He's co-chairman of a convened by the National eHealth Collaborative that's studying current best practices and flow rates of patient-generated health data. The Patient Generated Health Data Technical Expert Panel to a workgroup on patient empowerment of the federally chartered Health IT Policy Committee.
What's the difference between the information gleaned from a patient during an office visit and this new order of data flow? Wald explains that these new types of information that could be coming from the patient at home. The data could be pre-visit, they could be a series of blood pressure readings the patient has been asked to provide, or they could be tracking device data.
How common is it for healthcare organizations to receive electronic data from patients? “It's a lot more common than anybody thinks,” Wald said. “There are practices and little projects and experimentations going on all over the country all of the time.”
In its presentation to the feds July 18, Wald's panel of experts came up with a list of best practices for patient-generated data handling. It recommended that organizations establish clear policies and procedures, and then start small with a pilot project. Pick a group of patients—those with a common ailment and who are prior users of any existing online services—to test the concept, they said.
One policy example the panel gave was that providers should, as part of their data-handling practices, set an acceptable response time within which a patient will be sent a message that their information has been received. The panel also suggested that marketing of an organization's capabilities to handle patient-generated data is essential, and must be ongoing, and that marketing plan must include progress measures, the panel said.
Staffers working with patient-generated information must be properly trained to send and receive information, as well as how that new data fit into the workflow, the panel said. Training resources need to be made available to patients, providers and staff. For example, staffers reviewing incoming patient messages should be trained to triage the information, determining whether the data is or is not within a normal range. Reviewers should be guided by protocols.
The technology must provide for secure, encrypted messaging capabilities between patients and providers through an electronic health record system, patient portal or mobile applications, the group said.
The technology also needs to be equipped with notification functions on both ends, it said. For example, a message might be sent to a patient that their physician will be away for an extended period of time. Or it might be sent to providers, that a new message from a patient is in, or an action is required.
In the future, as the volume of patient-generated information increases and as the technology advances, automation may play an important role in assisting providers handle the flow through real-time analytics, such as identification of data points that are outside a normal range.
Questions of legal liability occurred throughout the panel's discussions about these new technologies pumping streams of data at providers, Wald said. “The concern was palpable,” he said. But “people who approach patient-generated health data thoughtfully and with some planning, can substantially reduce concerns about liability and burden.”
Wald said the aim of panel was to inform the federal policy committee work group of best practices, to inform the full panel, the ONC, and eventually, the CMS, in their deliberations on Stage 3 meaningful-use criteria for EHRs.
Follow Joseph Conn on Twitter: