You may have seen (or at least heard about) the recent article published by researchers at Stanford University. The article, published in the Annals of Internal Medicine, concludes that current usage of electronic heath records and clinical decision support technology may improve administrative efficiency, but does “not appear to translate to better outpatient quality of care”. (You can read more here). Your first thoughts may have been, “Wow, that doesn’t bode well for the ICT industry! How can that be”? But before you bail out of your job or start looking for work in another industry, let me share some of my own thoughts on what this really means.
I’ve been telling HealthBlog readers, and anyone else who will listen, pretty much the same thing for several years now. The electronic record all by itself, really doesn’t add all that much value. I’ve said it before, and I’ll say it again, “it’s what you do next that counts”.
Electronic health record systems are a necessary first step in transforming health and healthcare delivery. I hope no one uses the Stanford report as an excuse for not buying or delaying the implementation of an electronic record solution. We must get health information in digital, electronic format. Yes, I know that electronic records are a pain in the butt for most clinicians. Data entry remains a barrier to adoption. Compared to pen and paper or dictation, using an EHR, especially for clinicians who lack keyboard skills, does nothing but slow down productivity—at least initially. However, we can no more continue to use pen and paper in the practice of medicine than we can go back to the days of the horse and buggy. The time has come. America is already significantly behind most other developed countries in the digital transformation of paper records. But didn’t I just say that the researchers at Stanford are right; that electronic records don’t necessarily improve quality? Yes I did.
You see, the true improvements in healthcare cost, quality and access to be gained from digital health information don’t come from the EHR itself. They come from everything else that digital data enables. And, this is why the future for Microsoft in the health industry remains blindingly bright.
The opportunity to transform health and healthcare delivery depends on improving caregiver, patient and consumer communication and collaboration. It depends on more intuitive and powerful ways to aggregate, analyze and share data. It depends on the ability to surface pertinent information to whoever needs it at exactly the right time and place. It depends on new ways to deliver information and certain kinds of medical services beyond the hospital or clinic and right into the consumer’s own home. It depends on intelligent software that can help manage the complex relationships between all the players in the ecosystem of care. It depends on robust information worker tools because clinicians are information workers. It depends on solutions that extend to mobile devices or whatever devices the user happens to be using. It depends on lower cost, highly scalable, more manageable, and very flexible cloud-based solutions.
If you don’t see where I’m going, perhaps you had better change careers or go work for one of those big EHR companies. If you totally get what I’m saying, then put a smile on your face and go confidently out into the world. Tell your patients, partners and customers about the gigantic opportunity that is before those of us who work in healthcare, health and global ICT; an opportunity and a challenge that few companies besides Microsoft (and perhaps a few others on the globe) are even capable of taking on.
Bill Crounse, MD Senior Director, Worldwide Health Microsoft
I write this as someone who does BOTH clinical medicine and works with health care IT.
The current crop of EHRs that are out there do not provide meaningful clinical decision support. They provide "alerts" which most people blow past (much like we blow past the pop ups that occur on the websites we surf). Unfortunately, I feel that health care IT is a few steps back from the rest of computer technology.
When I can go on amazon and buy a TV and scrolling right below the TV are "98% of people who bought the TV also bought this HDMI cable and , this DVD player. Would you like to add to cart?" That's decision support.
When my bank emails or texts me when bills come in or my balance hits a threshold that's decision support.
When our EHRs see a K+ of 6.5 and say click this button to order Kayexalate, Lasix, albuterol, insulin+glucose, and a STAT EKG. That's decision support. We're not there yet.....
However, I think that studies like one cited in AIM are no reason to give up on EHRs. We should rather push our EHR vendors to make the technology work for us.
Thank you so much for your accurate and informative comments. You are spot on.
Bill Crounse, MD
Communication isn’t the only thing, it is everything. This is especially true in the health field, where life and death, sickness and health, can be contingent on communication or lack of it. Whether it is the emergency room of a hospital, or a simple patient encounter in the office, communication is paramount.
I'm on the same page. Thanks for your comment.
Bill Crounse, MD
Here is a vision of the future of healthcare communication technology:
1. New patient arrives into patient room
2. Patient has RFID bracelet which automatically associates him with every device in room, patient info is populated in every device database.
3. Nurse enters patient room
4. Nurse has RFID bracelet which automatically associates her with every device in room, giving authentication and authorization priviledges.
5. Nurse uses communication device (either badge or smartphone) to issue command (either voice recognition or utilizing smartphone touchscreen) to override television to access hospital network and validate patient information. Upon validation nurse issues validation command.
