|
|
 |
|
|
Our
relatively young specialty of neonatology has been characterized by remarkable
discovery and innovation. Collaboration among neonatal units and health
professionals has been essential to this rapid progress, evidenced by
willing participation, both funded and voluntary, in randomized controlled
trials of new therapies. The creation of neonatal networks has facilitated
such collaboration. Seventeen years ago, Lucey described the organized
study of surfactant replacement therapy –from development to clinical
trials to implementation — as an example of “getting it right.”1 Despite
the significant advances made, few would argue that the quality of neonatal
care is the best it can be. For example, after improvements related to
widespread adoption of surfactant treatment in the 1990’s had been realized,
outcomes for vulnerable, extremely premature, or very low birth weight
(VLBW) neonates have continued to improve only slightly, if at all. During
this time another approach to achieving better outcomes began to be applied
in medicine. QI techniques, largely developed in industry, employ the
concepts of standardization, measurement of processes as well as outcomes,
and habits of structured introduction of changes. An appealing aspect
from industry, QI is the idea of benchmarking—finding out who is
getting the best results and learning from them. These concepts have found
fertile ground in neonatology. The papers reviewed herein represent several
different models of how networks and collaborative projects have tested
applications of QI in neonatology. The mixed results represent both a
sense of optimism and opportunities for continued learning.
The Northern
New England Cardiovascular Disease Study Group (NNECDSG), a multidisciplinary
regional voluntary consortium of cardiovascular programs in Maine, New
Hampshire, and Vermont, is one of the first examples of a collaborative
for QI in medicine. Using feedback of outcome data, training in QI techniques,
and site visits to other medical centers, the consortium reported a 24%
reduction in hospital mortality rates from coronary artery bypass graft
(CABG) surgery.2 The
NNECDSG has also contributed over 75 articles to the peer-reviewed literature,
leading to greater understanding of individual patient risk for mortality
from CABG, as well as team and process variables linked to outcomes. The
first Vermont Oxford Network QI collaborative project was modeled on the
NNECDSG. In the reports of this project, the Horbar and Rogowski studies
showed a modest reduction in the incidence of coagulase-negative staphylococcal
sepsis (BW 501-1500 gm), a reduction in rates of oxygen supplementation
or death at 36 weeks postmenstrual age (BW 501-1000 gm), and an intriguing
reduction in the costs of care when collaborative centers were compared
to contemporaneous controls.
The cluster-randomized trial of benchmarking
in participating National Institute of Child Health and Human Development
(NICHD) Neonatal Research Network centers (Walsh et al) failed to show
significant improvement in rates of survival free of bronchopulmonary dysplasia
(BPD) in infants of birth weights <1250 grams, in spite of evidence
that intervention centers were able to change practices. The authors raise
an important caution that the interventions chosen were supported by very
weak existing evidence. The only three potentially better practices intended
to be implemented in all study centers (higher PaCO2 target;
lower oxygen saturation goals; and high-saturation alarms set at 95%) were
classified as having indeterminate or no supporting evidence. Reduced oxygen
saturation targeting has shown promise in reducing the incidence of retinopathy
of prematurity (ROP) and BPD, but large scale multicenter randomized controlled
trials which include longer-term neurodevelopmental outcomes are lacking.
On the other hand, the cluster-randomized
trial by Horbar et al promoted implementation of a practice with strong
evidence (surfactant use) but inconsistent clinical usage. That study showed
success in changing practice, but failed to demonstrate the improved outcomes
predicted by the existing evidence. The approach used in the study is similar
to the Breakthrough Series Collaborative, popularized by the Boston-based
Institute for Healthcare Improvement (IHI). The IHI conducts Learning Sessions
for hospital or clinic teams on a focused topic area. Evidence-based practices
are taught along with quality improvement methods. The IHI reports improved
outcomes in many of the topics of the series, such as dramatically reducing
hospitalizations for adult patients with congestive heart failure by 50%.3
Finally, the study by Bloom et al represents
an innovative Best Demonstrated Process methodology used successfully
by the Pediatrix® Medical
Group. Following a process of isolating and sharing meaningful differences
in care between centers in the top and bottom thirds for weight gain, they
demonstrated an overall increase in average daily weight gain during the
first 28 days after birth (BW 401-1500 gm).
Neonatal QI collaboratives are becoming quite
popular. The California Perinatal Quality Care Collaborative (CPQCC), founded
in 1997, has virtually all NICUs in California participating. New collaboratives
are evolving in states and regions of the US and internationally. It is
clear that we need to understand more about the interaction between processes
of care or evidence-based practices and the contextual environment of individual
centers. Rather than becoming discouraged by the mixed results reported
in the articles summarized herein, we should realize that there are great
opportunities, through collaboration, to develop new knowledge about how
prevalent mental models affect adoption of new practices; how complex organizations
work; how to analyze and change processes in a sustainable manner; how
to rigorously measure processes and outcomes; and how to study and communicate
results.
The science of healthcare quality improvement
is in its infancy and must be nurtured. We are more likely to achieve and
sustain better neonatal outcomes by taking advantage of the synergy between
the science of clinical trials and the science of QI than by relying on
either approach alone. Multi-institution collaboratives are ideal settings
to create such synergy.
References
|
|
|
|
|
|
