Consumer Focused Health

Quality is a buzzword in medicine today–unfortunately, its application is anything but a high-quality, well-designed approach.

There are many reasons why quality is hard, and many reasons why defining quality well may not be fair to existing players who are trying hard to do the right thing. In settings like these, there is no one right approach that fits everyone. Ironically, by having quality efforts led by the government (P4P) and major payers, we’re less likely to get it right than by having a number of smaller players seek to play objective, third party roles that maximize benefits for specific constituents.

Structurally, quality is hard because:

1. There are no established benchmarks
2. The only metrics that matter–morbidity, mortality, and patient experience– are complicated and seem largely uncontrollable
3. There is significant money at stake for the “chosen”
4. Politically, no doctor believes they are inferior
5. Society favors heroic intervention over statistically avoided event

I’ll deal with the first 2 below:

Benchmarks
The medical profession likes to think that all of its practicioners are better than average (of course a statistical anomaly). While licensing and training make all quite good, some will be better than others– and recent evidence shows there is a strong correlation to better quality with a minimum volume of procedures done, by hospitals and by surgeon. There is also strong evidence that physicians differ significantly in rates of treatments recommended, with few added benefits to societal health. Where benchmarks have been applied and evidence of high and low performers made transparent (not the case in recent hospital mortality rankings), low performers quickly adjusted to close the gap with high performers (see NY state cardiac surgery rankings)

Metrics
There are two types of metrics, process and outcome metrics. Process metrics are seen as means of getting to desired results, with outcome metrics being the desired results. Outcome metrics are the things that really matter– in medicine this would be rates of morbidity and mortality, with good patient experience also potentially a desired outcome.

The issue we have with quality reporting today is that most of the metrics available focus on process: did they use an EMR, did they give pill X in Y time, did they check body part A for B sign. While process is important, process metrics often forget about context and they forget that there are many roads to the end-goal. Unfortunately, process can only incorporate disease features we think we understand– and as with drug-eluting stents, we may be maximizing to solve intermediate steps but creating a worse end outcome.

Outcome metrics, while politically less viable (everyone wants a reason why they’re not #1–and will complain they need to stop seeing sick patients), are truly measures ensuring that treatments work as well as promised. Risk adjustment will never be perfect and segmentation will always be a rough estimation. This is where a single approach linked to payment will get tremendous pushback, whereas a few third party approaches (e.g., Morningstar for financial products) can be triangulated to determine appropriate risks and risk-payments. At the same time, spending a ton of resources on sick people who die soon is ultimately not a great use of society’s funds if its not benefiting research– or coming out of that individual’s own wallet (yes, its not fair to the poor–and nothing else in life is fair to the poor either).

Outcome metrics also seem most out of the hands of providers. However, in the broader sense, providers have done a poor job of managing longer term adherence to therapy and some outcome metrics (blood pressure, sugar levels) can reflect those changes– and have been linked to significant improvement in outcomes. Perhaps the issue isn’t that physicians aren’t able to control those metrics, but that we’ve been asking them to perform instant miracles instead of helping patients understand and control their disease over time.