How To Control the Wildfire of Patient Portal Messages

Brad Crotty MD MPH
Inception Health
Published in
11 min readJun 20, 2022


Ever since the first patient portals emerged circa 2000, clinicians and healthcare organizations have been dealing with the smoldering challenge of incorporating patient messaging as an integral part of the care delivery lifecycle. Portal messages from patients (e.g. MyChart® messages) have been perhaps the most difficult to fit into the workday of clinicians; many healthcare systems are struggling to respond to the imperative of a better experience for patients while adapting their business model, and more broadly the culture of care delivery.

This effect has been compounded in the last couple of years with the COVID-19 pandemic which obliged the vast majority of healthcare organizations to ramp up efforts to sign patients up for online patient accounts to manage care digitally and to offer telemedicine services. The net result is that most clinicians feel that they have a fire in their inboxes; the winds of digital transformation are fueling a rise in messages, and many feel that there’s not a great strategy to gain control. Short-term fixes feel like they are not enough or backfire (usually in predictable ways).

It’s sort of like climate change and wildfires, now that I think about it.

How may we gain control?
Wildfire experts have devised strategies for managing fires, using innovative and unorthodox approaches using techniques such as controlled burns, for example. I have thought a lot about this topic and it seems that the analogy may apply here. I don’t see the climate changing any time soon to be less conducive to digital connections, but I can envision a few ways that we need to adapt and manage the landscape of digital communication to avoid being burned.

Let’s start with understanding some of the trends and data.

Background, Data, and Trends

Healthcare (and I’m including the entire ecosystem of government, payors, and providers) has had more than two decades to figure out how to incorporate patient messages into our work and business models.

Health Affairs article published in 2014 reviewing data from 2000–2010 in patient messaging

Nearly ten years ago, we wrote a paper in Health Affairs reflecting on the first decade of patient-clinician messaging. The writing was on the wall… as more patients signed up for portal accounts, clinicians would see more messages in their inboxes. But some clinicians adopted messaging as a big part of their practice, while others did not.

At that time, we wrote in our policy implications:

As the use of secure messaging becomes more prevalent, a mechanism for reimbursing physicians and accounting for the workload of electronic messaging will be important. For instance, if physician reimbursement systems move toward fixed population-based payments, instead of visit-based payments, physicians will be better able to incorporate telephone and e-mail time into their workdays.

Undoubtedly, even ten years after this was written, most clinicians are straddling the fee-for-service and the fee-for-value world. A few places, like Kaiser Permanente (where co-author Yonas Tamrat works), salary their clinicians and can build virtual care (including messages) into their structure. For most others, these types of messages are an add-on to the day.

Previous research has shown that patients are satisfied with using patient messaging features if they can get a response within 48 hours. We also know that the turnaround time for many messages exceeds that for a meaningful response. Triage by nurses and then routing to clinicians sometimes adds additional delay. Forty-eight hours seems like an eternity nowadays when we can request a service instantly on our phones.

One of my colleagues back at BIDMC summarized how they quickly got through their messages (reply “Yes”, “No”, or “Please, make an appointment”). But even this sometimes breaks down in the era of high-deductible plans where patients will (reasonably) ask if they can come to get their labs drawn and schedule any preventive services they need even without scheduling a visit. Or people will develop a clinical need and it is far easier for them to send a note to a clinician that they trust rather than navigating where and how to get the care they need on their own.

I was looking at our own practice’s data just recently, observing that even over the last year (after our initial large increase in portal users) we have continued to see message volumes increase by a third.

Internal Data

What People Want

What I take from these data (both historical and recent) is that people find value in electronic communication. Generally, patients are in the doctors’ offices for maybe 2–4 visits per year, but a lot happens in between these visits. But we haven’t found a way to differentiate between a quick follow-up question and a more involved workup for a new, undifferentiated problem. They are both office visits in the contemporary care model, or else phone or portal messages at the end of the day.

Others are solving this problem…

Look at companies like Hims & Hers. They offer asynchronous care for a range of problems; their more interesting area, though, is their move into mental health, care that is more similar to the longitudinal care provided by health systems versus the more transactional nature of erectile and hair loss medications.

Additionally, several new primary care companies are baking in digital interactions as their first point of entry. New entrants such as or Babylon are starting to meet the needs of their users virtually first. Even large companies in the industry such as United Healthcare are offering virtual-first primary care services.

So where does that leave health care organizations that are not part of venture-backed new care models or who are ‘payvidors’ where they control both the care and the payment, such as UHC?

My perspective is that we have to pivot to better meet our customers more where and how they want to be met, even if the payment model is not entirely ‘all there.’ We have to run towards the disruption.

One of my Friends advising our course of action

We have to run towards the disruption

Changing Our Models

So, how exactly may we change our models to quench the wildfire?

We can’t change our climate, at least not overnight. And we are resource-constrained, where we can’t add more people easily to address the messages. We know the landscape is not always favorable. And the wind is blowing.

There are a couple of ways we can approach this. Following our wildfire analogy, we can look at removing heat, shaping the land, and fighting fire with fire.

Artist’s rendition of the Fire Triangle — Source: USFWS Alaska via

Removing Heat

In a forest fire, you may see fire crews dousing a fire with water or flame retardant, or even aircraft deploying treatments from the air. The goal here is to remove heat. Heat is needed both as an ignition source, and also to promote the combustion process, drying out fuel ahead of the fire and warming the air to help the spread.

In our analogy, we’ll use the demand for simple portal messages as heat. We can do a few things to remove that heat. We can put character limits to reduce the length of messages. We can put up a toll booth and charge for messages, either to people directly or to insurance with appropriate consent.

