HIStalk Interviews Todd Plesko, CEO, Extension Healthcare

Todd Plesko is CEO of Extension Healthcare of Fort Wayne, IN.

Tell me about yourself and the company.

My career began in the ambulatory space in the mid-90s. It was a very interesting time where CMS, Medicare, Medicaid, etc. had mandated that ambulatory care move from paper-based billing and scheduling, primarily billing, to electronic billing. That created a huge boom right around 1996 in the first wave of HIPAA for every ambulatory practice in the country to switch to an electronic practice management system. Then as we know, EMRs came 10 years later, with Meaningful Use and the Recovery Act.

Extension Healthcare is my third startup. We’re well past the startup stage now. We focus on acute care. At Extension Healthcare, we believe fully that the enemy is alarm fatigue.We believe that that enemy will be beat over the next 770-plus days as the Joint Commission focuses on solving that problem via their National Patient Safety Goal on alarm safety.

Today we’re just under 200 hospital clients. Right around 90 staff and four registered nurses. We’re poised to grow very, very quickly as the problem of alarm safety and alarm fatigue in particular becomes more and more relevant, becomes more and more of a discussion point, and of course with the Joint Commission focusing on eradicating this problem, or helping to solve it, with a National Patient Safety Goal on alarm safety.

The ECRI Institute also recently named alarm safety as its number one technology hazard. With all that attention, what’s been the response from the monitor manufacturers, the companies like yours, and the hospitals themselves?

There are only a handful companies that can solve the problem of alarm fatigue. In fact, that’s a very small amount. What’s important for your readers to understand is there’s a very distinct difference between an alarm and an alert. There are many companies out there focused on alerting, which is low priority — something that may not be clinical in nature and doesn’t require a response to that event. Alarming is very, very different. 

As it relates to the monitor manufacturers, some of the EHRs and other companies that are just on the outside or adjacent to the middleware space – which is a word traditionally used to describe what we do — we see some of them entering the market. But most of them are leveraging tried and true companies, like Extension Healthcare, to deliver those alarms and alerts and allow a knowledge worker — a nurse or a physician or someone else in the hospital clinical — to respond to those alerts. Event response is a very, very important topic for us. It’s something that’s not talked about a lot. But in our view, it’s at least as important as delivering an alert with context to the caregiver.

Most of the companies understand that middleware, alarm safety, alarm management, and event response is a business all its own. It’s taken us years and many, many millions of research and development dollars to get to the place where we support every major device on the market, every EMR on the market. Every input that you can imagine, we support. Every output you can imagine, we support as well, which is equally important.

As the world moves towards smartphones, specifically iOS and Android, what people often overlook is that the majority of devices in place at a hospital today continue to be voice over IP devices. We believe that the only way to effectively begin to solve the alarm fatigue problem is to recognize that most communication begins with an event, an alarm or an alert; recognize the fundamental difference between an alarm and an alert; be able to support hybrid environments from pagers to voice over IP phones to smartphones; and be able to work inside the four walls of the hospitals and outside as more and more workers begin to work outside of the hospital.

Those several statements alone are enough to deter a lot of would-be companies from entering the space because it’s just a daunting challenge, not to mention the regulatory environment. We are a Class II regulated medical device focusing on alarm safety in the alarm safety category. That’s a daunting challenge for any company and something that obviously we take very, very seriously.

Just to put the market in perspective, who are your top two competitors?

I don’t consider any company in the space to be competitors with what we do because we go about it a different way. In the traditional middleware space, Emergin and Connexall are probably our top two competitors. Two companies that I have a lot of respect for.

Emergin created the industry. They changed dramatically when Philips bought them. We have many, many — upwards of 20 — ex-Emergin staff with us now, something that I’m proud of. They bring with them tremendous knowledge.

The way we handle data is very, very different from anyone else on the market. We believe that context is king. Context means everything when it comes to solving the problem of alarm fatigue, truly solving it. 

The two companies I mentioned, I would consider first generation middleware companies. We consider ourselves the next generation of alarm safety and event response companies because of the way we handle data, because of inside the four walls, outside the four walls, and most importantly, because of the way we enable event response. That’s very, very important.

I had mentioned earlier that the most important thing, we believe, to solving the problem of alarm fatigue is delivering context with an alert.  We’re running a clinical study now where we’ll soon share with the community exactly what that number is — what percentage of clinical communication begins with some event, an alarm or an alert. We believe it’s very, very high. Soon we’ll have those data.

If you believe that, and you believe that context is king as we do, that means that the only way to truly solve the problem of alarm fatigue is to deliver the five Ws — the who, what, why, where, when — in the form of an alarm or an alert to the appropriate caregiver at the right time on the right device, whether that’s a smartphone or a voice over IP phone, and whether it’s inside the four walls or outside the four walls. Then the event response piece occurs. That event response today is predominantly secure text messaging. 

Those are the full components required to solve the problem of alarm fatigue. If you don’t have context, you are sending an unintelligent alert. If you are not sending the who, what, why, where, when, the user has to ask those questions. That’s just yet another interruption that contributes to the problem of alarm fatigue. That’s why we believe that those first generation companies or competitors are missing the boat on actually solving the problem. Evidence exists over the last five years that hospitals that have installed first generation alarm safety middleware have indeed contributed to the problem and not solved it. 

