Identifying improvements for delivery room resuscitation management: results from a multicenter safety audit

An estimated 10 percent of all newborns require assistance breathing after birth,
with one percent of infants needing more complicated resuscitation measures 1]. In particular, very low birth weight infants, those weighing less than 1500 grams
at birth, are more likely to require intubation, chest compressions, or administration
of medications during initial resuscitation 2] and subsequent admission to a neonatal intensive care unit (NICU). High-quality care
during resuscitation lays the foundation for a successful transition to the NICU and
extrauterine life.

Delivery room resuscitation and stabilization is a complex activity involving multiple
health care professionals from different disciplines 3]. Effective teamwork is essential to achieve optimal outcomes in health care 4]-9]. A call to action for quality patient care in labor and delivery, endorsed by seven
influential professional organizations, specifically recommends training clinicians
in the principles of teamwork and shared decision-making to improve outcomes 10]. Thomas and colleagues identified optimal teamwork behaviors for neonatal resuscitation
including sharing information, evaluating plans, and prioritizing and distributing
the workload among team members 11]. Teaching team behaviors in conjunction with a skills-based curriculum such as the
Neonatal Resuscitation Program (NRP; 12]) can significantly improve teamwork 13]-18] and quality of care 13],15]. To facilitate effective teamwork, the 2010 International Consensus Guidelines on
Neonatal Resuscitation recommend briefing and debriefing, the process of reviewing
and communicating pertinent facts about the resuscitation before and after events
19]. The most recent revision of the NRP added simulation-based education and post-scenario
debriefing to help trainees practice teamwork skills 12].

An additional important step to improving team behaviors is systematic documentation
and evaluation of standard teamwork practices 9]. To function effectively, team members need to know who should be present at a neonatal
resuscitation, each team member’s roles and responsibilities, and the skills required
to fulfill those roles 9]. To reinforce optimal teamwork behaviors, briefing and debriefing should be a standard
part of every resuscitation procedure. Documenting such policies and guidelines reinforces
organizational support for quality care and emphasizes a culture of safety 4],5],9].

As part of ongoing efforts to help patient care teams improve the quality and safety
of medical care for newborn infants and their families, Vermont Oxford Network, a
non-profit voluntary collaboration of health care professionals established in 1988
20], coordinated an online VON Days quality audit of guidelines and practices for teamwork
during delivery room resuscitation. The quality audit was part of iNICQ 2012, an Internet-based
improvement collaborative in which multidisciplinary teams from around the world worked
together under the guidance of expert faculty to make measurable improvements in the
quality and safety of respiratory care. A key component of this collaborative, as
with other face-to-face collaboratives organized by VON, was the use of potentially
better practices (PBPs), practices derived from the best evidence available, whose
implementation through Plan-Do-Study-Act cycles is likely to improve the processes
or outcomes of interest. PBPs are improvement ideas with logical appeal and practical
application that can be implemented, tested, and measured at the local level after
modification for local context 21]-23]. We call these evidence-based practices “potentially better” rather than “better”
or “best” because until the practices are evaluated, customized, and tested at the
local level, teams will not know if they are truly better.

iNICQ faculty identified three potentially better practices (PBPs) for delivery room
resuscitation and respiratory care of very low birth weight infants during the first
hour of life: teamwork and communication in the delivery room, use of a team approach
to respiratory care during the first hour of life, and maintenance of normal temperature
in very low birth weight infants. Over the course of the 2012 iNICQ, clinical and
quality improvement expert faculty conducted five web-based sessions. Before each
iNICQ session, teams completed a VON Day Audit to assess local practices.

The VON Day Audit for resuscitation focused on the first PBP, improving teamwork and
communication in the delivery room. Of 144 centers eligible for this audit, 84 centers
participated (Additional file 1). The audit asked about guidelines or policies at the unit level and whether they
were applied at the infant level. Unit level policies or guidelines included: which
deliveries required attendance by an individual or team responsible for the infant
after delivery; required personnel; required training; pre-delivery briefings; debriefings;
and family communication after delivery. Data collectors reported whether resuscitation
teams attending deliveries routinely used checklists to ensure the quality and safety
of the deliveries, whether the unit had a program for simulation-based neonatal resuscitation
training, and whether the unit routinely videotaped deliveries/resuscitations of high-risk
infants.

Each unit was asked to evaluate up to ten deliveries of subsequent NICU admissions
from February 6 to 17, 2012. Eligible infants were those stabilized or resuscitated
in the delivery room, defined as the place where the preponderance of stabilization
and resuscitation occurred, and subsequently admitted to the neonatal intensive care
unit. For each delivery, auditors reported the infant’s gestational age at birth and
delivery characteristics, how many members of the resuscitation team attended the
delivery, who attended, whether the composition of the team met unit policies or guidelines,
whether the team performed a briefing, a debriefing within 24 hours of delivery, and
whether a member or representative of the resuscitation team spoke to the mother or
other family member within 30 minutes of initial resuscitation. Units could answer
“family member or parent not available” if the mother was under general anesthesia
and it was known that no other family member authorized to receive medical information
about the infant was in the hospital within 30 minutes of the delivery. If the delivery
was attended by a single individual responsible for the initial resuscitation, a briefing
was defined as a systematic communication prior to the delivery between the individual
and one or more members of the obstetrical team to help prepare the individual for
the delivery, and a debriefing was defined as a systematic communication within 24
hours after the delivery between the individual and one or more members of the NICU
team responsible for monitoring the quality of delivery room practice. Delivery audits
occurred either by direct observation in the delivery room or by interview within
24 hours after admission of infants requiring delivery room resuscitation.

Vermont Oxford Network provided a manual of operations and data collection forms.
The audit employed a custom online data collection and reporting system built on the
Microsoft ASP.NET application framework and integrated into the Vermont Oxford Network
member website. Each unit received a confidential online report of results immediately
following the audit and a summary report of all participating centers’ results shortly
after audit completion.

For this analysis, measures relating to infant characteristics were reported within
the context of the audited deliveries. Measures relating to NICU characteristics came
from the last available annual Vermont Oxford Network member survey. Teaching hospitals
were defined as those that had pediatrics residents, neonatal fellows, or other residents
involved in care. Accordance with audit objectives was defined as the proportion of
audited deliveries for which all four of the following were done: team composition
met unit policy; briefing was performed; debriefing was performed; and family communication
occurred immediately following delivery. Units without guidelines or policies were
considered not to be in accordance with audit objectives. Deliveries for which a family
member or parent was not available was considered to be in accordance. Only descriptive
statistics are reported. All analyses were done in SAS 9.3 (Research Triangle Park,
N.C.). The VON Day audit was approved by the institutional review boards of the University
of Vermont and each participating hospital.