Utilization of Non-US Educated Nurses in US Hospitals
Utilization of Non-US Educated Nurses in US Hospitals
The characteristics of the nearly 1.3 million surgical patients included in the study are given in Table 1. The average age across all patients was just over 61 years. Forty-three percent of the patients were men, and 28% were emergency admissions. Overall, 2.2% of these surgical patients died within 30 days of admission, and among the 35% of the patients who developed complications during their hospital stay, 5.4% died within 30 days of admission (failure-to-rescue). The most common operations undergone by these patients were orthopedic operations (52%), followed by operations for digestive system diseases and disorders (22%) and hepatobiliary systems diseases and disorders (11%). The most common comorbidities among this group of patients were hypertension (48%), chronic pulmonary disease (15%), uncomplicated diabetes (15%) and congestive heart failure (14%).
Demographic characteristics of non-US and US-educated nurses are given in Table 2. The majority of non-US-educated nurses in the study were educated in the Philippines (63.5%). Compared with US-educated nurses, non-US-educated nurses were slightly younger, more likely to be male, and decidedly more likely to have obtained a baccalaureate degree in nursing (68 vs. 40%). A significantly larger proportion of US-educated nurses (23 vs. 17%) were dissatisfied with their jobs; however, despite their greater dissatisfaction, US-educated nurses were slightly less likely than non-US-educated nurses to report intentions to leave their current jobs (13 vs. 14%). There were no significant differences between the two groups regarding burnout or their ratings of the nurse practice environment and the quality of care in their hospitals.
Across 665 hospitals in the four states for which we had patient data, non-US-educated nurses as a proportion of all registered nurses ranged by hospital from 0 to 77%. The percentage of non-US-educated nurses was <5% in 224 study hospitals (or in 34% of them), from 5 to 25% in 327 (49%) of the hospitals and >25% in 114 (17%) of the hospitals. Hospitals with >25% non-US-educated nurses are referred to below as 'high utilization'. Table 3 also shows there were no significant differences between the percentages of non-US-educated nurses in hospitals by technology status, though non-teaching hospitals and small hospitals in general were less likely to employ higher proportions of these nurses.
Mortality and Failure-to-Rescue.Table 4 shows the effects, estimated by odds ratios, of nurse staffing, the nurse practice environment and the percentage of foreign-educated nurses on mortality and failure-to-rescue, derived from unadjusted (bivariate) models and from two fully adjusted models. One of the fully adjusted models allows only the main effects of staffing and the percent of non-US-educated nurses whereas the other includes two significant interactions involving the nursing characteristics. The fully adjusted models control for patient characteristics (age, gender, transfer status, diagnosis-related group and the comorbidities specified by the Elixhauser Comorbidity Index), nurse characteristics (including the nurse practice environment, proportion of nurses with bachelor's degrees (BSN), proportion of nurses in ICU, proportion of nurses in medical-surgical units) and hospital characteristics (bed size, teaching status and technology status).
When we estimate the effects on the mortality measures of staffing, percent of non-US-educated nurses and the nurse practice environment from bivariate models, only the practice environment is significant. In the main effects model with controls (middle panel Table 4 ) both the effects of nurse staffing and the practice environment are significant, on both outcomes, while the main (or direct) effect of the percentage of non-US-educated nurses is not significant. However, the main effects model is not the appropriate model to use to describe the effects of the three nursing factors, because the fully adjusted model with interactions (final panel of Table 4 ) suggests that there are highly significant interactions between staffing and the practice environment, and between staffing and the percentage of non-US-educated nurses. The interaction between staffing and the work environment involves a more pronounced effect of nurse staffing in hospitals with better work environments, as we have described in detail in a previous paper. The interaction between staffing and the percentage of non-US-educated nurses can be best understood by considering Table 5 .
Table 5 shows the relationship between the proportion of non-US-educated nurses and mortality and failure-to-rescue in hospitals with different staffing levels. For hospitals that are well staffed, there was no significant relationship between high non-US-educated nurse utilization and mortality. For hospitals with higher ratios of patients to nurses (5 patients per nurse or more), high non-US-educated nurse utilization was associated with increased mortality. The odds on 30-day mortality and failure-to-rescue were both significantly higher in hospitals with 25% or more non-US-educated nurses than in hospitals with >25% of these nurses by factors of 1.1, 1.2 and 1.3 in hospitals with 5:1, 6:1 and 7:1 patient to nurse ratios, respectively, even after controlling for selected nursing factors (% bachelor's nurses and nurse practice environment) and hospital characteristics (bed size, teaching status and technology).
In our sensitivity analysis, we reassigned the 358 Canadian nurses (7.4%) in our sample to the US-educated group and reran all regression models. We found no differences in the significance of all odds ratios in Table 4 and Table 5 .
