Kovner, C. T., Brewer, C. S., Fatehi, F., Jun, J. (2014). What does nurse turnover rate mean and what is the rate. Policy, Politics, & Nursing Practice, 15(3–4), 64–71. doi: 10.1177/1527154414547953

The authors wish to make a correction to the above referenced article. Table 4 provides very misleading data, because all the settings and hospital only rates were calculated differently and therefore the rows should not be compared. Table 4a.

One-year RN Job Turnover Rate Comparison among Three Cohorts in All Settings.

	2004–05 Sample	2007–08 Sample	2010–11 Sample
Percentage (N)	13.4% (18/135)	20.6% (56/271)	22.8% (N = 106/463)

The first row (now Table 4a) includes RN jobs only in any setting. The numerators are those RNs who had started their first jobs at least 13 months prior survey completion and had left that 1st job. The denominators are those RNs who had started their job at least 13 months prior to survey completion. To calculate the numbers in Table 4a we used the dates that respondents gave us for start and stop dates of the job.

Corrected Table 4 are shown in Tables 4a and 4b Table 4b.

One-year Job Turnover Rate Comparison between Two Cohorts Working in Hospitals.

	2004–05 Sample	2007–08 Sample	2010–11 Sample
Percentage (N)	n/a	26.5% (76/287)	20.7% (64/311)

Note: n/a-not available

The second row (now Table 4b) includes RNs working in RN or Non RN-jobs in hospitals. The numerators are those RNs who had started their first job in a hospital at least 13 or more months prior to survey completion and had left that job. The denominators are those RNs who had started their job at least 13 months prior to survey completion. To calculate the numbers in Table 4b, we used a variable that specifically asked if the RN had left his or her first job or not and if he or she left what was the setting for that job.

Table 4a and 4b were calculated differently due to limitations in the data. We did not ask the 2004–05 responders the setting of their first job if they had left that job prior to the survey. From the dates we knew when that job started and stopped. We did ask the setting question of the 2007–8 and 2010–11 responders. When we tried to calculate the hospital turnover rate using dates from responders and combining that with the setting question in some cases we saw inconsistencies in the data. For example, using the dates’ variable a responder said that she had left her first job prior to the first survey, but in the setting question she said that she had not left her first job prior to the first survey. In addition, in the original table we used as a denominator all responders who were in their first hospital job. To make the numbers more readily comparable when we redid the tables we used the same type of denominator for Tables 4a and 4b.

Published Table 4 Table 4.

One-year Turnover Rate Comparison between Three Cohorts.

	2004–05 Sample	2007–08 Sample	2010–11 Sample
All Settings (N)	13.4% (18) (N = 135)	20.6% (56) (N = 271)	22.8% (106) (N = 463)
Hospitals only	Not available	5.2% (78) (N = 1500)	5.9% (64) (N = 1093)

SOURCE: Author’s unpublished data

Notes: One year Percentage Turnover for New Nurses from 14 states who had started their job as a registered nurse at least 13 months prior to data collection for three different cohorts compared to turnover rates of those nurses whose first job was in a hospital

The changes that we made to the tables necessitate changes to the narrative in the article. The replacement sentence is in bold below. The page 68 right column, paragraph starting with “To make the point that turnover varies by setting …” in light of the new tables should read, “To make the point that turnover varies by setting, we also calculated the turnover rate for those whose first job was in a hospital for the 2007–2008 and 2010–2011 samples. In Table 4 , note that the hospital-specific turnover rates are higher than all setting rates for the 2007–2008 sample and hospitals rates are lower for the 2010–2011 sample, although it is a small difference and the calculation procedure was similar, the calculation procedures used different data sources. This is consistent with our findings (Kovner et al., 2014) that new RNs from the latest cohort were less likely to work in hospitals than earlier cohorts because they were unable to get hospital positions rather than because they preferred nonhospital settings.

While we wish that we had focused on the differences in calculation methods prior to publication of this article, finding this after publication adds to the argument about how important it is to look at how turnover numbers are calculated and to what settings do the numbers refer.