Nurse Staffing and Work Environments: Relationships with Hospital-Level Outcomes

by Ann Tourangeau | Mar 01, 2006

Key Implications for Decision Makers

  • While it is important to acknowledge that both death and readmission to hospital are unpreventable outcomes for some patients, persistent wide variations in death and readmission rates (adjusted for differences between patients) across hospitals suggest that some of these outcomes are in excess and are preventable.
  • Hospital decision makers and policy developers can use the evidence from this study to make plans and decisions to minimize 30-day mortality for acute medical patients.
  • To minimize unnecessary deaths of acute medical patients, medical unit staffing should:
    • maximize the proportion of registered nurses in their nursing staff mix;
    • hire and retain baccalaureate-prepared nurses in their nursing staff mix;
    • provide adequate staffing to deliver safe and effective care; and
    • consistently use care maps or protocols to guide patient care.
  • Conclusions cannot be drawn about the determinants of unplanned readmission to hospital as analytic models were statistically non-significant.

Executive Summary

Context

Hospital mortality and readmission rates are used as indicators of the quality of hospital care. The rationale for their use is that some hospitals have structures and processes that minimize unnecessary patient deaths and readmissions better than other hospitals. The challenge is to discover and explain what facet of hospital structures and processes affect patient readmissions and deaths.

In Canada, readmission and mortality rates for hospitalized patients vary widely across hospitals, even after accounting for differences in patient characteristics and mix (1). Once we control for the impact of patients' own characteristics on the outcomes of mortality and unplanned readmission, what is affecting the outcomes? Why do some hospitals have much lower rates than others? What are these hospitals like or what do they do to attain such low rates? Tolerating such variations is an enormous cost and quality of life burden for hospitals, patients, and society. While it is important to acknowledge that death and readmission to hospital are unpreventable outcomes for some patients, persistent wide variations in risk-adjusted rates across hospitals suggest that some portion of these outcomes are in excess and are preventable. Determining the characteristics of hospitals with lower risk-adjusted 30-day mortality and readmission rates is a necessary first step to developing strategies that modify those characteristics to prevent unnecessary death and readmission.

The purpose of this study was to determine what nursing-related and other hospital characteristics affect two hospital quality of care indicators: 30-day mortality and 30-day readmission to hospital. Nursing-related hospital characteristics such as nurse staffing, nurse experience, nursing care delivery system, condition of the nursing practice environment, and nurse response to these environments are included.

Approach

This retrospective study occurred in two phases. In phase one of the project (year 2003), we surveyed all registered nurses and registered practical nurses working in acute medical and combined medical-surgical units in Ontario teaching and community hospitals. Nurses were surveyed to collect data about their work environments, their responses to these work environments, selected processes of care, and their terms and conditions of employment. At the time of this study, there were 75 Ontario adult teaching and community hospitals. The final sample for this study consisted of 3,886 medical nurse respondents working in medical clinical areas in Ontario acute care hospitals.

In phase two of the project (2004-05), we linked nurse survey data with secondary sources of patient and hospital data to test our hypotheses of what determines mortality and readmission. Patients had to have been discharged from an Ontario teaching or community hospital between April 1, 2002 and March 31, 2003 with a most responsible diagnosis of acute myocardial infarction (heart attack), stroke, pneumonia, or septicaemia (blood poisoning). Patients also had to be Ontario residents and 20 years of age or older. Patients discharged from pediatric, psychiatric, and small hospitals and those with a pre-existing diagnosis of cancer or HIV were excluded. The final sample consisted of 46,993 patients.

Results: Key Study Findings

Determinants of 30-day Mortality

  • Lower 30-day hospital mortality rates were associated with hospitals that had:
    • higher proportions of registered nurses (specifically, a 10-percent increase in the percentage of registered nurses is associated with six fewer deaths for every 1,000 discharged patients);
    • higher proportions of baccalaureate-prepared nurses (specifically, a 10-percent increase in the percentage of baccalaureate-prepared nurses is associated with nine fewer deaths for every 1,000 discharged patients);
    • lower amount of nursing staff;
    • higher proportions of nurse-reported adequacy of staffing and other resources (specifically, a 10-percent increase in nurse-reported adequacy of staffing and other resources is associated with 17 fewer deaths for every 1,000 discharged patients);
    • increased routine use of care maps and other kinds of protocols to guide patient care;
    • higher nurse-reported quality of care;
    • lower nurse-reported adequacy of manager ability and support; and
    • higher nurse burnout.

Determinants of 30-day Unplanned Readmission to Hospital

  • Regression models exploring determinants of unplanned readmission were less informative and yielded statistically non-significant results.

Nurse Outcomes

  • Medical nurses report low overall job satisfaction and being more likely to leave their jobs compared to surgical or critical care nurses. Medical nurses are most satisfied with co-workers and least satisfied with work-related control and responsibility.
  • Medical nurses on average report experiencing moderate levels of burnout.
  • Medical nurses report weak professional nursing practice environments.

Implications

This project furthers understanding of how nursing-related structures and processes of care affect two hospital quality of care indicators: 30-day acute medical patient mortality and 30-day unplanned readmission. Our analytic models exploring the determinants of 30-day patient mortality were very informative. Hospital decision makers and policy developers can use these results to make evidence-based plans and decisions to minimize 30-day mortality for acute medical patients. Our analytic models exploring determinants of 30-day unplanned readmission to hospital yielded unstable results and do not lead to evidence-based planning.

Based on study findings, we conclude that to minimize unnecessary acute medical patient 30-day deaths, medical unit staffing should maximize the proportion of registered nurses in their nursing staff mix, hire and retain baccalaureate-prepared nurses, provide adequate staffing to deliver safe and effective care, and consistently use care maps or protocols to guide patient care.

Further Research

Although a respectable 45 percent of variance in 30-day risk- and case mix-adjusted hospital mortality rates was explained by eight hospital care structures and processes, 55 percent of variance remained unexplained. Additionally, our analyses of models exploring determinants of 30-day unplanned readmission to hospital for acute medical patients yielded statistically non-significant results. Clearly, there is much to discover about the determinants of 30-day mortality and readmission rates for hospitalized patients. Future research should also include measures of the environment, including access to in-hospital and outside-of-hospital care.