Abstract
When nurses say that low overall staffing numbers and low proportions of RNs adversely influence patient outcomes, the basis for that claim is stronger than individual impression and opinion. There is research evidence supporting it. A recent study commissioned by the American Nurses Association (ANA) confirmed the relationship between low staffing and poor patient outcomes. These results add to the evidence of prior studies in which a strong inverse relationship between nurse staffing and adverse patient events has been found. In this article, a synopsis of the ANA study is presented and a suggestion regarding how to build on the study's methods and findings is made.
In many settings nurses are troubled by issues of staffing. A major part of that concern stems from the belief that staffing mix and numbers affect the occurrence of errors, complications, falls, and other adverse events. A second part of the concern is that staffing deficiencies affect patients' level of comfort, rate of recovery, and transition to home because they don't get the nursing care they should. Third, staffing deficiencies affect the quality of nurses' work lives because they increase stress and generate angst about making mistakes.
The concerns about the effects of deficient staffing on patient outcomes and complications is reflected in the results of NAON's 1997 Delphi study. That study was conducted to determine the specific issues the membership thinks need to be examined by research. Four of the top eight questions identified in that study pertain to the ramifications of staffing (NAON, 2000).
Well-designed studies looking at the relationship between staffing patterns and patient outcomes are difficult and costly to conduct. For that reason, we look to federally funded agencies, endowed foundations, large professional organizations, and federally funded research teams to provide us with knowledge about the effects of various staffing patterns. To support the conduct of studies on this issue, NAON gives them priority consideration when awarding NAON and AAOS grants.
Some nurses have said in essence, "Why do we need research about this? It's a no-brainer. Low proportions of RN staffing and low overall numbers clearly cause errors, oversights, and gaps in care."
At one level this is true. However at another level, we need to understand the whole issue better. We need to be able to quantify the risks and the benefits of various staffing models. We need to know what staff mixes work best with different patient populations and acuity levels. We have to learn more about how staffing and system characteristics (such as medication ordering, dispensing, and administration procedures) interact with nurse staffing to influence patient care and contribute to, or avoid, adverse events.
Perhaps most important, we need to document how nurses: (a) contribute to good patient outcomes, (b) prevent complications, and (c) keep patients from experiencing adverse events.
A recent study commissioned by the American Nurses Association (ANA) documents the association between nursing staffing and patient outcomes. The study provides objective, large-scale evidence regarding the linkage between staffing and patient outcomes. The results enable nurses to go beyond personal opinions and anecdotes when claiming that deficient staffing patterns can adversely affect patients' well-being (ANA, 2000).
Purpose
This study sought "to quantify the relationships between nurse staffing and patient outcomes" by collecting and analyzing data from a large cross-section of U.S. hospitals (ANA, 2000, p. 1).
Methods
Data was collected from very large data sets in nine states, and included a total sample of nearly 13 million patients in approximately 1500 hospitals (p. vii). A special analysis of the Medicare elderly was done because they are expected to have higher rates of complications (ANA, 2000, p. 19).
Patient variables examined were length of stay and 16 patient outcomes including: pneumonia, postoperative infections, pressure ulcers, urinary tract infections, joint effusion, transfusion reactions, vascular complications, and adverse drug reactions. The expected adverse outcomes rates were based on the average adverse outcome rate for a diagnosis-related group (DRG) and each hospital's mix of patients by DRGs. Each hospital's adverse outcome rate then fell at, above, or below the expected rate. The nurse staffing variables studied included the amount of nursing time provided to patients and the skill mix of the nursing staff (ANA, 2000, p. 16). Setting variables such as type of hospital and rural versus urban setting were also analyzed.
Results
Collectively, the results "clearly show that relationships exist between the amount and mix of nursing staff and patient outcomes" (ANA, 2000, p. 25). Five of the outcomes (length of stay, pneumonia, postoperative infections, pressure ulcers, and urinary tract infections) showed statistically significant relationships with nurse staffing.
In particular, lower than expected rates of four complications were associated with a higher mix of registered nurses across all three data sets (one all-payer data set and two Medicare-only data sets). Shorter lengths of stay and lower rates of pressure ulcers rates were related to higher level of overall staffing after adjustment for acuity; lower rates of pressure ulcers were also related to RN mix.
Lower than expected rates of postoperative infection rates were associated with more total licensed personnel hours per day in the all-payer data set but not in the two Medicare data sets (ANA, 2000, pp. 19-20).
Paradoxically, higher rates of pneumonia were associated with total licensed hours per day. Many of these findings are consistent with those of other studies (Aiken et al., 1999; Blegen et al., 1998; Knauf et al., 1997; Kovner & Gergen, 1998; Scott et al., 1999).
Use of the Findings
The demonstration of relationships between nurse staffing and patient outcomes across several studies is strong evidence in support of the claim that what appears to be cost savings from reductions in nursing staff numbers and mix may not be real cost savings when the additional costs related to treating complications and increased length of stay are considered.
The outcomes examined in the study were chosen because they are sensitive to nursing actions and interventions. If a health care facility or system were to collect data about the patient outcomes included in this study and analyze them in relation to nurse staffing, they would gain insight into how their staffing patterns affect patients' well-being and the cost of caring for patients.
Beyond the better financial decision making made possible by such knowledge, the members of the public who are considering whether to enter a particular health care system or facility have a right to information about nurse staffing and nursing-sensitive quality indicators.
Linking Staffing and Patient Outcomes in Your Hospital
* Cite the strong evidence showing that good nurse staffing can prevent adverse patient outcomes.
* Collect and analyze same data for own unit or department.
* Include nursing-sensitive outcomes in hospital's quality improvement data collection.
* Include nursing-sensitive outcomes in hospital's clinical information system.
* Promote the use of information about good nurse staffing and good patient outcomes in hospital's public relations materials.
References