6. Nurse needs to hook patient up to infusion pump, there is no device located in room. Nurse issues command to locate nearest infusion pump that is not in use. Voice responds and data on television indicates a free infusion pump by room number, located two rooms down. Nurse either looks at TV screen or issues voice command to find out if this device has been sanitized recently. Also because this patient is an extremely critical patient, nurse issues command to inquire as to the last time this pump had preventative maintenance performed and if it has been broken anytime in the past 6 months. Once nurse is satisfied with this pump, she issues command for nurse call system to contact a technician to bring the pump to her.
7. During this process, nurse is alerted to a red or critical alarm going on with another one of her patients in another room, she immediately issues command to see patient monitor data and waveforms from that patient’s room on the TV in the room she is currently in. Upon realization, that this was a false positive alarm, she issues command to have alarm cancelled in patient monitor, she also issues command to have screenshot from the camera installed in the patient room transferred to the TV in the room she is in. The patient looks a little nervous based on the alarm episode, so she issues command to activate nurse call in that room and directly talks to the patient via video and audio on that patients television.
8. Meanwhile in the patient room next to this nurse, a disgruntled family member is who is techie, has decided to try and access the hospital backbone through a patient monitor, this patient monitor is windows based, and the hacker has had limited success. Immediately, the intelligent network switches have detected data flowing from this patient monitor that does not match its data footprint for that particular make and model of patient monitoring equipment. The intelligent switch controller immediately segregates this device to its own VLAN, and allows the normal data that it recognizes from the monitors data footprint, but disallows the extra data it is not familiar with. An alert is sent out three ways, one to the system analyst that indicates the situation and explains that this device has been segregated and will not be allowed back on its normal VLAN, until proper steps have been taken. Another alert is sent to the security department, and another alert is sent to the nurse. All three immediately issue voice commands to view the patient room in question, to see if there is someone physically interrogating the monitor.
Healthcare systems utilized in this scenario: Nurse Call Server and hardware, Alarm Integration server, VoIP or Cell Communication Server (Vocera or Voalte) and communication devices, Patient Monitor Database and devices, Biomedical Device Database, A/V Communications Server and TV/Monitor with access to both Guest network (patient) and Hospital network (staff), Hospital Wireless and Wired backbone including intelligent Layer 1, 2, and 3 network hardware, RFID Server and tags.
- All the technology and hardware listed here, already exists in the healthcare environment. What is needed is intelligent software that will not only extract data from different systems, but will also push commands to other systems. EMR integration is great, but it only addresses extraction of data, we need to address the interaction of hospital systems. Interoperability standards are great, but that will take too long for vendors to implement and may require replacement of existing hospital systems. We need software that extracts and infuses data in multiple healthcare systems. Come on Microsoft, hurry up :)
Minor Correction for typos - please delete prior post
PICU MD –
"(Most of ) The current crop of EHRs that are out there do not provide meaningful clinical decision support"
I would be pleased to demo a system does puts best practices in-line using an Orders Management System where tasks that post must be 'committed' in a timeline manner otherwise things start to escalate at executive dashboards.
Alerts do post in this system and yes, they can be bypassed (i.e the following fields are mandatory Continue?/Do not Continue?).
The only problem is background compliance checking will eventually trip up the user or others at key process points and post a roadblock, Yes, you can go to a gatekeeper to have the roadblock removed or invoke 'break glass' yourself and skip over the roadblock, but they/you will, over time, tire of writing up formal incident exception reports.
For years there were problems in medicine with rigid methodologies that could not accommodate the frequent need to deviate from 'best practices' through accommodation of ad hoc steps. However, things have advanced today to where users can perform any mix of unstructured/structured work they like. I follow the agency 'best practices', you ignore the best practices and hopefully get the same result making exclusive use of ad hoc steps or interventions.
All of this may be new to healthcare but in the area of aerospace engineering, the customers issue very exacting technical specifications and test protocols but given there are multiple suppliers each using different protocols/test procedures to arrive at the same end products or face rejection of parts/assemblies, hybrid methodologies such as Adaptive Case Management/Business Process Management needed to come along (2010) to accommodate real life situations such as found in healthcare..
The 'new' methodologies have relevance across multiple industries, particularly healthcare where complexity is probably an order of magnitude higher than what you find at a space station although engineers are getting better at fixing things without having to shut them down.
Anytime someone tells me they work on machines that are more complex than humans, I ask them how they would go about fixing one of their machines while it is running.
K Walter Keirstead, P. Eng.
If you read the study, the survey was taken based on patient visits in 2005-2007. That is like decades old compared to the evolution of features and functions inside EHR systems.
I certainly agree with the premise that just populating check boxes and making drop down selections do not in and of itself enable better patient care. It is the decision support, alerts, and meaningful trend analysis resulting from the digital entry which provides the greatest benefit.
There have been substantial improvements in these areas in many EHR systems since 2007. I encourage all those considering an EHR system to research the current state-of-the-art.