4 Lessons Learned: How to Get Doctors to Email w/ Patients
Many have stated that lack of reimbursement, fear of patient "flooding" and HIPAA were the three primary barriers to getting physicians to adopt email as a way of communicating with their patients, but we discovered that the lack of an intellectual challenge (the "hum drum" of answering emails) was at the real heart of the issue.
Upshot: Eventually we got most of our doctors to go from not replying to patient emails at all, to providing very short answers (most of which ended with "...just make an appointment") to finally providing answers like this -
In 2013 one of my engineers had an interesting theory - He had recently enrolled in Kaiser Permanente, where staff physicians are required to accept patient emails."The emails from my doctors are so meaningless," he moaned. "If doctors are forced to accept email from patients, there's never going to be meaningful dialogue other than a prescription refill."
My first reaction was, 'Wow, my twenty-something year old employees are using the health plan - we're not paying for nothing after all!'
My second was, "OK, but you're asking for our docs (who mostly own their own practices) to do something where there's nothing in it for them. The last thing a non-Kaiser doc wants is to have to email patients at the end of their day." He replied, "But John, our doctors want better marketing. Our patients want better access. Why can't we figure out a way to pay back our doctors who accept email in "social currency" by leveraging the happy patients they email?"
They drew this for me on the whiteboard -
My third reaction was rolling my eyeballs (I find that I do a lot when I meet with millenials). "Uh, duh, ok," I said (sounding a bit millenial myself). "I'll get right on that."
The engineering team had a plan to turn our current mobile-based secure messaging system into a collaboration tool that would eliminate the things that doctors hated about patient email and a way to turn a doctor's email answers into marketing (social currency) for our doctors who opted into this feature. I was skeptical - I gave them a small budget, 6 weeks and another 2 weeks to show what we in Silicon valley call "traction" (results).
"If it doesn't work in 8 weeks, it's over as a failed project, " I told them.
I didn't think it was going to work, but the one thing I learned about young engineers is that if they get passionate about something, you've got to let them fail (and occasionally succeed) on their own. WARNING: The previous link leads to a Youtube video where all the young people (should) have good health insurance.
It turns out they were right. I was wrong. But before I go off and give credence to why most Silicon Valley VCs don't like investing in people over 30 years of age (which makes me a 40 year old dinosaur) in fairness the initial model conceived by my engineers wasn't the one that ended up working.
So for anyone curious how to get doctors in independent practices to accept email (or do anything else as far as physician alignment for that matter), here are the 6 things we learned on our way to building a platform that eventually did this -
[caption id="attachment_4917" align="alignnone" width="606"] Despite the hype (and incorrect date range) from the marketing dept, the graph (and traction) are real. Current 2014 usage is nearly double that of the 2013 peak.[/caption]
Lesson 1) You can't fake love. As social proof that email begets great reviews online, another one of my engineers ran a keyword study on Yelp's highest rated doctor's offices in the Bay Area. It turns out that the word "Email" is the most frequently occurring word in 5 star reviews.
Even one of my favorite healthcare Twitter'ers (is that the correct lingo?) Rob Lazerow had once chimed in on the subject -
We've seen many doctors and healthcare executives try various things to "pump up" great reviews of their brands online - using reputation services or asking patients for great reviews (both shady IMHO) or even handing patients a Google tablet to write a review before they leave the office (asking may be illegal in some states).
Before we wrote a single line of code, we ascertained through our research that giving patients email access (even if it's limited in scope and expectation) was at the core of great patient experiences and five star reviews. Everything else was going to be short-lived marketing fluff.
Lesson 2: Doctors can learn to love email, but not administrative business email. When we had our first feature set (called "Minimal Viable Product" by startup geeks) we went into a "Beta" where we asked 20 of our current clients to use our new mobile secure messaging feature. We were a fast growing digital health startup with revenues, customers and bright ideas - and we were experienced mobile developers! What could go wrong?
Decline of usage is what went wrong. Horribly, terribly wrong.
Out of the 20 clients we had in beta, 16 of them stopped using our feature after 3 weeks. What was worse was that patients did not stop using our system to message their providers, and we had to "go dark" on these patients by shutting off this feature (and some were very upset we gave it to them, then took it away). The other 4 doctors, it turned out, simply switched from their existing email clients (Hotmail, Outlook, etc) to PANDA (our software) - so in fact, we had affected nothing except to annoy people in our social experiment.
It took us many rounds of interviews and even knowing which questions to ask (this is where I would like to think my 'old man wisdom' comes into play) before we figured out why doctors, even the ones who were initially excited about the idea, stopped replying to patient messages - replying to patients was boring.
Sometimes (most of the time) we get so caught up in the day to day operations of this company that we forget why doctors became doctors. Sure its to help people, make a decent living (sort of) but to also challenge oneself intellectually - whether its studying for the USMLE or CE or handling challenging cases - doctors want to be intellectually challenged.
What we eventually discovered was that the third or fourth time a doctor got a question from a patient about billing, appointments or insurance - a switch shut off in the doctor's mind as if to say, "Experiment failed. This might be ok for Kaiser docs, but it's not worth it for me."
That's what led to -
Lesson #3: Traditional email doesn't work between doctors and patients. We were 5 weeks into my self-proclaimed 6 week code sprint to get this project off the ground, and we had to completely retool and redesign our secure message function. If it was indeed true that doctors didn't continue to stick with patient emails because they too gotten got routine questions that were not medically or intellectually challenging, then we needed a message system that was very unlike email. We ended up going from building 'Secure Email' (which was just mobile-friendly messaging behind a password firewall) to something we called "Secure Message Collaboration." It looked like this -
And..... what do you know? It worked!
When it all finally started to come together at the end of week 7 (better late than never?) I realized I had not only underestimated my engineers - I had grossly underestimated the size of our doctor's hearts and minds. As the CEO (allegedly) of any company (my last one had 300 employees, my current one has 30) you hear a lot of complaints. Complaints from patients, from doctors, from staff, from analysts, from investors, from employees....Sometimes you forget how wonderful human beings can really be.
It turns out that more and more of our doctors - when given convenience and an intellectually stimulating environment where they know their time on email is spent delivering care - will do it. But it has to be a 'message triage" system where admins and mid-levels can easily help patients with more routine treatment (Rx refill, Lab results interpretations) and admins can help patients with insurance, billing etc. Leaving the doctors to do what they do best - provide care efficiently.
When "Critical Mass' hit - it hit hard. We had to move from our existing data center provider on the East Coast to Amazon Web Services - where we could 'burst' our data and usage pipes on demand. We saw all these spikes in usage and lull periods that we just didn't understand (example: why do so many women hit up their OBGYNS and PCPS on Monday mornings, while doctors like replying on Sunday evenings?)
So we conducted a second set of surveys on the first 100 medical offices using our new 'Secure Message Collaboration' feature, but this time I had the luxury of asking "Why do you use it?" instead of the dreaded, "Why are not you using it?"
It turns out Office Managers hate voicemail retrieval (and spending time on the office phone in general) and doctors are (for the most part) seeing that patient emails are worth it to them if they feel like the administrative and lower level medical service functions can be taken care of by their staff in a 'tag team' environment.
So much so that more offices were asking us how to roll it out to their patients in time for them to achieve Meaningful Use Stage II's "5% requirement." And that's where we learned -
Lesson #4: Your competitor is not another startup technology - it's the telephone!
Oh, we tried all the things that Mayo Clinic and others must've tried. The sign behind the receptionist letting patients know they can email their doctor, a newly revised opt-in form during patient intake, a newsletter blast, etc. Nothing moved the needle. It's no small wonder - having been through this lesson - that even the mighty Mayo Clinic can't get 5% of their patients to enroll in a basic patient portal.
You're competing with the telephone!
It turns out the funnel by which most patients book an appointment online is to Google the doctor's name on their phones, look for the phone number in the local listings, then call the office to book an online appointment.
They key was to divert this 'funnel' and place them into a path patients could consider and adopt - conveniently - as a passive diversion from the funnel they knew.
It ended up looking like this in an early beta version (it has since become simpler) -
But compare that to what we replaced (this form goes one and on...) -
Like a lot of things in healthcare tech - things seemed so simple, until they weren't. Until we figured it out, then it seemed so simple again. Until... an enterprise client wants their own implementation?
* In the next blog post I'll cover exactly how we did Part II - connecting "Social Currency" with a "Message Triage" system to nail the 'Viral Loop.'
* I am the author of this article and the opinions here are my own.