The push for what now is being termed "evidence-based practice" began in the clinical areas of nursing and has moved into how care is delivered. The groundbreaking research by Knaus, Draper, Wagner and Zimmerman in 1986 that showed a correlation between collaboration among nurses and physicians in the ICU and the lower incidence of patient morbidity and mortality started the movement to identify organizational issues that make a difference in patient care outcomes.

Consider: In 1992, Baggs et al. studied the association between interdisciplinary collaboration and patient outcomes in an MICU. After controlling for severity, the risk of negative outcomes (readmission or death) was reduced from 16 percent when the nurse reported no collaboration in the decision-making about patient transfers out of the MICU, to 5 percent when there was full collaboration.

In 1994, Aiken, Smith and Lake compared Medicare mortality between 39 Magnet hospitals and 195 control hospitals. When the data were adjusted for differences in predicted mortality, the mortality rate in the Magnet hospitals was 4.6 percent lower than in the control hospitals. Also, in 1994, Shortell et al. studied 17,440 patients from 42 ICUs and found that caregiver interaction was significantly associated with lower risk-adjusted length of stay, higher technical quality of care and greater ability to meet family members' needs.

In 1999, Aiken et al. studied organization and outcomes in inpatient AIDS care by looking at hospitals with dedicated AIDS units, hospitals with scattered-bed AIDS care and Magnet hospitals. The 30-day mortality was 60 percent lower in Magnet hospitals and 39 percent lower in dedicated AIDS units. There also was a strong association between organizational control of care by bedside nurses and patient satisfaction.

Two studies by the American Nurses Association in 1997 and 2000 show that increased RN hours of care and increased RN staffing mix both result in lower infection rates. The Department of Health and Human Services' study on nurse staffing and patient outcomes (released in February 2001 and highlighted in the May 30 issue of the New England Journal of Medicine) showed strong and consistent relationships between both level and mix of RN staffing and negative patient outcomes such as UTIs, upper GI bleeding, pneumonia, length of stay, shock and failure to rescue.

Most recently, the results of a study by Clarke et al. reported in the June issue of the American Journal of Infection Control show that the likelihood of needlestick injuries and near-misses to nurses are increased 50 percent to 200 percent in hospitals with poor organizational climates and high workloads.

The evidence continues to mount. Studies consistently show that three key elements are associated with improved patient outcomes: staffing adequacy, support of nursing care by hospital administration, and collaboration between nurses and physicians.

Nurses have been building this evidence base for the last 10 years, but only recently has increased attention been paid to the results. One reason is the spotlight on nurses as we become a scarce resource. The other is the growing awareness that negative patient outcomes, in addition to being bad for patients, also are bad for health care facilities. Money and other resources needed elsewhere must be diverted to fixing the problems and often fighting or settling lawsuits; reputations become tarnished when patient outcomes do not meet the expectations of patients, insurers and providers.

Health care executives, patients and their significant others, insurers and many other groups need to understand the value of registered nurses. Our task is to make sure that this evidence gets into their hands and is translated into terms that each group understands.

Discuss this and other topics with your colleagues at www.nurseweek.com/rnvillage.