Collectively, people around the world are working together to stay home whenever possible. At the same time, many critical care departments at hospitals are overwhelmed, while other department staff covering elective surgery and non-emergent visits are reassigned or outright furloughed as part of social distance and quarantine measures.
It makes sense: the sickest people in the midst of a global pandemic are centered around healthcare facilities. Anybody potentially susceptible to infection or risk of death should steer clear of these locations.
The problem, though, is people managing chronic conditions—vulnerable populations—aren’t suddenly any less sick, or less needing of care. As a company that supports government-funded populations, we looked for a way to harness our expertise to assist our payer and provider partners provide continuity of care to their chronically ill patients. We decided to help providers identify and prioritize those patients who would be best served by telemedicine1 via a simple and quick report enabled via Lumanent Pre-Encounter Prep.
While every organization is unique, and different technologies can solve this in a myriad of ways, if you don’t work with Health Fidelity, you can still follow the same steps we did to generate your own telehealth suspecting tool.
1. Evaluate Your Inputs. The immediate and most actionable data available to us was claims data, showing us demography, confirmed conditions, and past behaviors. Look for source material that includes factors that could influence a patient’s ability to engage with telehealth (past use, any guardianship, any diagnoses that may inhibit technology use, like dementia or motor impairment), as well as any chronic conditions that create a higher susceptibility and contra-indication with COVID-19.
2. Layer in Parameters & Filters. Once you have amassed a useful dataset, you’ll need to whittle it down to those populations and patient conditions that are telehealth eligible according to the latest guidance from CMS. We focused on Medicare patients, and of the subset of telemedicine diagnosis codes included in the report, we suspect 97% are able to be addressed via telehealth. We also layered in additional ranking for patients who are due for their next annual wellness visit, frequency of visits over the past few years (as well as a population’s average) to find high need patients, or those that may be chronically avoidant of care.
3. Build the Logic. Our report logic is fairly straightforward, employing multiple binary algorithms to parse and then prioritize the aforementioned factors. Once the analysis is complete, a score is generated for each patient, which corresponds to high, medium, and low tier for ranking telehealth compatibility.
4. Evaluate the Output. Despite the shortened timetable, we ran the report more than a dozen times, slightly altering the logic and filters until we had a viable list of suspects. Our provider clients have indicated the attention to non-CMS related factors (social determinants, care access, etc.) have been very valuable additions to the list’s efficacy.
5. Disperse the Findings. We then documented and distributed the report to our providers for informing doctors, care managers, etc., those best equipped to address the patients’ conditions. By taking that last step, we’re taking additional data processing burdens off of hospitals and practices by allowing them to directly route to the service line department that needs to see who, when, and why.
How long telehealth remains a priority in a post COVID-19 world remains to be seen, but in addition to this suspecting report, we’ve also been working to make it as easy as possible to capture and code for telehealth encounters with Lumanent. Critically, we did so in a way that impacts Post-Encounter Review and Retrospective Review solutions, neither of which will place any additional burden on clinical staff at a point where their focus is acutely necessary on either treating patients with COVID-19, or handling the surge of non-infected patients that have gone without necessary care for several weeks. Because CMS continues to redefine and clarify the position of telehealth and risk adjustment under pandemic conditions, these updates to Lumanent will help organizations gather the data they need to adjust for telehealth encounters and presumed covered conditions, should they so choose.
NLP and the associated technologies that support its use are still relatively young. As today’s challenges continue to evolve, so will our commitment to solving problems that interfere with the sustainable delivery of care. If you are interested in obtaining a copy of the technical specifications for Health Fidelity’s Medicare Telehealth Outreach report parameters, send an email to email@example.com. We look forward to the opportunity to support your care efforts.
1 = Note: the terms “telehealth” and “telemedicine” are used synonymously, but both refer to HIPAA compliant, interactive audio and video telecommunications systems that permit real-time interactive communication.