This publication is provided for historical reference only and the information may be out of date.
A shortage of registered nurses, in combination with increased workload, has the potential to threaten quality of care.1–3 Increasing the nurse to patient ratios has been recommended as a means to improve patient safety.4, 5 However, the cost effectiveness of increasing registered nurse (RN) staffing is controversial.6, 7
This systematic review analyzes associations between hospital nurse staffing and patient outcomes with consideration of variables that could influence the primary association. The basic research questions were:
How is a specific nurse to patient ratio associated with patient outcomes (i.e., mortality; adverse drug events, nurse quality outcomes, length of stay; patient satisfaction with nurse care)? How does this association vary by patient characteristics, nurse characteristics, organizational characteristics, and nursing outcomes?
How is a measure of nurse work hours (hours per patient or patient day) associated with the same patient outcomes?
What factors influence nurse staffing policies?
What nurse staffing strategies are effective for improving the patient outcomes listed in question 1?
What gaps in research on nurse staffing and patient outcomes can be identified to address in future studies?
Questions 1, 2, and 4 are addressed in the systematic review using meta-analytic approaches. The literature associated with question 3 does not lend itself to meta-analysis.
Questions 1 and 2 address the same basic association but employ two different measures of nurse staffing. The nurse to patient ratio relies on a general ratio, which may include all nurses assigned to a unit, including non-clinical time, whereas nurse work hours look specifically at nurses involved in patient care. Even beyond this distinction, the varied ways staffing rates are calculated complicates pooling data.
Observational studies from from 1990 to 2006 from the United States and Canada were reviewed for questions 1, 2, and 4. Studies for question 3 addressed implications for nurse staffing policies. No studies primarily empirically examined a specific nurse staffing policy. Sources included journal articles, administrative reports, and dissertations.
For questions 1, 2, and 4, we present the relative risks of nurse staffing levels on various patient outcomes adjusted for measured confounding factors. Meta-analysis was used to test the consistency of the association between nurse staffing and both patient outcomes and economic outcomes (e.g., length of stay); the analyses were conducted separately for classes of patients and hospital characteristics.
Of the 94 eligible studies from 96 reports, 7 percent were case-control studies; 3 percent were case-series; 44 percent were cross-sectional studies; 46 percent assessed temporality in the association between nurse staffing and patient outcomes. The overall quality of the studies averaged 38 (of a possible 50).
Patient Outcomes and Nurse Staffing Ratios
Consistent evidence from observational studies suggests that an increase in Registered Nurse (RN) to patient ratios was associated with a reduction in hospital-related mortality, failure to rescue,1 and other nurse sensitive outcomes, as well as reduced length of stay (LOS), after adjustment for patient and provider characteristics but does not establish a causal relationship. The effect size is greater in surgical patients; ratios less than 2.5 patients per RN per shift in intensive care units (ICUs) and less than 3.5 patients per RN in surgical units were associated with the largest risk reduction based on quartiles of nurse staffing ratios.
Pooled results showed that every additional RN full time equivalent (FTE) per patient day was associated with a relative risk reduction in hospital-related mortality by 9 percent in intensive care units and 16 percent in surgical patients.8–21 If the relationship were indeed causal, we estimate that an increase by one RN FTE per patient day would save five lives per 1,000 medical patients, and six per 1,000 surgical patients. Reducing the workload from more than six to two or less patients per RN per shift would save 25 lives per 1,000 hospitalized patients and 15 lives per 1,000 surgical patients. A further reduction from two to four patients to less than 1.5 patients per RN would save four lives per 1,000 hospitalized patients and nine lives per 1,000 surgical patients. However, staffing rates of this magnitude may not be realistic.
Every additional patient per RN per shift was associated with a 7 percent increase in relative risk of hospital acquired pneumonia,13, 14, 22 a 53 percent increase in pulmonary failure,13, 14, 23, 24 a 45 percent increase in unplanned extubation,13, 14, 23–25 and a 17 percent increase in medical complications.13, 23, 24 The increase in relative risk of unplanned extubation and pulmonary failure was higher and in hospital acquired pneumonia was lower, corresponding to an increase in patients per nurse ratios. We estimated that if the relationship were causal, one additional patient per RN per shift would result in 12 additional cases of failure to rescue, six cases of pulmonary failure, and five accidental extubations per 1,000 hospitalized patients.
