A critique of “Education levels of hospital nurses and surgical patient mortality”

Article Outline

• References
• Copyright

Aiken, Clarke, Cheung, Sloane, and Silber (2003) concluded that surgical patients receiving care from nurses educated at the Associate level were at higher risk for postsurgical mortality in Pennsylvania hospitals. To quote the authors, “In hospitals with higher proportions of nurses educated at the baccalaureate level or higher, surgical patients experienced lower mortality…” (p. 1617).

Aiken et al. (2003) based their analysis on hospital-level data, and the authors admit that hospitals with more ADNs were characterized by fewer technological resources and higher workloads. One would expect resource-deficient hospitals to incur higher mortality rates, regardless of staffing characteristics. Yet, Aiken et al. cite nursing education levels as the primary cause of higher mortality, which is a somewhat surprising inference (see Table 1).

Table 1. Percentage of the hospital nursing workforce with a BSN degree or higher

The data reveal that, while a relationship exists between workforce characteristics and mortality, a strong relationship also exists between average patient age and mortality (see Fig. 1). In fact, average patient age accounts for 95.1% of the variation in hospital mortality rates (r = .975). More important, the high correlation between average patient age and nurse staffing characteristics introduces a high degree of multicollinearity that must be corrected, controlled for, or both.

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 1. 

  Scatter diagram and trend line of hospital mortality and average patient age.

Because the difference in average patient age is a significant covariant, an examination of background mortality rates for the general population is strongly indicated. An estimate of U.S. mortality rates (Centers for Disease Control and Prevention, 1999) reveals a sharp rise in general mortality (all causes) between the ages of 57.3 and 61.3, which are the lowest average age and the highest average age of the hospital groups surveyed, respectively (see Fig. 2, Table 2).

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 2. 

  Plot of general U.S. mortality: all causes.

Table 2. U.S. mortality

When hospital mortality rates are age adjusted for background mortality rates of the general population (i.e., the ratio of Pennsylvania hospital mortality to U.S. mortality; see Table 3), a weak relationship that counters Aiken et al.'s (2003) findings emerges (see Fig. 3).

Table 3. Ratio of Pennsylvania hospital mortality reported by Aiken et al. (2003) and U.S. mortality

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 3. 

  Scatter diagram and trend line of adjusted mortality and average patient age.

Namely, as average patient ages increase (and, concurrently, the percentage of ADN and Diploma nurses also increases), patient mortality, when age adjusted for the background mortality of the U.S. population, actually decreases slightly (y = −0.0279x + 3.6327, r = −.536). Although there is no sufficient evidence to support an alternate hypothesis, the analysis provides strong evidence that shows that the inference of Aiken et al. (2003) is not supported.

Back to Article Outline

References 

1. Aiken LH, Clarke SP, Cheung RB, Sloane DM, Silber JH. Education levels of hospital nurses and surgical patient mortality. Journal of the American Medical Association. 2003;290:1617–1623
   • View In Article
   • CrossRef
2. Centers for Disease Control and Prevention, National Center for Health Statistics Data Warehouse . Death rates from 113 selected causes, United States, specified Hispanic origin, race for non-Hispanic population. Atlanta, GA: Centers for Disease Control and Prevention; 1999;
   • View In Article