Factors related to the intention to leave and the decision to resign among newly graduated nurses: a complete survey in a selected prefecture in Japan

Table 4 Multivariate logistic regression model for the upper quartile of intention to leave (n = 493)

	Odds ratio	95 % confidence interval	P value
Employment and organizational characteristics
Desirable position (in a word)^a	1.281	0.744–2.205	0.372
Working more than 51 h per week^b	1.224	0.684–2.188	0.496
Break facilities and amount of permitted rest time in the hospital^c	1.464	0.831–2.581	0.187
Hospital type (establishment)^d	1.796	0.999–3.229	0.050
Basic charge for hospitalization^e	1.476	0.717–3.038	0.291
Individual factors
Being personally suited for nursing work^f	0.661	0.550–0.794	<0.001
GHQ-12^f	1.075	0.993–1.164	0.076
CF-13^f	1.112	1.025–1.207	0.011
Psychosocial factors in the work environment
Supervisor support^f	0.824	0.691–0.981	0.030
Coworker support^f	1.000	0.871–1.149	0.994
Quantitative demands^f	1.017	0.993–1.042	0.166
Cognitive demands^f	1.001	0.975–1.028	0.940
Emotional demands^f	0.989	0.973–1.006	0.216
Demands for hiding emotions^f	1.007	0.991–1.024	0.382
Sensory demands^f	1.001	0.979–1.023	0.953
Presence of a role model^g	3.932	1.987–7.781	<0.001

The result of the multivariate logistic regression analysis for independent variables (individual factors and psychosocial factors in the work environment) were entered into the equation, along with the variables of employment and organizational characteristics, using a stepwise method (backward selection method). Hosmer–Lemeshow goodness of fit χ ² = 11.089, df = 8, P = 0.196
^a1 = disagree, 0 = agree
^b1 = agree, 0 = disagree
^c1 = inadequate break facilities and an insufficient amount of permitted rest time, 0 = adequate break facilities and a sufficient amount of permitted rest time
^d1 = not public hospital, 0 = public hospital
^e1 = ratio of the number of patients to that of nurses is 7:1, 0 = ratio of the number of patients to that of nurses is not 7:1
^fContinuous variable
^g1 = lack of a role model, 0 = presence of a role model

ISSN: 1347-4715