We’re taking a very, very different approach, which includes delivering context inside the four walls and outside and allowing event response via the form of voice or secure text messaging – point-to-point, point-to-group, etc., to truly finally solve that problem. It’s killing people, it’s costing a lot of money, and it’s a big dissatisfier for nurses and for physicians. 

We believe over the next 778 days, the time between now and the Joint Commission mandate, that the problem can mostly be solved by intelligent, contextual systems that allow for event response.

A lot of the work with alarm management seems mostly to be routing and prioritizing an excessive number of alarms or notifications that weren’t significant to begin with. Can monitors be made smarter so that they do more than just display information and make noise all the time?

That’s what our system does and that’s what other systems do. It’s not just our system that can solve that problem. To take data in, parse out what’s relevant and what’s not relevant, determine what’s actionable and what’s not actionable. That’s really a small sliver of the problem.

Imagine stripping out some of the data that the alarm is sending, the physiological monitor in this case. Stripping out what’s relevant and what’s not relevant, packaging what’s relevant with the other who, what, why, where, when. Typically that’s not coming from the monitor. That’s going to come from the EHR and from other systems like nurse call systems. Often the “who” comes from there.

That’s going to, in delivering an intelligent alert, someone who can be actionable with it. What happens today is a lot of those alerts go to someone who’s on break. The system is not intelligent enough to understand presence and whether someone is actually available, or whether that person can actually solve the problem or act on it. We don’t see the monitors doing that any time soon. That’s why we work closely with those companies and we’re proud to do it. That’s precisely the problem we solve. 

Most importantly, it’s just a fraction of the problem is getting that monitor alert to the right person at the right time. That’s a sliver of the problem. The bigger problem is context and how the user will interact with those data, something that we call event response.

Has anybody done statistics on how many of the alerts that go through your system or other systems are found clinically useful by the clinician?

There is a cacophony of bells and whistles going off in a hospital. Walk through one someday and it doesn’t take long to get a headache. You can imagine what those nurses do day in and day out, God bless them.

To my knowledge, the clinical studies, as it relates to alarm safety, are lacking. I’m really glad that you asked this question. One of the things that we’re doing with a new program that we call Extension Evaluate, a free service designed to collect those data for a hospital. Think black box recorder. We put Extension Evaluate in. Because of the way we handle data, it works out of the box. 

As opposed to sending alarms, triggering alarms, and communicating with endpoints, Extension Evaluate sits and listens. It listens for 30 days. And at the end of the 30-day period, our consulting group sits down with the hospital and shows them a very deep and illustrated picture of what’s happening with their alarming and alerting environment. Those data are incredibly valuable, especially spread over time. Nobody to my knowledge ever in the space has collected those data longitudinally over time and reported on them. From an academic, clinical study standpoint, that’s exactly what we intend to do with Extension Evaluate. 

We’re solving two problems. One is allowing hospitals for free, no risk, to get a very good and deep picture of their alarming environment. Then of course a gap analysis between where they are today and what they need to do to be compliant with the new Joint Commission mandate. But also building a compendium; building a library of data that can be used and regressed to answer the question you just asked. To answer how many alarms are actionable. How many alarms beget a clinical communication. We believe that number’s incredibly high.

That’s another clinical study that’s currently underway. If you have to communicate with someone as a nurse or a physician, how often does that begin with an event, an alarm, or an alert? We believe the number is very, very high, well into 80-90 percent. Soon we’ll have that exact number. That’s something that we’re very excited about — contributing to the academic community on true statistics taken from real-life hospitals longitudinally over time.

Nurses are on the hook to not only set up and adjust the monitors, but respond to the messages they issue. Are problems caused by nurses not having the time or knowledge to perform as monitor maintenance techs?

While some of that may be true, we would never, ever blame the nurse. Our view at Extension Healthcare is truly the nurse and the physician are the most important knowledge workers in the country. Nurses in particular have an incredibly challenging job, maybe the most challenging of any job in America. I believe it’s incumbent upon companies like us and the monitoring companies, perhaps biomed and IT, to design clinical workflows that truly contribute to solving the problem.

That is where a lot of the first generation alarm safety middleware companies have not spent enough time – pausing to evaluate what is truly causing the problem and what’s contributing to it. It’s very easy to send out, for instance, a Code Blue alert to a code blue team when someone is in asystole. It’s easy to send that via a phone or a pager or overhead. What’s not easy is to do it in a silent way and allow the first responder to respond in a silent way,and inform everyone else on that team who’s in different areas of the hospital, perhaps even outside the hospital, of exactly what’s going on — the who, what, why, where, when. That is what we call event response and probably the most important thing. 

For me, for us at Extension Healthcare, it’s about educating and informing the nursing community about which workflows make sense and which ones don’t. Because a lot of the time, tried and true methods that are in place today are actually contributing to the problem and not solving the problem of alarm fatigue.

Do you have any final thoughts?

The future is very important. Our space is dynamic. It continues to evolve. Data handling will become more and more contextual. Alarm management systems will continue to become much more advanced in terms of rules engines and complex rules processing. Clinical algorithms will become part of the system. All of this will advance patient safety and complexity even further. 

It’s very, very important to take into account not only where we’re at today in lessons learned from the past, but also where the industry’s going. Not only in terms of which device a nurse will use, but which data to deliver to a nurse or a physician, the context, and how they’ll interact with that. Not only now, but in the future, to drive down this evil, evil problem of alarm fatigue.