Center to Advance Palliative Care

You probably should NOT try and do a comparison of your small PC population with a matched group, or try to replicate the statistical rigor of the cost study (Archives of Internal Medicine, 2008). That is why Sean Morrison did the multi-site study - it took a lot of patients and a big comparison group to get to a scale where reliable matching could occur. Don't try and do it unless you have a lot of patients (>1000 at least) and a lot of analytical resource with the skill base. It just will not get a useful result.

Concurrent Session B, Using Cost Avoidance Data, from the October 2010 CAPC Seminar has some info on this topic; check the CAPC store for the CD on this. This should orient you to some of the issues. Focus on slide 15 re "before and after" and slide 20 which discusses some of the data decisions you will need to make.



We do recommend running the "before and after" comparison to use your own data, and running the comparison ONLY for patients that died in the hospital. This will allow you to generate local info to compare to a couple of the figures in the cost article.

Good luck; this is not for the faint of heart. Critical step is to engage a finance partner in planning this out, and use the cost study as a strong anchor.