Joyoti Goswami

Predictive analytics in EHRs aren't yet effective enough for clinical decision support at the point of care.

While understanding and reconciling drugs after discharge from the hospital can be challenging, it's a necessity for greater efficacy of care delivery. HL7's Da Vinci project can help.

In the COVID-19 era, health systems recognize that existing data infrastructure is inadequate. Here are three things large datasets need to be useful.

Technology has proved to be a blessing for consumers who today have a plethora of care options and improved access to care. This is not the case with frontline providers who continue to suffer from burnout and fatigue. 

While traditionally deeply skeptical of artificial intelligence in clinical settings, in today's fast-changing care delivery landscape many physicians are thinking more proactively about how AI can improve quality and patient experience.

The right technology applications can turn the COVID-19 vaccination program into a game-changer for the healthcare industry.

CIOs need to look beyond just EHRs and explore stand-alone platforms to enhance care delivery – keeping focused on reliable patient data and streamlined clinical workflows.

At Geisinger in Pennsylvania, a pilot program to bring care to the homes of older patients with complex healthcare needs has shown a 35% reduction in visits to the emergency department visits, a 40% drop in hospital admissions and an average annual savings of nearly $8,000 per patient. The future will see chronic disease being managed more from homes and physicians' offices than in a hospital setting. Chronically ill patients may not always require visiting the hospital frequently if their follow-ups and routine checkups can be managed remotely. Especially in times of a pandemic, they are better off staying at home.  However, that should not affect their care. Given today’s technological advancement in healthcare, most follow-ups can be well managed remotely. While nothing can beat an in-person experience, physicians can check in on them digitally, through remote patient monitoring tools or video consults. Technology is rapidly changing the bedside examination. Video visits, infrared temperature sensors, mobile health, and wearables replace the need for clinic visits. While remote patient monitoring and telehealth have been on the radar for a few years now, defined workflows and processes around it are continually evolving to improve the care delivery experience. The pandemic has only accelerated this process and the need to innovate better infrastructure and device investments.   Infrastructure to manage chronic diseases remotely  Technology components that can help better organize chronic care management remotely include: Devices and Sensors: Medical devices and sensors help a physician measure the vitals and other patient parameters. For example, as a patient walks into the clinic, a physician notices his gait. The technology equivalent of this can be gait recognition sensors using mobile accelerometers. Similarly, sensors for measuring blood pressure, temperature, glucometers, pulse oximeters, and respiratory rate sensors, all integrated with mobile applications, can be used to monitor patients remotely. Many healthcare facilities and smaller clinics offer remote monitoring devices or kits, as per a patient’s need. Platforms: Platforms are required to aggregate data and facilitate communication between patients, physicians, and caregivers. These platforms are HIPAA compliant and offer integration with devices and EMRs. Integration with EMR: While EMRs are part of the technology infrastructure in any health system, integrating wearable data from devices into the EMR can help physicians maintain data continuity with the initial in-person visit recorded in the EMR. Facilities and providers can choose to reduce the noise created by large volumes of data from personal devices by requesting data at specified intervals only. With EMRs offering healthcare-standard APIs (such as FHIR) for integration, patients need to select devices with integration capabilities using healthcare standards. Data analytics for risk stratification and decision support: Devices and sensors can generate vast amounts of data stored in big data repositories. Algorithms on top of this data can help in risk stratification, prediction, and improving outreach. AI techniques like machine learning, cognitive computing, and deep learning can play a critical role in identifying chronic diseases using predictive modeling techniques. Preferred functionalities to manage remote care Type 2 diabetes, coronary artery disease, atrial fibrillation, chronic obstructive pulmonary disease, congestive heart