Hello,
I'm curious about the various methods other hospitals are using to track their
sepsis mortality rate. I am aware of several possibilities:
1. Using administrative data only: measuring the hospital mortality rate among
patients who have one of several several sepsis-related ICD-10 codes. If you
use this:
a. which codes? We use the list of codes that CMS samples from for SEP-1.
b. Do you require the code to be in only the Primary Diagnosis, or any of
the diagnoses?
c. Do you "scrub" this list to remove any patients based on any other
criteria? For example, comfort care patients
2. Measuring mortality rate among those cases that were manually reviewed for
the presence of severe sepsis and/or were evaluated for the bundle?
3. For hospitals who participate in the University Health Systems
Collaborative, the Sepsis Mortality Index is available which provides some risk
adjustment (Observed/Expected).
4. In the ICD-9 era, we used to use the Dombrowski Method, in which the
mortality rate is measured among the population defined by those with either
one of 3 sepsis codes, or both an infection code and an organ dysfunction code.
I'm not aware of a methodology to recreate this using ICD-10 codes.
Others???
Our experience: We historically used the Dombrowski Method for internal and
external reporting - and post ICD-10 conversion we switched to #1 above (using
the code list from SEP-1, in any position, with no scrubbing). The difference
has been striking - as with this method, the month-to-month variation of our
measured mortality rate is extremely high, changing 50-100% between months.
This obviously makes trend identification more difficult. So I'm curious what
others are doing and how they're thinking about it. We also follow the Sepsis
Mortality Index, which helps with benchmarking.
Appreciate any thoughts or advice.
Sincerely,
joe
Joseph Clement, MS, RN, CCNS
Clinical Nurse Specialist
San Francisco General Hospital
ph: 628 206-6174
pg: 415 327-0220
office: H5841
_______________________________________________
Sepsisgroups mailing list
[email protected]
http://lists.sepsisgroups.org/listinfo.cgi/sepsisgroups-sepsisgroups.org
