STEPP IN: A Multicenter Quality Improvement Collaborative Standardizing Postoperative Handoffs

Participation criteria included CHNC membership and agreement to contribute data for the entire project. Twenty CHNC hospital NICUs joined and were committed to improving postoperative handoff. The project management team consisted of interdisciplinary experts from 14 CHNC hospitals including nursing, neonatology, anesthesiology, pediatric surgery, and process improvement experts familiar with QI and guideline development methodology. Systematic literature review and expert opinion helped to develop a handoff clinical practice recommendation (CPR) reference document (Supplemental Fig 5), using the principles of the AGREE II instrument. It was not intended to be adopted as a bundle but rather a framework for sites to choose individual practices that best fit their local environment.¹⁶  Practice recommendations were specifically focused on standardizing postoperative handoff communication, including components of effective communication tools and required participants.

The collaborative intervention was primarily focused on implementation of a structured postoperative handoff process. The project provided infrastructure to help centers develop, implement, and evaluate and monitor compliance to processes. Teams were expected to review the CPR and engage local stakeholders to develop systems and processes for an ideal handoff as described by Sandlin and Segall et al.^17,18  The primary drivers included team member participation; standard communication method; and providers’ receipt of accurate, complete information (Supplemental Fig 5). The project was focused on highest-risk infants who returned from the operating room intubated.

The QI framework, modeled on previous CHNC QI collaboratives and the Institute for Healthcare Improvement’s Model for Improvement, was focused on team learning and sharing as well as communication and support (monthly webinars, bimonthly huddle calls, an active listserv, and QI faculty advisors).^19–22  The Institute for Healthcare Improvement extranet supported data reporting and document sharing allowing teams to track data and monitor improvement. Monthly webinars included QI education, collaborative data updates, and local team presentations to share barriers and successes including Plan-Do-Study-Act (PDSA) cycle interventions. Bimonthly huddles fostered small group discussions. Assigned faculty advisors supported center progress.

Centers were expected to implement handoff processes for all patients who underwent any type of surgery and ventilated postoperatively regardless of type or timing of the procedure, patient demographics, or where the surgery occurred. Discussions on monthly team calls and huddle meetings reinforced the importance of random monitoring including weekdays and weekends as well as day and night shifts. Patients who underwent induced hypothermia for surgical indications were excluded. The sampling plan was by local audits and self-reports on outcome and process measures for at least 10 surgeries per month.²³  If the surgical volume was <10 per month, centers were expected to monitor all surgeries for that month.

The primary outcome of care failures was defined as missing information at the time of postoperative handoff that directly affects the patient or the failure of knowledge transfer that impacts the staff’s ability to care for the patient.^24,25  The care failures were self-reported and identified within the first 24 hours after surgery. These care failures were divided into 2 categories: respiratory and all other communication failures and recorded categorically as present or not. Respiratory care failures included the communication of elements directly related to respiratory care, such as endotracheal tube position, status of endotracheal tube cuff inflation, ventilator settings and respiratory medications. All other issues were categorized as communication care failures, such as inadequate communication of intraoperative fluids, medications, complications, and care of wounds or drains. To reinforce understanding of the definition of a care failure, huddles and monthly meetings offered regular reviews and an extensive list of examples for teams to reference (Supplemental Fig 5).

There were 3 process measures: center-defined staff presence at handoff, use of a center-defined handoff tool, and interruptions (defined as any pause that interfered with handoff excluding emergent patient care). The goal minimum compliance with staff presence and use of a handoff tool was >80%, with an ideal target of ≥90%. We aimed to eliminate all handoff interruptions. The denominator for all measures was total number of handoffs monitored. The numerator for the outcome measure was the number of handoffs in which a care failure was identified. The numerator for each process measure was number of handoffs free of interruptions, number of times center used the handoff tool, or number of handoffs in which required staff were present. We used statistical process control (SPC) rules to analyze process measures. Team member satisfaction with the handoff process served as the balancing measure which was assessed by a standard 5-point Likert scale, with 5 being the highest satisfaction score (the number of participants with 4 or 5 score over total number of responding participants). Each center developed processes for assessment that were completed by the individual staff member, anonymous, and separate from the other handoff audits.

