciplinary briefings to capture staff concerns, with organisational facilitation and feedback. The primary outcome was excess length of stay (eLOS): an admission more than 24 hours above the local average for comparable patients. Secondary outcomes included safety and teamwork climates, and incident reporting. Mixed-effects models adjusted for time effects, age, comorbidity, palliation status and ward admissions. Safety and teamwork climates were measured with the Safety Attitudes Questionnaire. High-fidelity PCTS delivery comprised high engagement and high briefing frequency. Implementation fidelity was variable, both in briefing frequency (median 80% working days/month, IQR 65%-90%) and engagement (median 70 issues/ward/month, IQR 34-113). 1714/6518 (26.3%) intervention admissions had eLOS versus 1279/4927 (26.0%) control admissions, an absolute risk increase of 0.3%. PCTS increased eLOS in the adjusted intention-to-treat model (OR 1.32, 95% CI 1.10 to 1.58, p=0.003). Conversely, high-fidelity PCTS reduced eLOS (OR 0.79, 95% CI 0.67 to 0.94, p=0.006). High-fidelity PCTS also increased total, high-yield and non-nurse incident reports (incidence rate ratios 1.28-1.79, all p<0.002). Sustained PCTS significantly improved safety and teamwork climates over time. This study highlighted the potential benefits and pitfalls of ward-level interdisciplinary interventions. While these interventions can improve care delivery in complex, fluid environments, the manner of their implementation is paramount. Suboptimal implementation may have an unexpectedly negative impact on performance. ISRCTN 34806867 (http://www.isrctn.com/ISRCTN34806867).