Five questions for the patient-centered data home

HIEs are back and in a big way. Looking beyond interoperability, HIEs are now focused on creating nothing less than the Patient-Centered Data Home, accessible nationally. The timing is, paradoxically, good and bad. There’s no question that data silos continue to stall initiatives that could save money and lives. And there’s no question that EHRs have utterly failed to create such nationally accessible patient-centered data homes.

But events like the Facebook/Cambridge Analytica scandal have propelled data privacy to the forefront of media attention and public discussion. Before contributing to the Patient-Centered Data Home, HIES and other stakeholders must ask themselves some hard questions.

1. Who owns the data? In the fallout from the Facebook scandal, we at least know we can download the data that Facebook has been sharing and cancel our Facebook account if we wish. Will patients have the same ability with data collected in a national patient data home? Will patients have the authority to eliminate their records?

Interestingly, HIEs are one of the few entities in healthcare that must conform to patient opt in/opt out consent policies, which vary by state. These policies typically mandate the patient’s right to fully or partially opt out of having their data shared, but even so, there are exceptions. Some states may require that HIEs participate in public data sharing; presumably using all patient data, regardless of individual patient opt-in/opt-out status.

Beyond HIE requirements, the legality surrounding patient data ownership is hazy. CMS requires Medicare providers to retain patient records for a certain number of years, while various states have patient record retention laws in place.  Sharing de-identified data is also allowed in many instances; and while selling patient data is largely restricted, here again, there are some exceptions.

But bottom line, it’s the patient’s own data. Requests to delete or omit data should be handled respectfully and fulfilled in a way that answers the patient’s underlying wish for privacy and ownership.

Educating the patient on what he or she is legally entitled to, plus the data home’s privacy and security safeguards, are critical first steps. Following closely is educating the patient on the value of his or her data for individual and public health.

 2: Who owns data errors? A patient-centered data home will only be as good as its data, yet patient records are full of misspelled names, outdated addresses, wrong phone numbers and other errors that make it difficult to determine if it’s the right record for the right patient. There are also countless duplicate records, further adding to the confusion and data sharing risk.

All of the above is the normal state for records housed in a single system used by a single provider.  Now imagine how these problems will mushroom when patient records are coming from multiple providers in multiple states.

As noted during the patient matching track at the annual SHIEC conference for HIEs, up to 25 percent of patient records are incorrectly matched—a headache that HIEs inherit from health systems. Up to 50 percent of patient records are then subsequently mismatched during transfers. It was also cited that HIEs deal with unresolved patient record duplicate rates at 26 percent or even higher.

It cannot be overstated: unless HIEs know for sure that their own duplicate identity rate is minimal, and patient matching errors are extremely rare, they are putting patients at risk every time they submit data to the Patient-Centered Data Home or to state prescription drug monitoring programs.

3: How to match patients to the right records? As noted in a letter sent by several sitting U.S. senators to the Government Accountability Office, patient matching is an essential strategy to reduce risk of harm to patients. HIEs must have a means for assuring that patient data is going to the right master patient record. DIY solutions won’t cut it, and neither will conventional patient matching technology that processes patient identities in “Batches.” That’s too small scale for a national Patient-Centered Data Home.

4. What about using blockchain to solve the patient matching challenge? Blockchain can theoretically play a role here. As a distributed ledger, blockchain can update changes to a patient’s record. It can even assign a patient ID to the record. What it can’t do is assign a number that all systems exchanging patient data recognize as a patient’s unique identity—and that is the core challenge of patient matching today.

5. What is the best patient identity matching strategy in the absence of a national patient ID? Given what’s at stake, an advanced, tried and tested master patient index solution is the best bet for swiftly matching patient data with the correct patient identity.

Note that “advanced” should encompass the following criteria: the solution is paired with or includes a unique patient identity number or key—one that all systems recognize as the patient’s ID; enterprise scalability, as HIEs are dealing with millions of patient identities; real time matching (milliseconds matter in healthcare); and additional use of historical data when matching patient identities, including data that goes as far back as old college addresses and decades-old medical records.

With these criteria in place, the patient-centered data home can achieve the patient matching accuracy benchmark of 2 percent or less as recommended by AHIMA.

While these are not all the questions surrounding patient data integrity in the patient-centered data home, they are among the thorniest—and as such, should be addressed well in advance of creating this important repository for health information. Those HIEs that do so will establish themselves as trusted data stewards with patients and providers alike.

Photo: shylendrahoode, Getty Images

This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.