Friday, October 12, 2012

The Naiveté of mHealth

Last week I attended a seminar on mHealth sponsored by the Technology Association of Georgia (TAG).  The presenter was Arthur Lane, Director of Mobile Health Solutions at Verizon Wireless.  He gave a nice presentation and video of a system Verizon is designing to improve care of congestive heart failure (CHF) patients after hospital discharge.  CHF patients are treated effectively in the hospital setting with closely monitored vital signs and carefully administered medications / diet.  The problem is that once the patient goes home it is difficult to maintain the same level of monitoring and precision of the medication / diet regimen.  As a result re-admission rates for CHF are high, adding to the cost of care.

The Verizon system claims to correct this problem with smart phone technology.  The video showed a smart phone reminding a CHF patient to weigh himself before bed.  He has gained ½ pound since the morning.  When he wakes up the next morning the phone again reminds him to weigh himself.  He has gained another pound.  Weight gain day-to-day is an indication that CHF is getting worse.  The phone sends the weight data to a server, which in turn notifies a provider to call the patient and somehow prevent him from getting worse and showing up in the ER.   It was never clear to me how the provider was going to fix worsening CHF over the phone.

After Mr. Lane completed his presentation he joined 3 other panelists for a lively discussion moderated by a local physician whom I know.  Some of these panelists described their devotion to mHealth with near breathless excitement.  The physician moderator posed the ever-present question to the panel:  “How do we get doctors interested in this system (and mHealth overall)?”  The answers ranged from good – “Give doctors a product that is cost-effective” – to the ridiculous – “Align incentives by making physicians join ACOs.”  The silliest thought of the night was the suggestion from one panelist that health care is no different from banking.  I left the meeting with some concerns about who would pay for the Verizon system but decided to hold my reaction until I did a literature review.  After all, I am no cardiologist and have not treated a patient for CHF since med school.

My review did not yield good news for Verizon or mHealth.

Turns out physicians have been working on home monitoring for CHF patients for years.  Unfortunately their studies to not support remote home monitoring for CHF to reduce hospital admissions.  A study from Yale Medical School published in the New England Journal of Medicine in 2011 randomized over 1600 CHF patients to either a control group or a remote monitoring group for outpatient care following admission for CHF.  There were no differences in readmission rates for CHF or for any other cause over the 6-month study.  Several other studies, including comprehensive reviews of existing literature, reach similar conclusions.

So what would a more realistic mHealth video look like?

Our CHF patient is discharged from the hospital all tuned up with appropriate medications, diet and smart phone remote monitoring using a CHF app.  The monitoring app works well at first, feeding him periodic words of encouragement and reminders to take his meds, record his vital signs, weigh himself, etc.  After several days of his phone going off constantly with all the reminders, alert fatigue sets in.   After ignoring the alarms for a few days he gets fed up and shuts the CHF application off.  The monitoring network detects the data interruption, and a provider calls the patient.  At first the contact with a real human helps, but after several calls alert fatigue strikes again.  Our patient recognizes the caller ID and stops answering.

In the meantime he tires of his medication regimen and diet restrictions and succumbs to the urge to scarf down some pizza and beer with some potato chips for dessert.  His smart phone isn’t smart enough to change his behavior.  The salt and fluid load makes his heart failure worse.  In the middle of the night he wakes up short of breath and calls 911.  Back to the hospital he goes.

The mHealth community is so enamored with their toys they can’t see what is right in front of them:
1.     Peer-reviewed medical literature does not support the use of home monitoring for CHF patients.  Period.  LTE smart phones and glitzy medical apps do nothing to change that.
2.     Without supporting literature no one is going to pay for remote monitoring. 
Who is going to cough up the dough for all those smart phones, Bluetooth connected home monitoring devices, remote servers, and the army of providers that will be required to manage the terabytes of data that such a monitoring network would generate?  Neither ACOs nor any other ill-conceived “alignment of incentives” for physicians solve this issue.
3.     The mHealth folks fail to recognize that monitoring is not the endpoint.  The endpoint is changing patient behavior.  A smart phone constantly shrieking warnings and reminders is rendered useless by alert fatigue.  Patient behavior is a very tough nut to crack.  The Verizon video ends with a nurse talking to the monitored patient about his weight gain.  But that is NOT the end.  It is just the beginning.  No one knows what that nurse is supposed to say to change the patient’s behavior over the phone.
4.     Like many mHealth ideas this system creates unrecognized changes to the standard of care and thus changes medical liability.  What if our CHF patient who stops listening to alerts and stops answering the phone dies while he is in the monitoring program?  Who is liable?

So it’s the same thing all over again with health IT.  No proof of effectiveness.  No way to pay for it.  No understanding of the medical challenges involved.  Unrecognized changes in standard of care and liability.  Health care is not the same as banking.  Duh. 

Verizon has no business getting into health care beyond the LTE connection itself.  They are going to lose their shirt investing in a treatment the literature says doesn’t work.  Perhaps unwittingly, the physician moderator, Dr. James Morrow, said it best when he asked the panel, “Where is the app that slaps my hand when I reach for the bag of Oreo cookies?”