Baseline handoff data were collected for 8 months starting with the project rollout in October 2013. Teams initiated data collection as they engaged local stakeholders, mobilized their teams, and planned and instituted improvements. By June 2014, with the majority of teams having stable infrastructure, the collaborative proceeded into the 12-month action phase. During this period, teams focused on compliance monitoring, data sharing, and ongoing PDSA cycles to achieve targeted goals. Centers submitted monthly outcome and balancing measures as well as process compliance data. Collaborative respiratory and communication care failures were analyzed as a time-series variable by using SPC charts (Shewhart P charts). Special cause signals were identified by using standard control chart rules.²⁶  Ongoing collaborative sharing and learning continued for the 7-month sustain phase (June 2015 to December 2015).

Children’s Mercy Hospital, Kansas City, Missouri, Pediatric Institutional Review Board reviewed and determined that the project did not meet the definition of research involving human subjects.

Nineteen of 20 participating centers completed data reporting. Each center had unique barriers to overcome, including team assembly, development of handoff process, and staff engagement and education. Before implementation of the interventions, centers did not have processes nor monitoring systems in place to capture baseline data. After project rollout, centers developed and implemented new processes and instituted baseline data collection. Some centers were quicker than others to ramp up process change and monitoring systems. The baseline period continued through May 2014, to allow all centers to develop consistent data monitoring and reporting. During the action period, on average 181 (158–210) postoperative surgical handoffs and 320 (222–416) handoff satisfaction assessments were reported per month. In total, 4230 handoffs were assessed (baseline: 975; action: 2175; and sustain: 1080) (Fig 1).

Within the first 8 months, both care failures demonstrated special cause improvement signals. Communication care failures decreased by 49.4% from 17% to 8.6% (Fig 2). Respiratory care failures decreased by 43.9% from 8.2% to 4.6% and then another 52.2% to 2.2% after a second special cause signal for a total decrease of 73.2% from baseline through sustain (Fig 3). In total, 84% of 5626 staff members surveyed rated the handoff ≥4 on a 5-point Likert scale, indicating satisfaction with the process, which was maintained throughout the entire collaborative without a special cause signal for this balancing measure. High satisfaction assessments continued through the sustain period with that last 3 months reported 92%, 96%, and 94% satisfaction (Supplemental Fig 6).

Compliance with use of the handoff tool increased from a baseline mean of 74% to 80% by using SPC rules. For required staff in attendance at the postoperative handoff, compliance increased from 67% to a center line mean of 78% after a special cause signal in April 2015. The targeted compliance for these process measures was 80%. Handoff interruptions decreased over the project time frames to <5% (Fig 4). These improvements are reflective of individual centers’ PDSA cycles. Examples of local team interventions are shown in Table 1.

The STEPP IN multicenter collaborative achieved sustained success in improving patient safety for a high-risk surgical neonatal population. The period between verbal handoff and patient transfer is a particularly vulnerable time when patient care errors occur.²⁷  The impact of these errors may be amplified in the NICU perioperative setting. Participating CHNC centers improved handoff processes for >150 high-risk patients each month, demonstrating a decrease in respiratory and communication care failures. The strengths of the project included the engagement of interdisciplinary team members, implementing handoff processes previous not present, and using a handoff tool in a large cohort sample.

STEPP IN provided the QI framework that has been successfully used in previous projects and an evidenced-based CPR to support local CHNC NICUs to introduce a perioperative handoff process.¹⁹  The structure supported sharing ideas, lessons learned, and benchmarking, which could be adopted, tailored, and spread. Our infrastructure allowed centers to develop individualized processes with flexibility to promote success within an individual culture while achieving a shared improvement outcome. Every center worked within their ability to implement change. The CPR provided a resource for supportive literature and best practices and defined the characteristics of an ideal handoff. The ideal handoff incorporates receiving team preparation, inclusion of relevant team members, completion of critical tasks before handover, and concise information exchange specifically to allow only 1 provider to speak with adequate time for questions.^17,18 

A handoff tool can augment safe transfer of care. The implementation of the tool and associated processes depends on engaged team members. The collaborative expected STEPP IN centers to involve a team of stakeholders from all disciplines (surgery, anesthesia, and neonatology) caring for the patient during the perioperative course. There is strength in a multi- and interdisciplinary approach in detecting and preventing errors by using collaborative cross-checks among peers.¹⁰  Inclusion of different types and levels of health care providers enhances the handoff process.²⁸  In addition, limiting handoff interruptions is another key component to ensure exchange of information.

