On May 1, 2020, the Duke Margolis Center for Health Policy published a whitepaper titled Data Interoperability and Exchange to Support COVID-19 Containment. In it, the primary authors, Farzad Mostashari and Mark McClellan, lay out three strategies to help guide policymakers around the globe as they address the COVID-19 pandemic:
- Improve Commercial Lab Reporting
- Supplement Case Investigations with Clinical Data
- Enhance the Use of the National Syndromic Surveillance Program (NSSP)
Dr. Mostashari is a former National Coordinator for Health IT and previously served in the Epidemic Intelligence Service of the Centers for Disease Control and Prevention (CDC). He is now the CEO of Aledade, Inc., a company that supports independent primary care practices in value-based care.
Dr. McClellan served as the Commissioner of the Food and Drug Administration (FDA) from 2002-04, and as the Administrator of the Centers for Medicare and Medicaid Services (CMS) from 2004-06. He now directs the Duke Margolis Center for Health Policy.
As the Chief of Product and Strategy for a company that builds interoperability platforms used by healthcare organizations around the world, including labs and public health agencies, I believe Dr. Mostashari and Dr. McClellan’s work deserves discussion. Here, I’ll briefly cover each strategy they recommend in their report, as well as share my perspective.
Improve Commercial Lab Reporting
Most reportable disease notifications to public health are sent by laboratories, but as much as 50 percent of the time, the reports are missing key data, including basic patient demographics, which prevents an investigation from starting until data is manually tracked down.
The report suggests providing incentives for providers to submit complete data through two methods.
The first is by requiring basic information to be asked during order entry, so that the EHR or order–entry system collecting the order will force complete data before accepting the order. While this seems like a good approach, the reality is that sometimes samples are collected in the absence of complete data, and making certain fields required will just result in bad data being entered as a way to expediently complete the workflow.
The second, more scalable solution involves allowing labs to use APIs to query providers for the patient’s race/ethnicity, phone number, and address, using their MRN + DOB or payer + subscriber number — either directly to provider EMRs or through networks like CommonWell Health Alliance, payer clearinghouses, or Health Information Exchanges (HIEs).
Lyniate’s lab customers have been articulating this pain point for years, and I know they would welcome a solution to the problem. As McClellan and Mostashari point out, now might be the time to close this gap by pulling a fairly simple policy lever.
The CMS could provide financial incentives for laboratories and device manufacturers to ensure complete, accurate data. CMS could extend these incentives to include prompt notifications of reportable results to appropriate public health authorities.
Supplement Case Investigations with Clinical Data
Providers are permitted to disclose personal health information (PHI) to public health authorities under HIPAA, and, in fact, they are typically required by state law to notify public health departments of certain reportable conditions.
However, the ability and requirement to share PHI for public health operations typically does not extend to the providers’ business associates. Oftentimes, HIEs are business associates and therefore cannot serve as authoritative sources of patient information for public health departments, even though they frequently hold the necessary patient data.
For the COVID situation, the U.S. Department of Health and Human Services has said it will not impose penalties for the Business Associates Agreement to the public health use case. McClellan and Mostashari recommend that public health should assert authority to gain access to this data through HIEs.
Trust frameworks like CommonWell and CareEquality are geared toward provider treatment of patients. The proposal is to treat public health departments the same as providers to improve the data flow.
The recommendation that Mostashari and McClellan put forth in their report is for portal-based access for public health to information exchanges, “since most health departments have strained or limited ability or resources to establish and ingest new API-based data feeds.”
Lyniate’s position is that a portal-based approach is an expedient stopgap, but most public health departments should be able to achieve API integration with HIEs fairly rapidly.
Enhance Use of the National Syndromic Surveillance Program (NSSP)
The NSSP is a CDC project that collects daily data from thousands of facilities across the country. NSSP collects and collates more than 4 million admission, discharge, and transfer (ADT) messages per day, providing a good system for detecting outbreaks and patterns of emergency department trends. In spite of this scale, the data is difficult to use at a nationwide level due to differences in data sharing agreements among the states. In fact, coronavirus data is reported in systems outside of NSSP.
McClellan and Mostashari recommend undertaking the hard work of reconciling the state and federal concerns around the use of data within the NSSP. There is a balancing act between states having control over their own data and pooling it at a national level to react to outbreaks, and they outline a framework for taking both concerns into consideration.
They also recommend a number of technical changes to the Syndromic Surveillance that will improve the accuracy and consistency of reported data. Depending on how these changes are implemented, it could trigger changes to the Syndromic Surveillance messaging standard, which would need to be rolled out to all vendors who have certified against the existing standard as part of the Meaningful Use/Promoting Interoperability program. While the technical recommendations are worth pursuing, they may take time to roll out uniformly across the entire healthcare system.
Enhancements to NSSP will allow a single system to be used to track multiple different types of healthcare events now and in the future. Using a single, consistent system will improve the country’s epidemiological capacity.
High-Value, Low-Cost Improvements
This paper’s emphasis is on making high-value, low-cost improvements to the existing interoperability environment, rather than on investing time we don’t have to build new solutions. These seem like small, incremental improvements, but they will repair some significant problems that limit our ability to share and use data to combat pandemics.
Taking these actions now will help get us through the current crisis and prepare us better for the next one. Lyniate supports the recommendations in this paper, and we look forward to working with our customers to implement them to advance the state of the art.
For further reading: