News Release

Professional Nursing Care Related To Fewer Adverse Patient Occurrences, Study Finds

Peer-Reviewed Publication

University of Iowa

IOWA CITY, Iowa -- In a study that could have implications for how hospitals respond to competitive pressures and managed care, a University of Iowa researcher has found that inpatient care units with higher proportions of care delivered by registered nurses have lower rates of medication errors, patient falls and other adverse patient outcomes.

Mary Blegen, Ph.D, professor and associate dean of the UI College of Nursing, led two separate studies showing a relationship between the level of staffing by registered nurses and adverse occurrences affecting hospital patients. The most recent study used data from 39 inpatient nursing units in 11 hospitals in Iowa and will be published in an upcoming issue of the journal Nursing Economics. It confirms the findings of an earlier study and concludes that units with a higher proportion of care delivered by RNs had lower rates of medication errors and patient falls.

The finding is important because managed care and increased competition have prompted many hospitals to substitute unlicensed care givers for registered nurses, decreasing the overall proportion of patient care delivered by RNs in these settings. At the same time, Blegen said, shorter hospital stays mean hospitalized patients generally are more acutely ill and in greater need of skilled care.

Despite these trends, Blegen said, the question of what level and mix of staffing are appropriate for different types of inpatient units remains an open one. "Part of the reason we did these studies is that there's very little research to guide nursing administrators in determining how to staff their units," Blegen said.

"Currently, hospital restructuring responding to managed care is driving staff cuts at many hospitals," she added. "But the level of staff needed to deliver the best patient care should also be a factor, and that's what this research is about."

Both studies used data commonly reported by hospitals for quality control and accreditation reviews. Investigators gathered information on adverse patient occurrences in each unit as well as the number of patient days of care delivered. To determine the staff mix of each unit, they looked at the total hours of care by staff and what proportion of those hours was delivered by registered nurses.

The two studies found a direct relationship between the proportion of care by RNs and the number of certain adverse occurrences per thousand patient days or, in the case of some units, the number of medication errors per 10,000 doses. As the proportion of professional nurses increased, the adverse occurrences decreased.

One surprising exception to this finding was that medication errors decreased as RN care increased to 87.5 percent of total care hours, but increased when the RN mix rose above that level. Blegen and Thomas Vaughn, an assistant professor of hospital and health administration at the UI who co-authored the second study with her, suggested possible explanations for this phenomenon but wrote that additional research is needed to find out why an RN mix above 87.5 percent would be associated with increased medication errors.

Both studies examined the relationship between nursing staff mix and the rates of medication errors and patient falls. The first study, however, also looked at other indicators of nursing care quality, including bed sores, patient complaints and hospital-acquired infections.

Published in January in the journal Nursing Research, this study used data from 42 inpatient nursing units at the University of Iowa Hospitals and Clinics, an 880-bed teaching hospital. It found a clear statistical correlation between nursing staff mix and rates of medication errors, patient complaints and bed sores (technically known as decubitus ulcers, or decubiti). Less clear was the relationship between nursing skill mix and patient falls, hospital-acquired infections and deaths.

The second study, using data from 11 Iowa hospitals ranging in size from fewer than 100 beds to more than 300 beds, also looked at the rates of cardiac arrest in the 39 patient units but found no clear correlation with nursing staff mix.

The studies took data from skilled care, pediatric, psychiatric, obstetric, medical, surgical and intensive care units. Because the intensity of patient care in the different units calls for varying levels of skilled nursing, Blegen and her colleagues used a statistical method known as regression analysis to correct for this bias and give each unit equal weight for the studies' purposes. Blegen's research differs from previous studies in its use of data from specific units and of nurse staffing devoted only to patient care. Other researchers have gathered information on patient outcomes at the hospital level and included nurses devoted to administration and other non-patient care activities. This difference, Blegen noted, made the results of her first study "more detailed and specific but perhaps less generalizeable" than other studies.

Now, with the more recent investigation confirming the results of that first study, "it's becoming more clear that cost containment efforts that attempt to reduce the proportion of professional nurses on inpatient units may be ill advised," she said.

Co-authors of the first study are Colleen Goode, Ph.D., associate vice president for patient services at the University of Colorado Hospital in Denver and formerly director of nursing services at the University of Iowa Hospitals and Clinics, and Laura Reed, a departmental personnel specialist at UIHC. It was supported in part by a grant from the Midwest Alliance In Nursing/Midwest Nursing Research Society. The second study, co-authored by Thomas Vaughn, received support from the University of Iowa and from participating hospitals through the Institute for Quality Healthcare.

Blegen is associate dean for academic affairs of the UI College of Nursing. She can be reached by phone at 319-335-8210, and by e-mail at mary-blegen@uiowa.edu.

###


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.