2022 Mid-Year Healthcare Data Breach Deep Dive — Protenus

In this post, we’ll use the United States Department of Health and Human Services (HHS) public breach tool as well as data compiled by DataBreaches.net (“DataBreaches”) for Protenus.

I hope you’ll all read the mid-year blog post. In one respect, though, it turns out that it doesn’t matter if you use HHS’s statistics or DataBreaches.net’s larger dataset: the number of incidents was down in the first half of 2022 compared to 2021, as was the number of records impacted. But while the percent of incidents attributable to hacking increased slightly in 2022 from 75% to 80%, the percent of incidents attributable to business associates decreased from 42% to 36%. Whether those smallish changes will be maintained by the end of they year or when more data becomes available remains to be seen.

But let me use this post to take a deeper look into HHS’s breach tool when it comes to business associate reports.

Unfortunately, and although Protenus gets it right, a number of other sites misreport data on business associates based on HHS’s public breach tool. The misreporting occurs if they confuse the number of incidents reported by a business associate with the number of business associate incidents. Their mistake is somewhat understandable if you understand that while HHS’s reporting form asks the reporting entity if a business associate was involved, the entity’s response does not appear on the breach tool unless you either manually expand each entry to see each entity’s response or export the database and then look at a field in the Excel sheet that asks Yes or No for Business Associate?

Some sites who try to report statistics based on HHS do not seem to realize that simply counting the number of incident reports submitted by “Business Associates” significantly underestimates the number of incidents that were attributable to  business associates. Thus, for the first half of 2021, Protenus reported:

44 reports were submitted by business associates, but a deeper dive into HHS’s tool reveals that 156 of all incidents in the data set (42%) involved business associates.

So one take-home message is this: the number of incidents submitted by business associates is not the same as the number of reports involving business associates.  If you want to know the latter, you will have to dig deeper into HHS’s breach tool. But even if researchers know to expand the breach tool or export it,  they immediately encounter a second issue: how to differentiate the number of reports that involved  business associates from the number of unique incidents involving business associates?

Some business associate incidents may be reported, in part, by the business associate on behalf of some covered entities while other clients of the business associate may file on their own behalf.  Determining how many unique incidents there are in any time period requires the researcher to know whether each report on the breach tool is due to a business associate breach that has already been counted in any incident counter or is a new and previously unreported incident.

Imagine, for  a moment, if all 650+ clients of one business associate each filed individual breach notifications with HHS. The numbers we’d get might totally distort our understanding of how many breaches involving business associates occurred that year.

HHS’s public breach tool does not permit the kind of analysis researchers need because even in its expanded form or exported form, HHS’s breach tool generally does not name the business associate involved in a breach reported by a provider/covered entity. Or if the business associate is the reporting entity, HHS’s note may not name the covered entity or entities affected. We generally cannot determine from HHS’s breach tool how many reports might be related to the same business associate incident.

In a more helpful world, HHS’s breach tool would enable us to look at business associate breaches in a framework that allows us to cluster associated reports.

In 2022, discriminating between the number of reports involving business associates and the number of unique incidents involving business associates will remain challenging because of the number of large incidents involving business associates that we have already learned about this year.