News Article

RSI v2.0 - Part 3

Author: Rohan Cattell | Posted: 13/08/2019 | Views: 587

RSI v2.0 - Part 3

Rohan Cattell,
Chief Data Scientist

The Health Roundtable



This is the third part of a 3-part series on the new Health Roundtable RSI model. Part 1 looked at our rationale for updating the algorithm for expected length of stay. Part 2 was focused on describing the new model and part 3 will look at the anticipated impact on members.

We now have an intended release date for reports using the new RSI model. In early September 2019 our Jul 2018 – Jun 2019 reports will be released, and these will be updated to use RSI v2. The core reports that are affected are:

-          Hospital KPI report

-          Inpatient briefings

-          Top 10 report

-          Executive briefings

-          Departmental reports

Improvement group reports that show RSI will also update to RSI v2, except for any use of RSI for subacute episodes, which will remain on v1. We have not included subacute episodes in RSI v2.0 as we determined this requires a separate piece of work to get right.

In addition to moving to RSI v2 we are bringing the reference period forward to a single year period from Jul 2017 – Jun 2018. As RSI tends to come down over time, moving the reference data period forward means RSI results will go up. In fact, this effect is likely to be larger than any effect from the change of model, at least at the whole hospital level. We have moved the reference period before, so hopefully this process will be familiar to many.

In addition to the changes due to the reference period and the new Expected LOS (ELOS) values, there may be some additional exclusions from the RSI in some cases. The most important of these is that deaths are now excluded, so the number of episodes included and the bed days calculated, will exclude those episodes. This is unlikely to have a major impact except in individual DRG and department reports for cohorts with a large proportion of deaths. You may see some discrepancies though, between numbers of episodes in the RSI, and other calculations in the same report, such as for Average LOS, which will not exclude these episodes.

As the reference data is changing, it will be important to focus on your relative position to other hospitals and not on the absolute number that you land on in any particular report. Expect the number to change (go up) but your relative position less so. Having said that, if you previously had a high rate due to poor documentation (hence low ELOS), you will find that the more fine-grained analysis using ECCS will probably not be kind to you.

At the whole hospital level, large hospitals will move less, in relative terms, than smaller ones. This is because the differences between RSI models are more likely to be evened out over a larger set of data. As you drill down to finer levels, individual DRGs or departments will vary more between versions.

Your top 10 reports should be broadly similar but will have some changes due to particular DRGs being more or less affected by the new model. Expect bed day savings numbers to change.

What are the top things to do when the new reports come out?

1.       Check the RSI in your hospital KPI report.  You should expect that it will have increased, but look at where you sit relative to others. Compare the relative position to the previous report. Compare trends between the two reports.

2.       Check your top 10s for any major changes. Are there new opportunities that have been identified?

3.       Check DRG reports for the top 10 DRGs that are of interest to you. Again, expect to have increased in absolute terms but focus on the relative position.

4.       Check the departmental reports for any changes. Are there department heads who read these reports and who need to be prepared for interpreting the changes?

With the increased use of a detailed complexity measure, there may be some focus on the question of documentation. Poor documentation can affect the results but be careful to avoid an over-reliance on this explanation for a high RSI. Triangulate with other data, look at your average LOS. Does this reinforce the message or throw more doubt? Never allow perceived documentation issues to be an excuse for no action. Work in parallel on improving documentation and performance.

In the lead up to the change we’ll be releasing a short video explainer and publishing an FAQ about the changes. If you’re expecting to be the one who has to explain the changes to your colleagues, I highly recommend engaging with this material.

Our chat box staff will be available to answer questions, so if there are things you want to understand that aren’t covered by our FAQ, reach out or ask a question through our website chat box. Your local Health Roundtable Client Relationship Manager will also be able to cover the new RSI in any regular briefings your organisation has booked with them.


You must be logged in to submit a comment.

Sydney Office

Level 2
120 Chalmers Street
Surry Hills, NSW 2010

Call Us

In Australia: +61 (02) 8041 1421

In New Zealand: +64 3741 3123

© 1995-2019 The Health Roundtable   

Partners and Members