Running example: predicting adverse events in hyperkalemia
The methods are much longer than the background. Open by stating what type of study it was and where the data came from, then cover:
Tap each sentence to see which job it does.
This study had a single outcome, so there are no secondary outcomes to list. When a study has them, they follow the primary outcome. Here the model description comes after the outcome; describing the analysis last is common and reads naturally.
Common mistakes