Table 1 The Candidate Models to Predict no-show rate across healthcare centers
Model | Description |
|---|---|
0 | Predicting no-show status from a central intercept term, will always predict the most common outcome. |
1 | Allows the intercept to vary by facility and borough. |
2 | Allows the intercept to vary by facility and borough. Lead Time and Empaneled Provider have the same linear effect across all facilities and boroughs. |
3 | Allows the intercept to vary by facility and borough. Lead Time and Empaneled Provider can have different linear effect across all facilities and boroughs. |
4 | Allows the intercept to vary by facility and borough. Lead Time and Empaneled Provider have the same non-linear effects across all facilities and boroughs. |
5 | Allows the intercept to vary by facility and borough. Lead Time and Empaneled Provider can interact, so that the non-linear effect of lead time depends on whether respondents are visiting their empaneled provider or not. |
6 | Allows the intercept to vary by facility and borough. Empaneled Provider has the same effect everywhere, but lead time can have different non-linear effects in different facilities, but are shrunk towards a central, non-linear function. |
7 | Allows the intercept to vary by facility and borough. Empaneled Provider has the same effect everywhere, but lead time can have different non-linear effects in different facilities, without any constraints on variations between facilities. |