The associations vary by clinical settings and patient population. In ICUs, an increase by one RN FTE per patient day was associated with a consistent decrease across studies in relative risk of these patient outcomes: a 28 percent decrease of cardiopulmonary resuscitation,13, 23, 24 a 51 percent decrease of unplanned extubation,13, 14, 23–25 a 60 percent decrease of pulmonary failure,13, 14, 23, 24 and a 30 percent decrease of hospital acquired pneumonia.13, 14, 22 In surgical patients, an increase of one RN FTE per patient day was associated with a consistent reduction in the relative risk of failure to rescue by 16 percent,12, 15, 16, 20, 21 and in nosocomial bloodstream infections of 31 percent.
The data on other nursing personnel is limited and not replicable in the studies. LOS was shorter by 24 percent in ICUs and by 31 percent in surgical patients, corresponding to an additional RN FTE per patient day.8, 9, 13, 14
Patient Outcomes and Nurse Staffing Hours
An increase in total nurse hours per patient day was associated with reduced hospital mortality, failure to rescue, and other adverse events. The death rate decreased by 1.98 percent for every additional total nurse hours per patient day (95 percent confidence interval [CI] 0.96–3 percent).26–29 The association with RN hours per patient day did not show significant changes in mortality rates.26–29 The relative risk of death was lower by 1 percent per 1 additional RN hour per patient day in ICUs8, 9, 13, 14, 16 and in medical8, 10, 11, 17–19, 26, 27, 30–32 and surgical patients.9, 12–16, 20, 26, 27 The association between LPN/LVN hours per patient day and death rate was not consistent across studies.17, 20, 26, 27, 33, 34
The association between patient outcomes and RN and LPN/LVN hours was inconsistent across the studies. Pooled analysis showed that 1 additional RN hour per patient day was associated with a reduction in relative risk of hospital acquired pneumonia by four percent,13, 14, 22 pulmonary failure by 11 percent,13, 14, 23, 24 unplanned extubation by 9 percent in ICUs,13, 14, 23–25 failure to rescue by 1 percent in surgical12, 15, 16, 20, 26, 27, 30 and medical patients,26, 27, 35 and deep venous thrombosis by 2 percent in medical patients.27, 35
The LOS in hospitals was lower for additional total nursing, but not for licensed LPN/LVN and unlicensed assistive personnel (UAP) hours. The association between RN hours and LOS was not consistent across studies.
Other Attributes of Nursing
There was a significant negative correlation between the percentage of nurses with Bachelor of Science in Nursing (BSN) degrees and the incidence of deaths related to health care (r = -0.46, p = 0.02). Nurse job satisfaction and autonomy was associated with a significant reduction in the risk of death. An increase in nurse turnover increased the rate of patient falls by 0.2 percent.36
Staffing policies examined for this review related to the shift length, scheduling nurses to rotate to different shifts, mandatory overtime, weekend staffing, use of agency or temporary nurses, assigning nurses to nursing units other than those they are regularly assigned to work (floating), use of full-time, part-time, and internationally educated nurses (IENs), the nurse-to-patient ratio or nursing hours per patient day for nursing units, and the skill mix (licensed vs. unlicensed staff) of nursing units. Overall, few studies for any of these staffing policy variables limited drawing any conclusions. Trends in the literature suggested that rotating shifts may have negative effects on nurses' stress levels and job performance perceptions. Further, several studies indicated that nurses working longer hours may have a negative impact on patient outcomes and safety. No research provides guidance on the impact or effective use of agency/temporary staff. Research on the use and effectiveness of IENs in U.S. hospitals37 includes qualitative exploratory studies38, 39 and descriptive studies40–42 that examined IEN use in healthcare. No studies empirically evaluated the interaction of IEN staffing policies with organizational, nurse, or patient care unit factors.
