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

NURSES’ RETENTION AND HOSPITAL CHARACTERISTICS IN NEW SOUTH WALES

4. PROFILE OF PREDICTED NURSES RETENTION PROBABILITIES

An alternative method of examining the impact of changes in the explanatory variables is to use the
model to predict retention probabilities for hypothetical nurses. Table 6 gives an indication of the
range of the estimated impact on the predicted probability of retention for specific nurse types
designated as typical, stayer and quitter. For each type the table shows the contribution to the
predicted index of a group of variables. A zero contribution indicates a variable is set to the omitted
reference level.

The typical nurse characteristics have been set at the means of the continuous variables age hours
and years registered, with some modification for consistency. The dummy variables are set to the
omitted base levels. We have used the means for selected hospital types to produce a profile of the
three nurse types In the case of the typical nurse, the reference hospital is Principal Referral and the
mean of the continuous variables for Principal Referral hospital have been used for the continuous
hospital variables. The typical nurse has a predicted probability of retention of 0.852. The stayer nurse
has been located in a nursing home and the quitter in psychiatric hospital.

Two major contributors to the higher predicted probability, 0.910, of the stayer nurse relative to the
typical nurse, are the impact of country of birth, (foreign born) and mean characteristics of hospital
(nursing home), while the quitter is strongly influenced by age (young), country of birth (UK) and
hours of work (part time). The distribution of predicted probabilities across the sample is highly
skewed to the right. Consequently the stayer nurse retention probability is only slightly higher than
that of the typical nurse, while the quitter is considerably lower.

Table 6: Predicted probabilities for example nurses

Typical:    female, aged 38, Australian born, working 34 hrs, post basic qualification, registered 12 years, clinical job classification,

Mean characteristics of principal referral hospital

Quitter: male, aged 23, UK born, working 20 hrs, no post basic qualification, registered 1 year, activity psychiatric, mean
characteristics of rehab/psych hospital

Stayer:    female, aged 48, foreign born, working 45 hrs, post basic qualification, registered 25 years, job classification clinical,

Non-capital city metropolitan,.mean characteristics of nursing home

While the above discussion gives a picture of the range of variability that the model predicts, the
stereotypical stayer and quitter nurses may not be “near” large proportions of the nursing
population. For example, even though rehab/psych hospitals are significantly more male than the
average, they account for only 3% of the total workforce.

To give an alternative picture related to the distribution of nurse types Table 7 shows the impact of the
mean attributes of those in the top and bottom quartiles of the distribution of predicted probabilities.
The estimated coefficients for the whole sample are applied to the quartile variable means adjusted
slightly for consistency in quadratic terms.

14

<<   16   17   18   19   20   21   22   >>

More intriguing information