Results
The characteristics of the nearly 1.3 million surgical patients included in the study are given in Table 1. The average age across all patients was just over 61 years. Forty-three percent of the patients were men, and 28% were emergency admissions. Overall, 2.2% of these surgical patients died within 30 days of admission, and among the 35% of the patients who developed complications during their hospital stay, 5.4% died within 30 days of admission (failure-to-rescue). The most common operations undergone by these patients were orthopedic operations (52%), followed by operations for digestive system diseases and disorders (22%) and hepatobiliary systems diseases and disorders (11%). The most common comorbidities among this group of patients were hypertension (48%), chronic pulmonary disease (15%), uncomplicated diabetes (15%) and congestive heart failure (14%).
Demographic characteristics of non-US and US-educated nurses are given in Table 2. The majority of non-US-educated nurses in the study were educated in the Philippines (63.5%). Compared with US-educated nurses, non-US-educated nurses were slightly younger, more likely to be male, and decidedly more likely to have obtained a baccalaureate degree in nursing (68 vs. 40%). A significantly larger proportion of US-educated nurses (23 vs. 17%) were dissatisfied with their jobs; however, despite their greater dissatisfaction, US-educated nurses were slightly less likely than non-US-educated nurses to report intentions to leave their current jobs (13 vs. 14%). There were no significant differences between the two groups regarding burnout or their ratings of the nurse practice environment and the quality of care in their hospitals.
Hospitals
Across 665 hospitals in the four states for which we had patient data, non-US-educated nurses as a proportion of all registered nurses ranged by hospital from 0 to 77%. The percentage of non-US-educated nurses was <5% in 224 study hospitals (or in 34% of them), from 5 to 25% in 327 (49%) of the hospitals and >25% in 114 (17%) of the hospitals. Hospitals with >25% non-US-educated nurses are referred to below as 'high utilization'. Table 3 also shows there were no significant differences between the percentages of non-US-educated nurses in hospitals by technology status, though non-teaching hospitals and small hospitals in general were less likely to employ higher proportions of these nurses.
Mortality and Failure-to-Rescue.Table 4 shows the effects, estimated by odds ratios, of nurse staffing, the nurse practice environment and the percentage of foreign-educated nurses on mortality and failure-to-rescue, derived from unadjusted (bivariate) models and from two fully adjusted models. One of the fully adjusted models allows only the main effects of staffing and the percent of non-US-educated nurses whereas the other includes two significant interactions involving the nursing characteristics. The fully adjusted models control for patient characteristics (age, gender, transfer status, diagnosis-related group and the comorbidities specified by the Elixhauser Comorbidity Index), nurse characteristics (including the nurse practice environment, proportion of nurses with bachelor's degrees (BSN), proportion of nurses in ICU, proportion of nurses in medical-surgical units) and hospital characteristics (bed size, teaching status and technology status).
When we estimate the effects on the mortality measures of staffing, percent of non-US-educated nurses and the nurse practice environment from bivariate models, only the practice environment is significant. In the main effects model with controls (middle panel Table 4 ) both the effects of nurse staffing and the practice environment are significant, on both outcomes, while the main (or direct) effect of the percentage of non-US-educated nurses is not significant. However, the main effects model is not the appropriate model to use to describe the effects of the three nursing factors, because the fully adjusted model with interactions (final panel of Table 4 ) suggests that there are highly significant interactions between staffing and the practice environment, and between staffing and the percentage of non-US-educated nurses. The interaction between staffing and the work environment involves a more pronounced effect of nurse staffing in hospitals with better work environments, as we have described in detail in a previous paper. The interaction between staffing and the percentage of non-US-educated nurses can be best understood by considering Table 5 .
Table 5 shows the relationship between the proportion of non-US-educated nurses and mortality and failure-to-rescue in hospitals with different staffing levels. For hospitals that are well staffed, there was no significant relationship between high non-US-educated nurse utilization and mortality. For hospitals with higher ratios of patients to nurses (5 patients per nurse or more), high non-US-educated nurse utilization was associated with increased mortality. The odds on 30-day mortality and failure-to-rescue were both significantly higher in hospitals with 25% or more non-US-educated nurses than in hospitals with >25% of these nurses by factors of 1.1, 1.2 and 1.3 in hospitals with 5:1, 6:1 and 7:1 patient to nurse ratios, respectively, even after controlling for selected nursing factors (% bachelor's nurses and nurse practice environment) and hospital characteristics (bed size, teaching status and technology).
In our sensitivity analysis, we reassigned the 358 Canadian nurses (7.4%) in our sample to the US-educated group and reran all regression models. We found no differences in the significance of all odds ratios in Table 4 and Table 5 .
Source...