During the webinars and huddle calls, centers shared processes for engagement and team building. Members recognized that full multidisciplinary engagement removed barriers to optimal information exchange. Teams used creative ideas to educate frontline team members on handoff expectations (eg using influential team members in educational videos). Specific challenges that were reported included timely team member notification, supporting processes throughout surgical subspecialties, and education needs related to staff turnover and rotating trainees (Table 1). Key best practices reported multidisciplinary team engagement and buy-in to the value of a structured handoff, development of communication systems, member introductions at handoff, establishing a culture where interruptions were eliminated, and use of a structured but dynamic handoff tool.

Single-center studies recognizing the importance of handoffs have revealed error reduction, improved handoff process, and better clinical outcomes.^7,8,12,29,30  Furthermore, a multicenter project using I-PASS revealed reduction in medical errors and preventable adverse events without negatively impacting workflow.³  STEPP IN used these principles to optimize structured handoff in a larger number of centers with a unique population.

A common concern when implementing a standardized handoff is the perceived impact on workflow disruption. Comparing implementation of a standardized handoff protocol in a surgical ICU between pre- and postinterventions, Joy et al reported improved teamwork but no change in handoff duration.⁹  Sheth described improved efficiency with an I-PASS supported handoff process as well as improved scores for the handoff/transitions domain of a national culture of safety survey.²⁹  Satisfaction, as a balancing measure, was selected to assess how members valued the new handoff process weighed against time away from other essential work. High satisfaction with perioperative handoff in our study confirmed that a structured process enhances care without negative impact.

Success for QI projects lies in sustaining the outcome and spread to other settings. During the sustain phase, we maintained the collaborative meetings, shared experiences, and collegial relationships that were crucial component of ongoing success. Feedback from centers indicated that many adopted similar processes in other areas of their hospitals.

This collaborative intended to develop and implement a standardized approach for postoperative handoffs. Before STEPP IN, participating centers lacked structured handoffs. It is the processes that are the actionable items and critical measures for driving QI to identify and address problems that were not previously recognized.³¹ Implementation of a handoff process was not only integral to perioperative care, it was the actionable item critical to improvement success. However, the lack of a priori processes at any of the institutions to capture baseline data was a recognizable challenge. The short baseline time period and sample size during the initial data collection period may be considered a limitation. During the baseline period, each center had different resources and capabilities to begin reporting data and implementation of interdisciplinary handoff processes. We speculate the results underestimate the actual degree of improvement for the collaborative; however, it represents the contributions from and improvement in every center. Change quickly followed process implementation as evident by improvement during the action period and a second special cause signal for respiratory communication failure.

The potential inconsistencies in care failure capture would also be a limitation. We implemented serial education and open discussions on monthly meetings and huddle calls to align understanding of these definitions. Although granular details of care failures from sites were not tracked, some examples teams shared include error in reporting antibiotics administration, duplicate medication delivery, discrepant reporting of tracheostomy, and unexpected postextubation stridor in patients who was retrospectively noted to have intubation. Care failure capture rate may be subject to information or selection bias. Although this may have resulted in an over- or underestimated event rate, it was consistent throughout the project and within centers. Each center implemented processes to recognize care failures and then adapt to correct them. Nonrandom sampling, based on when the surgery occurred or related surgical complexity could also have impacted results. We did not track time of day or day of the week surgeries were monitored, but there was expectation to include nights, days, weekends, and weekday handoffs to assure representative cross-section of all surgical handoffs. Participating centers were expected to monitor a minimum of 10 surgeries per month per site, the minimum number approximated the greatest precision in gathering meaningful information while maintaining an acceptable level of workload.²³  Despite these potential biases, it is unlikely they impacted team improvement nor would alter our conclusions.

STEPP IN provided the QI framework to support local CHNC NICUs to improve the perioperative handoff process and decrease postoperative care failures with high participant satisfaction scores and evidence of sustainability. This work may be generalizable to other perioperative populations as well as other high acuity transfers of care.

