Answers are grounded in this paper and cite their sources.
Nosocomial pneumonia and its impact on length of stay are major healthcare concerns. From an epidemiological perspective, nosocomial pneumonia is a time-dependent event. Any statistical analysis that does not explicitly model this time dependency will be biased. The bias is not redeemed by adjusting for baseline information.
No takes yet. Share an insight, caveat, or question.
Beyersmann et al. (2009) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: