Can collaborative learning and performance benchmarking reduce infections acquired in neonatal intensive care units?

The coagulase-negative staphylococcus (CONS) bacterium is the most frequent cause of infections acquired by premature, very low birthweight infants in the neonatal intensive care unit (NICU). A collaborative quality improvement project among six NICUs resulted in a 44 percent lower incidence of CONS infection among such infants. This change was significantly different from the trend in infection rates among 66 other NICUs participating in a surveillance system.

Why is this important?

Very low birthweight infants being cared for in neonatal intensive care units (NICUs) are especially vulnerable to hospital-acquired (nosocomial) infections because of their immature immune systems and prolonged hospital stays (Harris 1997).

Sepsis affects up to one of five very low birthweight infants in the NICU and is associated with substantially higher death rates and longer hospital stays (Stoll et al. 2002). The coagulase-negative staphylococcus (CONS) bacterium is the most frequent cause of such infections.

Interventions

The Vermont Oxford Network (VON) convened multidisciplinary teams from six NICUs that volunteered to participate in a three-year Neonatal Intensive Care Collaborative Quality (NIC/Q) Project. Activities included training on quality improvement, agreeing on common improvement goals and metrics, reviewing performance data, and developing a list of "potentially better practices" for improvement (Horbar et al. 2001).

• Better practices were identified by sharing detailed analyses of care processes conducted by each site, reviewing the medical literature for evidence on prevention practices, and benchmarking the practices of participating sites and other superior-performing NICUs through a series of site visits.
• Each NICU selected specific practices for improvement that its team considered most relevant, in the following areas: handwashing, nutrition, skin care, diagnosis, respiratory care, vascular access (intravenous line management), and organizational culture.

A comparison group included 66 North American NICUs that contributed 25 or more cases to the VON outcomes database from 1994 to 1996.

Findings

Among six NICUs participating in the collaboration, the average rate of hospital-acquired CONS infections in very low birthweight infants decreased by 5.4 percentage points in absolute terms from 1994 (before the intervention) to 1996 (after the intervention). This change was significantly greater than a 0.9 percentage point decrease in the average infection rate among 66 comparison NICUs during this time (Horbar et al. 2001).

One year later (1997), the infection rate at the six collaborative sites continued to decline, to a level 9.7 percentage points lower than before the intervention, representing a 44 percent relative decrease since 1994. This change was significantly different from the trend at comparison NICUs. Infection rates improved at four of the six intervention sites, while worsening at the other two sites.

During the project, treatment costs were reduced by $10,932 per infant among the six collaborating NICUs, representing average savings of $2.3 million in annual patient care costs per NICU. Treatment costs rose at comparison NICUs. With an average resource commitment of $68,206 per NICU, plus grant-funded support, the quality improvement produced $9 in savings for every $1 invested (Rogowski et al. 2001).

Implications

An intensive, multidisciplinary, multicenter collaborative learning process conducted among highly motivated participants can support changes in local institutional practices that lead to improvement in average clinical outcomes and a reduction in patient care costs.

Improvement Ideas and Resources

Additional resources are available from the Vermont Oxford Network.

Measure:

Nosocomial infection was defined as the occurrence, after the third day of life, of an infection from a predefined list of bacterial pathogens. "Coagulase-negative staphylococcal infection (CONS) required the recovery of the organism from blood or spinal fluid, signs of systemic illness, and treatment for 5 or more days antibiotics." The study included infants weighing 501 to 1,500 grams who were born at or transferred to the NICUs within 28 days of birth, and who were hospitalized more than three days at intervention sites (N=745 born in 1994, 772 born in 1996, and 789 born in 1997) and at comparison sites (5,108 born in 1994, 5,528 born in 1996, and 5,572 born in 1997). Significant differences in magnitude of change in average CONS infection rates at intervention and comparison sites (group-by-year interaction) were found using a logistic regression model that controlled for birth weight, location of birth, multiple birth, assisted ventilation, and year of birth. Secondary analysis (based on year of birth) found a significant change in outcome over time among intervention NICUs from 1994 to 1997. There was a significant decline in the average rate of all measured nosocomial infections at intervention sites during this time, attributable to the decline in CONS infections (Horbar et al. 2001).

Limitations:

Participating hospitals were not randomized to treatment and control groups.

Source:

Outcomes data were derived from the Vermont Oxford Network Database, which tracks standardized information on very low birthweight infants for purposes of quality improvement, internal audit, and peer review at participating NICUs. Results were reported by researchers at the University of Vermont College of Medicine, Burlington, Vt.; RAND Corporation, Washington, D.C.; Paul E. Plsek and Associates, Inc., Roswell, Ga., and the participating hospitals (Horbar et al. 2001).

References:

* Indicates source of data used in the chart(s).Harris, J. A. 1997. Pediatric Nosocomial Infections: Children Are Not Little Adults. Infection Control and Hospital Epidemiology 18 (11): 739–42.

* Horbar, J. D., J. Rogowski, P. E. Plsek et al. 2001. Collaborative Quality Improvement for Neonatal Intensive Care. NIC/Q Project Investigators of the Vermont Oxford Network. Pediatrics 107 (1): 14–22.

Rogowski, J. A., J. D. Horbar, P. E. Plsek et al. 2001. Economic Implications of Neonatal Intensive Care Unit Collaborative Quality Improvement. Pediatrics 107 (1): 23–9.

Stoll, B. J., N. Hansen, A. A. Fanaroff et al. 2002. Late-Onset Sepsis in Very Low Birth Weight Neonates: The Experience of the NICHD Neonatal Research Network. Pediatrics 110 (2 Pt 1): 285–91.