Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Nurse staffing, mortality and burnout


An Audit Commission report in 2001 [1] could come to no conclusion whether nurse staffing levels on wards had any influence on clinical risk. Part of the reason was the paucity of data from UK hospital trusts that allowed that question to be addressed. An impressive large study from the USA [2] concludes that higher patient:nurse ratios leads to increased patient mortality, higher levels of nurse burnout, and nurse job dissatisfaction.


The study was conducted in Pennsylvania, and ultimately looked at staffing levels across 168 of 210 acute care hospitals with discharge data for surgical patients in targeted diagnosis-related groups of general surgery, orthopaedic surgery and vascular surgery. Hospital characteristics controlled for were size (≤100, 101-250 and ≥250 hospital beds), teaching status (none, minor or major teaching load), and high technology (open-heart surgery, or transplantation were measures of high technology).

Surveys were mailed to a 50% random sample of registered nurses in Pennsylvania, with a 52% response rate, and with 10,200 working in hospitals. There had to be at least 10 registered nurses returning a questionnaire from a hospital. Half the hospitals had more than 50 nurse respondents. The nurse staffing measure was taken from nurses who reported having responsibility for at least one but fewer than 20 patients on their last shift, regardless of time or specialty.

Discharge reports were obtained for 232,000 patients between the ages of 20 and 85 years over 18 months in 1998 and 1999. Outcomes used were 30-day mortality, and deaths within 30 days of admission among patients who experienced complications (pneumonia, hypotension, shock, for example).


The ratio of patients to nurses varied from 4:1 to 8:1, and the percentage of hospitals, nurses and patients in each category is shown in Figure 1. Almost 19 in 20 nurses were women, 4 in 10 had a nursing degree, and they had an average experience of 14 years in nursing.

Figure 1: Distribution of patients per nurse in Pennsylvania

Half the patients had undergone orthopaedic surgery, and nearly 40% digestive tract and hepatobiliary surgery. Some 54,000 (23%) experienced a major complication not present on admission, and 4,535 (2.0%) died within 30 days. In patients with complications 8.4% died.

Higher patient:nurse ratios were significantly associated with emotional exhaustion and greater job dissatisfaction. Increasing the ratio of patient to nurse by one increased burnout and job dissatisfaction by 23% and 15% respectively. An increase from 4:1 to 8:1 more than doubled job dissatisfaction. One in 10 nurses satisfied with their jobs intended to leave within 12 months. For dissatisfied nurses, this was closer to 1 in 2.

Higher patient:nurse ratios were significantly associated with patient mortality. Increasing the ratio of patient to nurse by one increased patient mortality by an average of 7%. Increases in nurse staffing from 4:1 to 6:1 or 8:1 would increase patient mortality by 14% and 31% respectively.

For Pennsylvania, the impact of nurse staffing on additional deaths for all patients and those with complications is shown in Table 1. With an average patient:nurse ratio of 4:1 to 8:1 there were 4,500 deaths. The implication of an overall patient nurse ratio of 4:1 would be 500 fewer deaths, and with a ratio of 8:1 there would be 500 more deaths over 18 months.

Table 1: Modelled effect of nurse staffing on patients

Additional deaths per 1,000 patients

Patient:nurse ratio

All patients

With complications

6:1 rather than 4:1 2.3 8.7
8:1 rather than 6:1 2.6 9.5
8:1 rather than 4:1 5.0 18.2


The background to this paper is that in 2003 California will mandate that the patient:nurse ratio in its hospitals will not exceed 6:1, falling to no more than 5:1 when fully implemented. Someone, somewhere, is taking these issues seriously. Extrapolation of the Pennsylvania experience to California would mean 2,000 fewer deaths over 18 months.

Are we surprised that poor staffing levels lead to burnout, dissatisfaction, and a desire to leave the job? Perhaps not, but job satisfaction and burnout were specifically measured in this study. There is a standardised tool, the Maslach Burnout Inventory, and even a Burnout Inventory Manual. Someone, somewhere, is taking these issues seriously.

The interesting discussion in this paper refers to the high cost of staff replacement in the USA, and speculates that increasing staffing levels could not only save patient lives and decrease nurse turnover, but reduce hospital costs. Again, someone, somewhere, is taking this issue seriously.

A literature is beginning to appear from the USA, where they do take these issues seriously. Bandolier 103 looked at another US study showing that lower nurse staffing in the ITU resulted in more patient complications. In California they are saying that staffing levels and safety are linked, and that below a certain level safety is compromised and will not be tolerated. The implications of that train of thought are far-reaching indeed.


  1. Audit Commission. Acute hospital portfolio: review of national finding: ward staffing. 2001: 3.
  2. LH Aiken et al. Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA 2002 288: 1987-1993.
previous or next story in this issue