failure, stroke, and chronic wounds are seven diseases where technology has been used to manage care for patients with promising results. These diseases impact at least half of the adult population globally and account for over 80% healthcare costs. Current care delivery for these chronic diseases is reactive. For proactive management of care, patient data at regular intervals is needed to provide real-time feedback and motivate patients to adapt to healthy behavioral changes. Chronic disease management needs a holistic approach focused on lifestyle changes and behavior modifications. It uses technology to gather data from wearables and adds AI and analytics to drive outcomes. While there are not many stand-alone solutions specifically  for managing chronic diseases,  it is important to make sure that these technology solutions or applications have the following functionalities: The ability to group patients: Data from an EMR can help group patients by disease conditions, risk factors, comorbidities, age, and active medications. Such groupings help provide valuable information for sending alerts, reminders, and prescription refills to patients suffering from chronic diseases. The ability to engage patients: Often, chronic diseases are due to health risk behaviors such as physical inactivity, incorrect nutrition, tobacco use, and excess alcohol consumption. Engaging patients throughout their healthcare journey can help achieve desired outcomes in chronic disease management. "Tell me and I forget, teach me and I may remember, involve me and I learn" (Benjamin Franklin). Improvement in behavior patterns can ensure continuous reminders, alerts, and engaging patients to track behavior patterns, diet, and lifestyle goals to improve health. A two-way communication flow from the patient to the care provider team and vice versa should be a necessary feature. The ability to integrate remote patient monitoring devices: IoT devices and wearables integrated with the EMR workflows can allow caregivers to provide advice at the point of care. Vital data points captured with wearables, smartphone apps, and home monitoring devices are used in between follow up visits to provide interim care and advice in chronic disease management.  The ability to aggregate data from other hospital information systems: These could be data aggregated from multiple sources, including labs, pharmacy, sensors, and valuable clinical-decision support (CDS) to track patient information to manage chronic diseases proactively. Population health management experts recommend closing gaps in care in high-risk populations for better health outcomes. Mobile applications, customer relationship management applications, patient portals, and patient education platforms are part of the technology infrastructure to engage patients. Integrating technology to improve access, affordability The CDC runs several structured intervention programs to manage chronic diseases. The National Diabetes Prevention Program and Chronic Disease Self-Management Program (CDSMP) is one such program with the CDC's necessary resources. The National Diabetes Prevention Program, for example, provides good economic value and could save an estimated $1,146 per participant for in-person classes and $618 for online classes over five years. Many of these programs are now being run online using video visits for interaction and wearable devices to capture vital data. Successful programs like Livongo offer connected glucometers, support from health coaches, and unlimited strips for diabetes management. Livongo’s success in the chronic disease management space can be attributed to its B2B partners, including employers, insurances, associations, and other providers. Integrating technology for chronic disease management needs planning for functionality, processes, and leadership policies in place for it to be successful and improve overall health outcomes. As more and more wearable devices are coming into the market, planning chronic devices' management is getting more affordable and convenient. To summarize, a technology solution for chronic disease management should include these capabilities: The ability to ientify the population eligible for intervention Communication with multidisciplinary teams that may include physicians, pharmacists, nurses, dieticians, and psychologists Algorithms for risk identification and stratification Patient education platforms Video visits Outcomes evaluation Tracking and monitoring the program Security and Compliance As healthcare systems move towards value-based care, creating the shift towards remote monitoring of chronically ill needs to happen. Dr. Joyoti Goswami is a Principal Consultant at Damo Consulting.

