We’ve been getting quite a few questions about the role of volume in the MPIRICA Quality Score. Mostly people want to know whether volume is a good indicator for quality, and I thought we could address it with a quick blog.
There is Conflicting Data About Volume
To give you a bit of context, there has been a lot of discussion since a “volume to outcome relationship” in healthcare was first reported over three decades ago by Luft, Bunker and Enthoven in the New England Journal of Medicine. Today volume is generally considered the only factor that have any potential correlation to quality, so most people who are concerned with quality look at volume.
At the risk of stating the obvious, healthcare quality is complex, and this is especially true in major procedures such as surgery. To quote the prominent Swedish surgeon Ingemar Ihse in the Annals of Surgery, “besides case volume, a variety of other factors underlie differences in outcome: selection of patients, preoperative preparation, skills of [the team], postoperative care, nurse staff levels, surgical judgment and skill, and others. Thus, volume is not an exclusive indicator of outcome….”
While there has been a wealth of research showing a positive correlation between volume and outcome, countered by research showing no or even negative correlations, the fact remains that volume is “not a formal indicator of quality, but rather a structural characteristic.” Put it in plain English: volume is a contributor to quality, but does not guarantee quality.
Accounting for Quality
With this understanding, MPIRICA does not explicitly account for volume when calculating quality. We let the outcomes speak for themselves. How volume plays out in our scoring is in the basic statistical tenant that “the more data you have, the more certain you are of the results.” Volume contributes to the statistical significance of any metrics calculated, no more, no less.
So from a practical standpoint, how should a non-statistician think about this?
To that, we say: “let’s talk baseball!”
Highest Batting Average vs Most ‘At Bats’
When you want to know the performance (aka outcomes) of your batter, would you rather know their ‘at bats’ (volume) or their ‘batting average’ (quality)? The obvious answer is that you’d rather have the batting average. While a greater number of at bats can improve a player’s batting average, you wouldn’t use that as a substitute for their performance if you have their batting average handy.
Same thing with volume and quality scores, in absence of all other data, volume is the only indicator that could potentially represent quality, but if you have a quality score which is calculated from ACTUAL performance, why would you use volume? Why would you even care?