Within the limits of scant literature, RN overtime is not associated with the location of the hospital, teaching status of the hospital, average hours in a nurses' work week, acute bed occupancy, acute average daily census, or financial margin of the hospital.37, 42–44 More overtime hours were associated with an increase in hospital-related mortality, nosocomial infections, shock, and bloodstream infections. The proportion of float nurses was positively associated with the risk of nosocomial bloodstream infections.45–47 More contract hours was associated with an increase in LOS.28, 45, 48, 50
This review confirms previous contentions that increased nurse staffing in hospitals is associated with better care outcomes,51 but this association has not been shown to reflect a causal relationship. Hospitals that invest in more nurses may also invest in other actions that improve quality. Magnet hospitals that are said to provide high quality care have better nurse staffing strategies.10, 52 Overall hospital commitment to a high quality of care in combination with effective nurse retention strategies leads to better patient outcomes, patient satisfaction with overall and nursing care, and nurse satisfaction with job and provided care.10, 52–59
Two general measures of nurse staffing were studied.60 One addressed hours of care provided by nursing staff averaging FTEs of different nurse categories at the hospital level,11, 18, 19 sometimes including only productive hours worked in direct care.28, 61, 62 The other relies on less precise data of total nurse staffing to patient volume derived from administrative databases61, 63–65 averaging annual nurse to patient ratios20 at the hospital or unit level.20 The ratio of patients per RN per shift ratio was more frequently used and provided greater evidence of the effect, but both showed generally the same trends.
The effect size varied with the nurse staffing measure. The reduction in relative risk of hospital related mortality was 16 percent for one RN FTE per patient day, and 1 percent for an additional RN hour per patient day in surgical patients. Assuming that every additional RN FTE per patient day would provide approximately 8 additional RN hours per patient day, the expected reduction should be more than observed in the studies that examined the risk of mortality in relation to nurse hours. The comparison of the effect size on patient outcomes among quartiles of patients per RN per shift ratio and nurse hours per patient day detected the same pattern; the maximum reduction in relative risk of hospital-related mortality and adverse events occurred when no more than two patients were assigned to an RN and more than 11 nurse hours were spent per 1 patient day. We did not find consistent evidence that a further increase in RN FTE per patient day ratio can provide better patient safety. The evidence of the effects of LPN/LVNs and UAP were limited and inconsistent.
It is difficult to transition between nurse hours and nurse-to-patient ratios. Nurse hours per patient day reflect average staffing across a 24-hour period and do not reflect fluctuations in patient census, scheduling patterns during different shifts (even the length of shifts varies),9, 13 and periods of the year.66, 67 They do not account for the time nurses spend in meetings, educational activities, and administrative work.
Nurse staffing could have a different effect in different hospital settings. The addition of one unit of nursing care may depend on the baseline rate. The effect of an additional nurse hour might be quite dissimilar in ICUs and typical hospital units. As shown in previous studies,26, 27 the present meta-analysis found consistent evidence that surgical patients are sensitive to nurse staffing.
The size of the nursing effect must be tempered by all the other factors not considered in most of these studies. No direct measure of other influences on outcomes is typically made. The traditional concerns about factors that affect quality of care, such as the nature of the primary medical and surgical treatment and the skill of the physician staff, are not addressed and are assumed to be evenly distributed to yield noise, but not bias. Many of the studies are performed on data collected at the hospital level over a long period of time. Adjustments for comorbidity depend on simple averages.
Skill, organization, and leadership undoubtedly play a role but are much more difficult to assess. Skill mix did not demonstrate consistent associations with tested patient outcomes in the present review. Nurse competence requirements include education, expertise, and experience68, 69 Nurse education was associated with lower mortality. The importance of nurses' professional competence and performance have been discussed with regard to developing standards of nurse performance to encourage high quality of care.70–73
Increased nurse staffing in hospitals is associated with better care outcomes, but this association is not necessarily causal. The effect size varied with the nurse staffing measure and sites of patient care (i.e., ICU, medical vs. surgical units). The size of the nursing effect must be tempered by all the other factors not considered in most of these studies.
Future observational studies will need to take cognizance of the many other factors that can affect the outcomes of interest, especially medical care, patient characteristics, and organization of nursing units and staffs. Larger multi-center studies will be needed. More studies should be conducted at the patient level to allow for better control of issues like comorbidity. Hierarchical models that control for both institutional and nursing effects could be employed. Nonetheless, it is unlikely that all the salient variables can be addressed in any one study. Future work will need to target specific questions and collect and analyze enough information to isolate the effects of nurse staffing levels.
The number of deaths in patients who developed an adverse occurrence among the number of patients who developed an adverse occurrence.
Agency for Healthcare Research and Quality (US), Rockville (MD)
Kane RL, Shamliyan T, Mueller C, et al. Nurse Staffing and Quality of Patient Care. Rockville (MD): Agency for Healthcare Research and Quality (US); 2007 Mar. (Evidence Reports/Technology Assessments, No. 151.) Executive Summary.