Director of Clinical Informatics in Neurology and Vascular Neurologist

Benjamin Kummer

Dr. Benjamin Kummer is the Director of Clinical Informatics in Neurology and a practicing vascular neurologist at the Icahn School of Medicine at Mount Sinai since 2018. He has experience building, implementing, and enhancing healthcare IT systems in neurological and non-neurological domains across challenging academic environments. He is triple board-certified in neurology, vascular neurology, and clinical informatics.

What are your roles and responsibilities as director of clinical informatics for neurology and vascular neurology at Mount Sinai Health System?

My focus is on the intersection between clinical informatics and neurology. By “clinical informatics” I mean the application of health information technology (HIT) to patient-level processes, which includes a wide range of areas including telehealth, analytics, electronic health record systems, interfaces, and so on. Overall, my time is split between operational informatics, clinical care, and informatics research. Within operational informatics, I am a central facilitator and HIT subject matter expert in conversations involving clinical and information technology stakeholders at my institution, which accelerates time to solution and helps both sides achieve their goals. I am responsible for lending subject matter expertise to development and execution of HIT-related projects, helping troubleshoot HIT system-related problems, advising on and vetting change proposals to HIT systems, and leading innovative asynchronous telehealth programs in neurology. In my clinical role, I treat hospitalized stroke patients from their time of arrival to the emergency department to their discharge to rehabilitation, and teach medical students, residents, and fellows. The remainder of my time is spent conducting research on the applications of information system tools to neurology, either through writing grants, articles, and/or mentoring medical students and residents on mutually interesting HIT-related projects in the neurosciences.

Have you observed any changes in the telehealth sector since the pandemic?

It’s undeniable that the pandemic pushed virtual care- but mainly synchronous video telehealth – onto the main stage of the healthcare sector. Synchronous video telehealth was still a novelty when I began working at Sinai in 2018, mainly because of CMS’ payment policies that limited coverage of video telehealth care at home. Those policies had done nothing for providers and/or administrators, many of whom were apprehensive about trying telehealth- and only a few people were interested in learning more about how to use video for their patients.

Despite the hesitation surrounding the technology, in early 2019 I launched a video visit pilot in Neurology – and was able to convince a handful of providers to join so we could begin experimenting with video visits, primarily with commercially insured individuals that had multiple sclerosis and headache disorders. The pilot was very helpful in allowing a rapid pivot when the COVID pandemic took hold about a year later and the institution decided to use the platform we had been piloting and made it the official institutional platform for video visits. The transition to virtual care was relatively seamless during COVID thanks to the fact that over the preceding year, we had already figured out the operating procedures for video visits that worked for our department and practices. Later in the pandemic, when video visits were running smoothly, I launched another telehealth pilot aimed at collecting more or less continuous data asynchronously from smart devices, in-between visits, to complement care decisions that were being made at each face-to-face episode.

What will be the impact of new technologies on the future of medicine?

In my opinion, the private sector is driving most of the HIT innovation and charting new territory. While several health institutions have very robust innovation programs, the vast majority of health care institutions that have the resources to innovate either orient themselves predominantly towards obtaining federal grant funds to support research programs, implement vendor software to address a specific need, or limit themselves by restrictions on how institutional patient data is used to develop commercially-oriented innovations. These things make it difficult to transfer many innovations that occur in institutions over to clinical use. 

New technologies will need to address interoperability – EHR systems and their data practically exist in silos right now. Although things exist like Epic CareEverywhere and HL7 FHIR, which is technically EHR-agnostic, I imagine the future will see technology evolve to allow EHR systems more and more efficient integration and in a more widespread fashion. I really think this is important to efficiency and accuracy during consultations and diagnosis as well as reducing cost throughout the healthcare system.

