Skip to main content

Table 2 Multinomial regression model of the associations of predisposing, enabling, and need factors with the utilization of nursing care (N = 1549)

From: Individual characteristics associated with the utilization of nursing care in the very old population: a cross-sectional study

  OdpNCU relative to noNCU InpNCU relative to noNCU
Odds ratio 95% CI Odds ratio 95% CI
Predisposing factors
Age 1.056* 1.005 1.109 1.143*** 1.069 1.224
Sex (Ref: female) 1.132 0.721 1.776 0.509 0.235 1.104
Socioeconomic status 0.995 0.985 1.006 0.983* 0.967 1.000
Migration background (Ref: no) 0.665 0.422 1.050 0.928 0.477 1.808
Community type (Ref: ≥500,000)
 100,000 to 499,999a 1.498 0.959 2.340 0.563 0.282 1.124
 50,000 to 99,999 1.178 0.647 2.146 1.253 0.548 2.867
 5000 to 49,999 1.355 0.765 2.399 0.216** 0.079 0.592
Enabling factors
Children (Ref: yes)a 0.713 0.390 1.302 3.827** 1.710 8.564
Partnership status (Ref: no partnership)a 0.577* 0.367 0.908 0.433* 0.195 0.962
Social isolation (Ref: no) 0.761 0.466 1.243 3.939*** 2.056 7.546
Loneliness (Ref: never/rarely)a 2.004 0.950 4.227 2.942* 1.581 7.892
Time spent with others (Ref: often/very often)
 never/rarely 0.867 0.507 1.481 0.434 0.178 1.061
 sometimes 0.939 0.600 1.471 0.800 0.405 1.581
Need factors
Self-rated health status 1.045 0.799 1.366 1.469 0.973 2.218
Basic activities of daily living 0.015*** 0.006 0.036 0.010*** 0.004 0.029
Instrumental activities of daily living 0.181*** 0.111 0.295 0.044*** 0.020 0.094
≥ 5 chronic conditions (Ref: no) 1.203 0.789 1.833 0.423* 0.204 0.878
Physical frailty (Ref: robust)
 pre-fraila 1.436 0.779 2.648 0.242** 0.101 0.577
 frail 1.136 0.528 2.445 0.085*** 0.026 0.272
Dementia (DemTect) 0.625** 0.439 0.890 0.477** 0.309 0.738
Pseudo R-squared (Nagelkerke)b 0.695***
  1. Weighted data
  2. NoNCU no nursing care use, odpNCU outpatient, day, and/or private nursing care use, inpNCU inpatient nursing care use, CI confidence interval, ref reference category
  3. *p < 0.05
  4. **p < 0.01
  5. ***p < 0.001
  6. aThe findings differed between the original and full imputed datasets. For more details, see “Results” section
  7. bCalculated as the mean value from all imputed datasets