The clinical manifestations of COVID-19 are varied, and patients are known to have rapidly changing signs and symptoms that must be tracked with laboratory testing. A patient may start his treatment journey with his primary care physician and will include lab centers, diagnostic centers, inpatient, and home quarantine centers.  It is crucial for the respective laboratory information or electronic health record systems to share the lab tests and diagnostic information with each other. Communicating the lab orders and results across multiple care settings and with treating physicians, quickly and effectively, is the need of the hour.   A patient encounter creates observational data such as vital signs, symptoms, diagnoses, and diagnostic information, along with prescription details. The observational data findings, and laboratory and diagnostic information, along with medical knowledge and clinical guidelines, forms part of a clinical database. The data here can be turned into information only when systems use the right terminology standards.  Terminologies are the heart and soul of a system  Terminology services help two health systems talk to each other and make the information stored within them, useful to both patients and practitioners across the care journey. Terminologies are the language in which computer systems can communicate physicians’ notes and observations with each other.  The HL7 2.x or FHIR API standards are the messengers that help deliver the language (read information) from one system to another. A simple example is the way different clinicians would write COVID-19 (is it COVID-19 or COVID 19 or just COVID?). As is always said, technology must adapt to its end users, and for a software to quickly understand and transmit information across systems terminologies packaged within HL7 2.x or FHIR standards are used. Laboratory testing is one of the significant parameters to track a pandemic like COVID. Public health management of the COVID-19 response needs complete laboratory testing data, including standardized test results and demographic details for contact tracing, mitigation, and control of the spread of the disease. One of the critical roles of technology is the usage of standard terminologies used by clinicians and laboratories while communicating patient information across different care settings and specialties. Two of the terminology standards used:  LOINC codes are used for the orders placed or the question that is asked of a viral test specimen, e.g. SARS-Coronavirus-2 RNA [Presence] in respiratory specimen by NAA with probe detection. The SNOMED-CT codes are used for the results or diagnostic answers, e.g. positive, negative, invalid, or specimen is unsatisfactory for evaluation. Specimen types, e.g., the serum, plasma, and nasopharyngeal swabs, are coded with SNOMED CT. The CARES Act mandates the use of de-identified data elements that must be reported for each test to the public health authorities, and these are available on the HHS site. The Regenstrief Institute has released new LOINC codes in the wake of the pandemic.  To increase semantic interoperability for laboratory reporting for detection of SARS-Coronavirus-2, the FDA, CDC, IICC, the Regenstrief Institute, and the Association of Public Health Laboratories (APHL) have developed a LOINC map for all SARS-Coronavirus-2 diagnostic tests listed on FDA Emergency Use Authorizations.  The CDC and FDA maintain the mapping of all the currently approved SARS-Coronavirus-2 in vitro diagnostic labs and their corresponding specimen types and results under the stewardship of the Office of the National Coordinator (ONC).  The ‘Observation’ resource in HL7 FHIR Real-world concepts in the healthcare system are linked together by FHIR packages called Resources. A Resource in FHIR is an entity with a URL by which it can be addressed and contains a set of structured data items as defined in the resource.  The ‘Observation’ resource in FHIR includes elements where the different components of terminology, i.e., the code, its description, values, units, and interpretation of the result, are contained within a resource or message. Systems transmit this resource or message to exchange information. Each resource includes elements where the terminologies and clinical concepts are included. This package, when exchanged across healthcare systems, helps to create interoperable information for the consumption of clinicians associated with the management of the patient. The FHIR framework includes four elements that have clinical terminologies: System: This is the URL that identifies the coding system. E.g. (www.loinc.org, http://snomed.org/) Version: The current version of the coding standard. E.g., LOINC V 2.68, SNOMED September 2019 Release Code: This is a unique identifier for the concept as defined by the code system Display: A description of the concept as defined by the code system Typically for lab tests, the ‘Observation’ resource within FHIR is used to convey the orders and results relevant to laboratory tests. The ‘Code’ and the ‘Display’ elements within the ‘Observation’ resource include the clinical terms for exchanging information.  In HL7 V 2.x data transmission, the Observation Request Segment (OBR) and Observation Result Segment (OBX) contain the laboratory tests and their corresponding results in lab test codes of LOINC or SNOMED CT. The LOINC and SNOMED CT codes are updated regularly, recently a set of new terms and codes for SARS coronavirus 2 have been released. Improve decision making and reporting of patient data across the care continuum In the scenario where a patient with COVID-19 moves from primary care to emergency care, quick decisions and tracking of laboratory results is necessary. The use of terminology standards can help reduce the time spent in exchanging, tracking, and reporting these tests across the patient journey.  This is a one-time activity that laboratory systems need to undertake to define all the tests in their catalogue correctly. Hospital systems will need to update their EHR systems with the updated orders for the COVID specific tests that they are ordering in collaboration with the referring labs.  This is a foundational activity not only towards data interoperability but also towards research with real-world evidence for the management of COVID. As Dr. Atul Gawande said in The New Yorker recently, the technologies are in place – it's the implementation that needs to be streamlined. Joyoti Goswami is a Principal Consultant at Damo Consulting, a growth strategy and digital transformation advisory firm that works with healthcare enterprises and global technology companies; she is a physician with varied experience in clinical practice, pharma consulting and healthcare IT.

How healthcare providers can leverage interoperability across the care continuum for improved care coordination and patient empowerment.