This blog post is the first in an ongoing insights series from our data team exploring various trends within our business and healthcare more broadly.

One of the challenges that we run into at Oscar is trying to find good benchmarks and comparison groups for the Oscar population so we can attempt to understand how and why we might differ from the “norm” (and accordingly what we might be doing well or poorly). This is not always easy, but even when we can find publicly available data about the US population, we discover that “the general population” is not actually the right comparison.

Below we’ve compared the distribution of maternal age for Oscar vs. the national average:

You can see that Oscar has many fewer young mothers than one would expect given national averages. The biggest reason for this is that young mothers are much more likely to have their insurance provided by Medicaid than any other source. If you only compare Oscar to other private insurers, we are closer to average:

Note that we still skew older; this could be Oscar-specific or could be more generally true of exchange insurers; it’s much harder to find these sorts of data for exchange populations.

These sorts of adjustments don’t always leap off the page, but because the exchange population (and Oscar’s population on top of that) is neither randomly-selected nor representative of the national population, it’s extremely important for us to try to quantify that as we analyze our population.

For instance, this observation came up in the context of research the Data team completed surrounding the establishment of a maternity-oriented program with our Population Health team. Understanding the age skew helps Oscar do a better job of determining how to approach our maternity program — in this case, recognizing that we rely on simple comparisons of common clinical metrics (like C-section rate) to understand our challenges, and that whatever work we do should be particularly attuned to the health needs and concerns of mothers over 35.



Source link

By