SNF Occupancy has been an ongoing concern for operators, regulators, and SNFists alike. Much has been written about the topic, and theories abound about rebound and the like.
I wanted to create a simple distillation of the relationship between star rating and occupancy during COVID. So, what I did was take the star rating of all SNF’s circa Nov-19 and then compare that to occupancy circa Nov-21. I then fit an ordinal regression to tease out the relationship between the Nov-19 star rating and Nov-21 occupancy.
As the plot shows:
- Star ratings (with the exception of 3 Stars) converged in probability at around 60% occupancy
- Star ratings of 4 and 5 showed the highest probability of high occupancy (especially above 65% occupancy
- A star rating of 5 was especially robust above an occupancy of 80%, even relative to a star rating of 4
- A star rating of 3 showed little movement across the spectrum of SNF occupancy – remaining at around a 15% probability
- Star ratings of 1 and 2 showed the highest probability of low occupancy (especially below 50% occupancy)
- A star rating of 1 was especially fragile below an occupancy of 40%, even relative to a star rating of 2