My worry with these approaches is that

  1. Roadblocks make the experience worse without finding a way to better address the problem. For example, character limits result in patients needing to send two messages, which are often now two separate encounters that staff need to straddle. Why would we want to make it more complex for patients to communicate their needs to us?
  2. By adding costs to patients for messages, we lower demand artificially. These tolls differentially impact low-income patients, precisely the patients that we have been trying to enroll to overcome the digital divide.

But there likely are some ways to remove lessen the burn and remove some of the heat of the messages. Ensuring that protocols exist within clinics to address simple clinical questions may reduce the number that requires clinician input. Team-based care approaches that designate a team member to review and address messages may also help. For example, teams may designate an advance practice provider or physician to be responsible for addressing messages for the team for any particular morning while they have reduced in-person duties.

Shaping the Land

In fire-prone areas, municipalities or the forestry service may take bulldozers to remove fuel or otherwise shape the land to be less favorable to fire.

For healthcare organizations, I think we must continue to shape our contracts with payors, such as taking on more risk or finding other ways to align incentives to care for patients in sensible ways. If we can meet a need virtually or asynchronously, why would we want people to come in for an in-person visit?

I believe that fee-for-service payments are like fossil fuels. They are still fairly plentiful and hard to wean off of. But weaning off is the right thing to do. The newer care delivery options described above that are digital-first are not consuming the same FFS diet. And as Hemingway wrote in 1926, humans tend to notice things happening “Gradually, then suddenly.” (Hemingway was talking about going bankrupt…) Payors have a role in this too, and can increasingly offer fee-for-value contracts that better align all parties. In some circles, patients are going the ‘direct primary care’ route or the concierge route, which is another way of accomplishing a ‘subscription-based’ care model that can absorb the digital and telephonic interactions that are difficult for most physicians to fit into the workday.

Working directly with employers to help manage healthcare costs in ways that are more convenient for employees (and less disruptive to employers) as well as governments (more convenient for citizens and less restrictive for providers) such as through some direct contracting arrangements or accountable care organizations.

It remains to be seen if individuals would be willing to pay for a subscription that includes digital access to their clinicians. Most people have come to accept digital access as a natural part of being empaneled with a care team, particularly in a medical home. People also may have a difficult time paying for services outside of their current premiums. And then there are the ethics of considering that subscriptions may provide differential treatment (turnaround time, for example) when we think of healthcare as (ideally)egalitarian.

Fighting Fire With Fire

This I think is the most interesting of the approaches. What does ‘fighting fire with fire’ even mean? In managing wildfires, ‘prescribed fires’ help control fuel supplies. For example, brush and debris may be burned in a controlled way so that the ground cover does not supply more fuel should a wildfire break out.

In the case of messaging, fighting fire with fire translates to better managing digital messaging through… better, more targeted digital messaging. Controlled experiences rather than the open text box. There will be a role for it, but what could be diverted or peeled off through other means?

Public Domain Dedication. Public Use Notice of Limitations:

Dedicated asynchronous communication may help with efficiency. While some organizations charge for this service, we have opted to use the service to drive efficiency rather than to generate revenue (see above re: health equity and engagement among our underserved populations). We can address 10 different areas of care for our existing patients through structured questions and clinical protocols. This enables a more efficient experience for patients and care teams.

Care navigation tools, bots, and chat-based services will also be important. We are beginning to explore how we may be able to parse clinical messages for intent through services like Comprehend Medical through AWS and to intelligently route messages, shifting people to the dedicated async channels or to other people who can more expeditiously meet their needs.

But there will be limitations to what we can automatically do. As I practice in internal medicine/primary care, it is very common for people to have nuanced questions or to have needs that do not follow simple heuristics. This is the natural complexity of medicine and people. But tools may make us much more efficient, and the ability to have synchronous chat to triage and solve problems I believe will be an important function of our work to come. But in these models, we have to fix the economic model in tandem.

The New Workday

The clinical workday is always evolving. In the 80s and even into the 90s, it was a mixed practice of outpatient and inpatient care (and teaching within academic systems like ours). Then, with increasing pressures to add patients and achieve metrics, workdays became even more packed. Hospitalists took care of inpatients, and ambulatory care providers were expected to add more patients into their days.

Newer, ‘futuristic’ models included time built in for responding to portal messages, maybe a 30 minute block in the morning and in the afternoon. But in general, these didn’t materialize.

If the business model was flipped, we may do things radically differently. Rather than seeing all people for preventive care each year in the office, we may have a nurse review all needed areas of prevention (like a hygienist manages our every 6-month dental care) and have the doctor there to solve any de novo problems. We might spend most of our time seeing (virtually or in person) patients where we need to make critical decisions (does this patient need to be admitted, see a specialist, or need their treatment plan adjusted today?) rather than spending most days in routine care and then fretting about these important decisions in the interstitial time between visits.

If we had a clinical engagement center, maybe we physicians would take over from our nurses when a patient was chatting with us and it was clear they needed additional insight added.

For people seeking our services, what we can provide them is quicker answers, more clear navigation, and hopefully, less money out of pocket spent trying to navigate care.Time to harness the blowing winds.


  1. As we have more patients enrolled in portals, we will have more usage of patient messaging features
  2. We see wide variability in how people use the messages historically
  3. Clear signal that patients want to interact digitally, solving their problems ideally without spending time coming in for a visit. Usually there is some form of medical decision making involved.
  4. We have not changed underlying business models, e.g. value-based care, subscription; Kaiser doesn’t have our problem
  5. While on fire, short-term solutions that reduce heat are not incredibly effective or could backfire, e.g. content limits, charging for messages. We need to get ahead of the fires, shape the land, and use digital in more sophisticated means to fight fire with fire.



Brad Crotty MD MPH
Inception Health

Chief Medical Officer, Inception Health | Chief Digital Engagement Officer, Froedtert & the Medical College of Wisconsin Health Network