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

NURSES’ RETENTION AND HOSPITAL CHARACTERISTICS IN NEW SOUTH WALES

From Table 3 we find that 8% of our sample is composed of male nurses, the average age is quite
high, just under 39 years old, and 23% of the sample is composed of individuals not born in Australia.
Hours of work per week for RNs in public hospitals average 36.2, which is in the full-time
employment range.¹¹ On average, these nurses have been registered for 15 years and 58% have post-
basic qualifications. Despite the age and the years of experience (as measured by years since
registered), 64% of these RNs are still at the basic job classification (level 1 RN), suggesting that
promotions are not that common amongst RNs.¹²

The hospital data includes many more hospital characteristics than those used in our analysis. Many
specifications were estimated and because of correlations across the characteristics, we found many
sets of variables with similar explanatory power. The model presented in this paper uses a
parsimonious set of variables which can be easily interpreted and which perform well. Detail on the
specification is provided in the following section.

Hospital characteristics were chosen to measure variations in size (separations and bed days),
complexity (acute versus non-acute separations and costs, length of stay, high cost procedures, and
ANDRG weight), intensity (separations per nurse, emergency admissions per nurse, wait time),
staffing (equivalent full-time number of nurses, non-nurse clinical staff, non-clinical staff, visiting
medical officers measured in expenditures) and expenditures (total expenditures excluding visiting
medical officers). Other characteristics were included because they mattered in explaining retention;
for example baby separations. In order to explain the meaning of the variables without taking up too
much space in the main text we include a glossary of the characteristics in Appendix 3.

11 Unfortunately, we do not have information on shift work.

12 The incidence of missing values is not high for each variable (e.g. 3% for age, 2% for job classification) but since they
involve different individuals, excluding all observations with missing values restricts the sample a lot more than is
apparent from Table 3.

<<   9   10   11   12   13   14   15   >>

More intriguing information