Early warning systems and rapid response to the deteriorating patient in hospital: A realist evaluation

Abstract

Aim: To test the Rapid Response Systems programme theory against actual practice components of the Rapid Response Systems implemented to identify those contexts and mechanisms which have an impact on the successful achievement of desired outcomes in practice.

Background: Rapid Response Systems allow deteriorating patients to be recognized using Early Warning Systems, referred early via escalation protocols and managed at the bedside by competent staff.

Design: Realist evaluation.

Methods: The research design was an embedded multiple case study approach of four wards in two hospitals in Northern Ireland which followed the principles of Realist Evaluation. We used various mixed methods including individual and focus group interviews, observation of nursing practice between June-November 2010 and document analysis of Early Warning Systems audit data between May-October 2010 and hospital acute care training records over 4.5 years from 2003-2008. Data were analysed using NiVivo8 and SPPS.

Results: A cross-case analysis highlighted similar patterns of factors which enabled or constrained successful recognition, referral and response to deteriorating patients in practice. Key enabling factors were the use of clinical judgement by experienced nurses and the empowerment of nurses as a result of organizational change associated with implementation of Early Warning System protocols. Key constraining factors were low staffing and inappropriate skill mix levels, rigid implementation of protocols and culturally embedded suboptimal communication processes.

Conclusion: Successful implementation of Rapid Response Systems was dependent on adopting organizational and cultural changes that facilitated staff empowerment, flexible implementation of protocols and ongoing experiential learning.

Keywords: early warning systems; enabling and constraining factors; escalation protocols; nursing; patient deterioration; patient recognition; rapid response systems; realistic evaluation.