Table 5.
Two-factor logistic regression analysis on MAT outcome
| General model type III tests of fixed effects | ||||||
|---|---|---|---|---|---|---|
| DFa | Fb | p c | ||||
| Year | 5 | 10.79 | < .0001 | |||
| Reason for sampling | 2 | 7.66 | 0.0005 | |||
| Interaction | 10 | 1.58 | 0.1049 | |||
| Effect | n | Positives % | ORd | CIe | p f | |
| lower | upper | |||||
| Year | ||||||
| 2011 (ref) | 5908 | 16.3 | 1 | – | – | – |
| 2012 | 5472 | 19.0 | 1.490 | 1.204 | 1.843 | 0.0023 |
| 2013 | 5568 | 19.3 | 1.251 | 0.996 | 1.572 | 0.1292 |
| 2014 | 5106 | 19.5 | 1.296 | 1.026 | 1.638 | 0.1935 |
| 2015 | 4733 | 21.3 | 1.951 | 1.549 | 2.457 | 0.0014 |
| 2016 / 01 to 09 | 3042 | 30.9 | 2.565 | 1.862 | 3.533 | 0.0240 |
| Reason for sampling | ||||||
| Monitoring (ref) | 1813 | 16.9 | 1 | – | – | – |
| Reproductive problems | 12,017 | 22.1 | 1.478 | 1.189 | 1.838 | 0.0017 |
| Unknown reasons | 15,999 | 19.1 | 1.326 | 1.068 | 1.646 | 0.0093 |
aDegrees of freedom
bF-test statistics for model parameter
cp-value for F-Test
dodds ratios adjusted for year, reason for sampling, and interaction
elower/upper bound 95% confidence interval for the odds ratio
fp-value for Wald test