Don’t get me wrong, folks.  Our practice has enjoyed great success with EMR in over the past 7+ years.  Our experience just scratches the surface of the awesome potential of health IT.  I want you to succeed.  But the health IT industry is headed in a direction that will guarantee failure.  To succeed you must stop chasing pipe dreams and focus on the one goal that must be met before anything else – HIEs, mHealth or anything else – can succeed:

Find a reliable way for doctors to succeed with EMR in the office setting.  Upgrade EMRs to reflect some understanding of the practice of medicine.  Design patient portals that actually work.  Demonstrate that EMRs are effective at improving care.  Design a business model that shows the path to a return on investment.

Until that goal is met, nothing else matters.

Friday, December 30, 2011

Deep Thoughts from the Meaningful Use Mountain Top

In some ways I am grateful to see 2011 end.  Several extracurricular projects have drained the life out of me, including our Meaningful Use (MU) project.  As near as I can tell we survived.  Our 90 days of compliance for phase 1 / year 1 are completed.  Last night I completed my attestation on line uneventfully.  We will get attestation completed for the rest of our physicians within a couple of weeks.  Then we will join the few (1%) of “eligible providers” that have complied with MU.  One would think that folks would be breaking down the doors of these “one-percenters” to learn the secrets of their success.  Yet a brief Internet search reveals no doctor testimonials on MU success beyond the second hand accounts offered by EMR vendors and consultants.  These are of little value.  So I am writing my testimony.

Over the past several months I have repeatedly criticized MU, with good reason.  But perhaps now that I have climbed the MU Mountain and my check will soon (hopefully) be on its way, I should soften my view a bit.  Sort of like final exam week…exams looked awful before you took them, but when you got done and you were somehow still alive, well, maybe it wasn’t so bad.

Well, sorry, it’s still that bad.  It took about 150 man-hours of work to complete this project.  And our EMR use, our quality of patient care and our practice efficiency is for the most part no better.  In some ways it is worse.  As a result of MU:
1.     We now take blood pressures on children.  This is almost never medically relevant in an ENT practice.  We can’t exempt ourselves from this requirement because of our adult patients, in whom blood pressure is often relevant.
2.     We waste volumes of paper printing clinical visit summaries that no one reads.  While the concept of a visit summary is OK, the document itself must include so much extra data it is useless.  Our web portal, which we are in the process of replacing, does not support this requirement so we have to use paper visit summaries for now.
3.     Patient waiting time is increased while we process data on pneumovax status, smoking status and body mass index on every patient.  In our practice these data are medically relevant for many patients, but not everyone.  Doing it for everybody is a waste.

To be fair, a couple of good things did happen:
1.     Use of EMR-based prescriptions and true e-prescribing (e-Rx) improved with those physicians that were still hanging on to paper scripts and/or were not using e-Rx.
2.     We were not maintaining true ICD-coded problem lists in the EMR before MU.  We had problem lists and diagnoses of course, and we were using ICD codes for billing.  But we had never combined the two processes before.

The entire process is complicated, confusing, and intimidating.  Not only are the guidelines themselves a mess, but also there is a surprising amount of inaccurate and misleading information out there.  Even the CMS publication Attestation User Guide is missing a page compared to the actual attestation web site.  After reading the User Guide I lost an entire night’s sleep thinking that the “children with pharyngitis” quality measure had been deleted because it is missing from that document.  I have 17 years of medical practice experience and 37 years of IT experience.  If I can’t figure this out there is something wrong.

The view from the top of the MU Mountain looking down is no better than the view from the bottom looking up.  Meaningful Use remains an expensive distraction that forces the true benefits of EMR to be overlooked in favor of regulatory compliance.  MU also creates an unhealthy alliance between government and the health IT community.  The government wants to own health IT just like it wants to own the rest of health care.  Don’t fall for it.

Sunday, February 20, 2011

Our Disaster Recovery "Fire Drill"

Last Friday our practice had an opportunity to practice our disaster recovery protocol.  This was actually good news; we are replacing our 6 year old servers with virtual servers and a storage area network (SAN).  The implementation plan required more down time than a weekend would provide so we added a Friday to it.

Unlike a real disaster we had the opportunity to prepare immediately before, knowing exactly when the system was going down.  

Friday, January 7, 2011

Over-Automation of EMR Note Creation Encourages Missed Diagnosis and Incurs Medical-Legal Risk

Over the past several months I have read several online discussions and comment threads on the medical-legal issues raised by EMR, including an HIMSS brochure on the subject.  Most of these discussions miss what I consider to be most important legal weaknesses of an electronic medical record. I finally came across an online discussion that comes closer to covering what I consider to be the most important medical-legal issues.

When we were setting up our EMR about 6 years ago many of our docs came to me with the same request:  “I want to create a chart note with a single button click.”  Although that was obviously a bad idea, their desire for it was understandable, given our inexperience at that time.  Templates are widely recognized as an effective method of documenting care and complying with CPT coding requirements.

Tuesday, November 2, 2010

Monday, October 25, 2010

We have Moved to

Sorry for the inconvenience.

Please go to where I have a new post entitled, "Bedside Manner in the Information Age"

Mike Koriwchak, MD