Nurses' retention and hospital characteristics in New South Wales, CHERE Discussion Paper No 52

NURSES’ RETENTION AND HOSPITAL CHARACTERISTICS IN NEW SOUTH WALES

Our analysis sample consists of RNs working in public hospitals (including nursing homes) in NSW
in 1996. The definition of working applied in this study uses the reported labour force status in the
NRB questionnaire. A nurse is regarded as working if she/he reported any nursing work in NSW.⁷ For
each person working in 1996, an indicator variable measuring retention is constructed as follows.
Using the individual identifier, we determine if this nurse is still working as a nurse in NSW in 1997, if
he/she is working then the retention variable is given a value of one. Otherwise, the variable has a
zero value. Note that the nurse is considered as continuing to work even if he/she has left the public
hospital sector and is working as a nurse in another premise. Also it is possible that a nurse no
longer working in NSW is still working as a nurse in a different state, we cannot fully identify these
individuals from those who have quit nursing altogether. Hence our definition of retention is not
restricted to staying on in a public hospital but it is restricted to working as a nurse in NSW.

Table 1 presents sample sizes and retention rates for the total sample of RNs in NSW and our
analysis sample. We find that 45,632 RNs worked in NSW in 1996⁸ and of these, 35,605 were still
working in NSW in 1997. The sub-sample of RNs working in public hospitals in 1996 numbers 25,568
or 56% of all RNs. The retention rate for this group is slightly higher, with 78.9% found to be working
in NSW in 1997. Most of the public hospital RNs still working in 1997 had remained in public hospital
(74%) and only 1,273 RNs moved to other work premises. Given the relatively small sample of those
leaving the public hospitals but still working as RNs in 1997, this decision is not modelled separately.

The last sample presented in Table 1 is the one used in the econometric model below. It consists of
16,393 public RNs who worked in 1996 and who provided a valid post code allowing a match with the
characteristics of their hospital.⁹ As reported in Table 1, their retention rate is slightly higher than the
total sample of RNs: 80.5% compared to 78.9%. We conducted checks for selectivity bias and these
are detailed in Appendix 1. The results suggest that selection bias is not a serious concern.

Table 1. Retention Rates- Total Sample and Analysis Sample

7 An alternative measure for participation, positive hours of work, is available but contains some missing values.
Using reported status is consistent with the measures presented in the Profiles of the Nursing Workforce.

8 As mentioned earlier, this number is believed to constitute over 90% of the total population of RNs in NSW in 1996.

9 The remaining respondents either did not supply a workplace postcode or the postcode could not be matched to a

4              hospital postcode.

<<   6   7   8   9   10   11   12   >>

More intriguing information