Along with COVID-enabled virtual care also came a big push to build out technology like remote patient monitoring (RPM) to support home-based acute care – in order to keep less severe patients out of the hospital, or monitor key clinical parameters for patients in-between ambulatory visits. The issues thus far limiting uptake in RPM have mainly been that providers are not totally comfortable with handling the increased amount of data these systems generate, and there is still some question regarding accuracy of measurements. I think we will see many HIT innovations enabling periodic remote vital sign or laboratory parameter measurements without requiring face-to-face, episodic encounters become mainstays in clinical practice –  by presenting this data to clinical decision makers in a digestible, relevant way that doesn’t overburden the clinician. I believe this trend will continue over the next several years, especially given that the US recently introduced a bill for the Telehealth Extension and Evaluation Act, which would extend Medicare reimbursement for several telehealth services for 2 years after the COVID-19 public health emergency ends in May.

In neurology specifically, I think the wearable revolution will be particularly revolutionary for lack of a better term. I think we’re going to see a lot of continuous monitoring of physical and physiological parameters with wearables integrated into smartphones and watches and rings, which I call passive detection, become more prevalent. Because so many neurological disorders have motor manifestations, we’re going to see these wearable physical activity monitors and sensors generate the “digital biomarkers” that may be surrogates for other clinically validated scores or assessment tools or measure a new aspect of neurological disease. We are going to see these become ready for prime time in the next several years. And this will be something new for neurology because our field still uses the neurological examination, which is un-quantified and serves as the “launchpad” for most of our clinical treatments. We don’t use lab values like endocrinologists or cardiologists do – but I believe we will see the rise of the “quantified neurological examination” in the years to come and that will change how we treat our patients.

What we are going to see in the next few years in my opinion is somewhat of a nationwide competition for patient bases between large healthcare systems and private companies and most of this is going to play out through virtual care models. We already see major tech players such as Amazon getting into the healthcare industry by positioning themselves as providers of access to healthcare. In this way, what’s innovative is not necessarily newfangled technology but rather a new business model for healthcare delivery, in which access to providers is made as easy as online retail or other app-based experiences.

I also think many clinical machine learning and artificial intelligence systems, which are still arguably in their infancy, will reach maturity and be incorporated or accepted into clinical practice to a much greater extent than they are today. This will also happen in the research realm – most clinically-oriented machine learning research is stuck in the development and proof of concept stage, but the evidence base will eventually include validation and operationalization. Our research team is working on a machine learning model to democratize access to cutting edge stroke diagnostics using low-cost systems and large amounts of data. This will hopefully be used at the point of care, eventually – but not before we conduct extensive validation, which we intend on studying formally as well.

What advice would you give your peers and colleagues regarding the use of virtual technology to optimize healthcare processes?

Every technology needs a champion. Providers that use virtual care can assist in propagating the benefits and educate peers on the use of technology to accelerate real-time patient consultations. Since the pandemic, we have in younger generations of providers gravitate more towards telehealth and virtual care more than the senior generation. We really need to better quantify the value of telehealth and virtual care systems in neurology and for the healthcare system as a whole. This is still a big open question for CMS and most healthcare systems, so I recommend anyone interested in these areas to try to find a way to measure the value. I think this will enable payors to continue paying for telehealth and virtual care. And it’s not just telehealth – any medical informatics intervention, be it clinical decision support tool, an order set, or a new integration – has to be framed in terms of value – so it can be articulated to the people who are paying for these systems.

Another thing that is imperative for the current generation of medical practitioners and students that are interested in health informatics and telehealth to consider is that most medical technology is primarily used by not very sick, wealthy,  well-educated individuals when it first comes out. We have to think more carefully about making helpful medical technology more easily accessible to needy populations, and less about creating newfangled technologies that target people who aren’t very sick and have a relatively less pressing needs for the technology. When it comes to the use of medical technology, there is always an element of unfairness in that access to technology tends to be correlated with measures of social well-being such as wealth, education, medical literacy, and medical comorbidity. As practitioners and informaticists it is one of the most important things we need to do to find ways of ensuring that high-quality care and the benefits associated with it are available to disadvantaged members of society.