Jeanette M. Asselin, MS, RRT-NPS, David J. Durand, MD (emeritus), Francine D. Dykes, MD (emeritus), Jacquelyn R. Evans, MD (Chair), Karna Murthy, MD, Michael A. Padula, MD, MBI, Eugenia K. Pallotto, MD, MSCE, Kristina M. Reber, MD, and Billie Lou Short, MD, members of CHNC. CHNC (http://www.thechnc.org) partnered with Children’s Hospital Association, Inc. (CHA; Overland Park, KS) to design, launch, and maintain the Children’s Hospital Neonatal Database (CHND). Children’s Hospital Association provided administrative and analytic support during this project. We would like to acknowledge Kate Conrad, Lorna Morelli and Tina Logsdon from Children’s Hospital Association whose support made this project a success.

We are indebted to the following institutions that serve the infants and their families, and these institutions have participated in this collaborative. Alfred I. duPont Hospitals for Children, Wilmington, DE (Judith C. Guidash, BSN, RNC-NIC; Kevin Sullivan, MD); All Children’s Hospital John Hopkins Medicine, St Petersburg, FL (Oscar Winners-Mendizabal, MD; Victor Mckay, MD; Corrie Long, RN, MSN); Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL (Karna Murthy, MD MSc; Megan Bankus, RN; Molly Schau, RN; Lisa Krick, RN); Arkansas Children’s Hospital, Little Rock, AK (Becky Rogers, MD; Allen Harrison, MD, BSN; Francesca Miquel-Verges, MD); Children’s Healthcare of Atlanta, Atlanta, GA (Sarah Keene, MD; Sarah Hash, BSN, RN, RNC-NIC); Children’s Hospital and Medical Center, Omaha, NE (Lynne Willett, MD); Children’s Hospital Colorado, Aurora, CO (Susan Moran, NNP; Sheila Kaseman, MS, RNC-NIC); Children’s Hospital of Los Angeles, Los Angeles, CA (Steven Chin, MD, MSc, Julie Evans, MISM, BSHCM, RN, HACP, James Stein, MD, MSc); Children’s Hospital of Michigan, Detroit, MI (Girija Natarajan, MD; Jay Ann Nelson, RN; Monika Bajaj, MD); Children’s Hospital of Pittsburgh of UPMC, Pittsburgh, PA (Trishelle Himmelrick, MSN, CCRN; Michael Scott, CNRA; Sara Angelelli, MSN, RN, CNOR); Children’s Medical Center, Dallas, TX (Becky Ennis, MD; Kerry Wilder, BSN, RN, MBA; James Moore, MD, PhD); Children’s Mercy-Kansas City, Kansas City, MO (Denise Smith, MSN, RN, NNP-BC; Meredith Kopp BSN, RN, CNOR; Susan J. Whitney, MD); Children’s National Medical Center, Washington, DC (Lamia Soghier, MD, MEd; Joyce Doering, RN); Florida Hospital for Children, Orlando, FL (Kathryn Mikulencak, MSN, PCNS-BC, RNC-NIC; Rajan Wadhawan, MD); Le Bonheur Children’s Hospital, Memphis, TN (John Ferguson, MD; Bobby Bellflower, DNSc, NNP-BC; Ramasubbareddy Dhanireddy, MD); Nationwide Children’s Hospital, Columbus, OH (Thomas Bartman, MD, PhD; Margaret Holston, BSN, RN); Primary Children’s Hospital, Salt Lake City, UT (Robert DiGeronimo, MD; Shrena Patel, MD; Cindy Spencer, RN); Rady Children’s Hospital, San Diego, CA (Brian Lane, MD; Ellen Knodel, RRT-NPS; Mark Speziale, MD); St Louis Children's Hospital, St Louis, MO (Renee Fishering, NNP-BC; Kristin Klockenga, RN; Amy Distiller, RN); The Children’s Hospital of Philadelphia, Philadelphia, PA (Missy Duran, MSN; Tamara Meeker, RN MSN, NNP-BC)

CHNC

Children’s Hospitals Neonatal Consortium

CPR

clinical practice recommendation

PDSA

Plan-Do-Study-Act

QI

quality improvement

SPC

statistical process control

STEPP IN

Safe Transitions and Euthermia in the Perioperative Period in Infants and Neonates