Table 3.
Logistic regression models for prediction of the Maintenance Care (MC) -recommendation
| Ordinary (one level1) Logistic regression | Multi-level (two levels1) Logistic regression | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
|
Factors |
OR |
95% CI |
p-value |
OR |
95% CI |
p-value |
|||||
| Many previous episodes |
3.5 |
1.6 – 7.7 |
0.002 |
4.1 |
1.5 – 10.7 |
0.005 |
|||||
| Long duration |
1.4 |
0.7 – 2.8 |
0.39 |
1.5 |
0.6 – 3.6 |
0.35 |
|||||
| Definitely better by the 4th visit |
1.4 |
0.7 – 3.0 |
0.33 |
1.2 |
0.5 – 3.0 |
0.62 |
|||||
| Akaike’s Index (AIC)2 | 244.69 | 225.41 | |||||||||
1 The levels are patients (n = 252) for the one-level model and patients (n = 252) and chiropractors (n = 33) for the two-level model. 2 A smaller value for AIC indicates better fit to data.