• quality improvement
• outcomes research
• neonatology
• performance improvement
• musiq

Neonatal intensive care units (NICUs) aim to provide safe, high-quality medical and surgical care for all newborns. The creation of, and widespread participation in, NICU quality collaboratives such as the Vermont Oxford Network (VON), Children’s Hospitals Neonatal Consortium (CHNC) and state perinatal collaboratives (such as California, New York, Ohio, Illinois, Massachusetts, Tennessee, Florida and North Carolina) has no doubt accelerated progress in the field of neonatal quality improvement and safety. However, there still remains limited evidence of overall efficacy of these collaboratives.1 NICUs strive to apply the best possible evidence in their local setting, known as ‘potentially better practices’, to achieve improved patient outcomes. However, similar interventions applied in a similar fashion in similar NICUs often do not produce the same results. This unexplained variability in outcomes between NICUs begs the question: What is the secret sauce? Why do some NICUs consistently outshine others in spite of the application of the same ‘potentially best practices’?

To answer these questions, it becomes necessary to determine those factors that contribute to success or failure of improvement programmes, including less tangible aspects such as NICU culture/environment, leadership, team work and organisational structure. Barron et al 2 identified eight critical success factors evident in places with strong performance improvement (PI) programmes:

1. Strong administrative executive and PI leadership. In creating this vision of excellence, a symbiotic relationship exists between executive leadership and the NICU. Leadership is supportive of PI efforts and is actively engaged with the NICU team; conversely, the NICU team strategically supports the vision of the organisation and exerts focus and effort in needed areas.

2. Active involvement of the Board of Trustees: the Board prioritises safety and quality and places organisational learning from data and parent satisfaction at its core, regardless of turnover in senior leadership.

3. Effective oversight structure: efforts are not duplicated within the system (eg, in the NICU, between NICUs or between the NICU and other departments), and a clear structure exists with capable leadership that sets goals, prioritises the order of important initiatives and is answerable for outcomes.

4. Expert PI staff trained in quality and safety who are capable of successful project execution, support the team and allow them to reach their full potential.

5. Physician involvement and accountability: physicians actively participate in planning and implementation of change with front-line clinical staff and are informal leaders—not merely bystanders. Physicians are involved in hospital-wide committees, and the job description of the unit director clearly featured quality and safety goals.

6. Active staff involvement: PI was viewed by front-line staff as a problem-solving tool, and grass roots initiatives proposed by them were rewarded and supported by the oversight team. Front-line staff are also encouraged to gain training in the use of PI tools.

7. Effective use of information resources (data used for decision making): the electronic health records and information technologists provided teams with adequate means of gathering data reports, and tools allowed tracking of performance measures or were modifiable enough to display new unit-specific measures.

8. Effective communication strategy to all stake holders: data were continually shared and disseminated at different stakeholder meetings or newsletters and benchmarked to other units or hospitals. This provided teams with feedback regarding their unit’s current performance metrics and where they needed to be.

Several models present an explanation for the connectivity and relationships between these factors, evidence-based medicine and patient and centre outcomes.3–5 The model for understanding success in quality improvement (MUSIQ)6 provides a conceptual model of contextual factors (the ‘where’, the ‘when’ and the ‘why’) and expands our understanding of the environmental, organisational (macrosystem), quality team and microsystem attributes that interplay with any quality initiative (‘the how’), to contribute to success or failure of a project (figure 1). Local elements such as team leadership, basic PI knowledge, motivation and unit culture were found to be direct influencers compared with more remote elements such as organisational and external factors.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The model for understanding success in quality improvement (MUSIQ).6

Kaempf et al 7 explain these centre variations through differences in contextual factors. The authors compare the characteristics of low and high performing NICUs caring for very low birthweight infants using a quantitative measure of performance, a risk-adjusted composite morbidity score called the Benefit Metric and the results of a 103 question contextual factors survey. All 39 NICUs participated in the VON and collectively showed a significant improvement in their yearly performance over a 15-year period, yet, 14 NICUs consistently outperformed the remainder. Contextual survey results from these 14 highest performing NICUs indicated that they all scored higher on questions related to the microsystem and quality environment. Physicians had good quality improvement knowledge and skills, a supportive culture that promoted teamwork and willingness to accept change, good staffing models, time resources that allowed for participation in PI activities and positive motivation for change were all features of high performers. External factors such as participation in collaborative PI seminars, participation in routine quality activities and communication of outcomes to front-line staff did not differ between high and low performing units. Similar to other models, this study shows that the microenvironment is one of the biggest influences on success of Quality Improvement (QI) projects. However, much of this work was through self-reported surveys that assessed 15 years of contextual factors which may have undergone changes in leadership, practice and environment.

Reflecting on our NICU, we continue to have the lowest rates of objective measures such as central line-associated blood stream infections (CLABSI) (0.33/1000 line days) and unintended extubations (0.35/100 ventilator days) in the USA, and our outcome measures, for example, bronchopulmonary dysplasia (BPD) fall within the 50th percentile compared with other centers in the CHNC collaborative. Yet we are a busy level IV NICU in a free-standing children’s hospital with a rapidly growing capacity, higher levels of complex patients, presence of trainees on rounds and routine 3:1 and 2:1 staffing models. We attribute our success to direct involvement of all levels of leadership in our unit in PI initiatives, a dedicated local PI team, quality trained medical unit director, engagement of front-line staff in PI, the presence of local subject matter experts, multidisciplinary diverse team both within the NICU and with other departments that bring an array of experiences and opinions and a supportive data infrastructure through local information technology and use of the Children’s Hospital Neonatal Database that allows benchmarking to other non-delivery NICUs. Our team finds motivation in solving local issues routine in our work, and leadership prioritises these issues and promotes engagement of front-line staff. Our outcomes undergo consistent review by senior leadership and are reported to the Board of Trustees. In response to questions raised by Kaempf,7 the secret sauce may lie in establishing systems that promote the culture of quality and safety rather than waiting for a reduction in morbidity to promote happier staff. Contextual factors play a critical role in enhancing performance and should be studied and reported in quality publications to allow readers to determine the applicability and generalisability of interventions to their local context. Metrics such as MUSIQ provide us with tools that can be used to report on these contextual factors and can be used to determine the likelihood of success at the beginning of a project.