EHR Inbox Reduction

Webinar (series)
A System-Level Approach to EHR Inbox Reduction
May 9, 2023

As the physician’s workload grows, so does the volume of the inbox, creating a burden that is impossible to manage alone. Physicians often spend extra time before and after clinic to complete between-visit clerical work, which can add one to two hours to their workday and contribute significantly to physician burnout. In this AMA STEPS Forward® Innovation Academy webinar, join Christine Sinsky, MD, and Jane Fogg, MD, MPH, in a live discussion about the benefits of taming the EHR inbox and how to take the first step to accomplishing that goal.

Learn more in the STEPS Forward® toolkit, EHR Inbox Management Toolkit.


  • Jane Fogg, MD, MPH, chair, internal medicine and family medicine, Atrius Health, Boston


  • Christine Sinsky, MD, vice president of professional satisfaction, AMA

Explore the AMA STEPS Forward® Innovation Academy on-demand library of webinars on physician burnout, digital health, private practice, BHI and more.

Dr. Sinsky: Good morning, everyone or good afternoon depending on where you are. It’s so great to have you joining us live and for those who are watching a recording, we’re glad you’re able to see that as well. It’s really a pleasure this morning to be talking about the inbox. One of the priorities that we’ve put at the AMA for the last two years has been to address the EHR inbox, burden and reduction. And so with that, it is my great pleasure to introduce Dr. Jane Fogg, who will be our presenter today.

Dr. Fogg is the chair of internal medicine and family medicine at Atrius Health in Boston, and she leads a team focused on primary care redesign for reliable systems that are team-based, patient-centered and return joy to the practice of medicine. Who could ask for a better mission statement?

She also leads programming in palliative care, geriatrics and hospital services and the organizational initiatives to advance care. I know that Jane is really committed to improving the patient experience and the physician’s and other health care worker experience. She’s committed to value-based care and she has been an innovator in practice both as a physician and as a health system leader.

I’m just so much looking forward to her presentation this morning and then our live interaction with questions that you ask in real time. And we’ll go over some of the questions that many of you asked at the time of registration. So with that, I’d like to turn it over to you, Dr. Fogg.

Dr. Fogg: Good morning, afternoon, evening. Thank you so much. It’s really a pleasure to be here and I’m extremely happy to see how many people have joined. This is a really important topic. So as Chris said, I’m Jane Fogg. I’ve been a PCP and a physician leader for many a year and my work has always centered on how we can create the conditions in primary care to deliver great care to our patients, but equally be able to do our work well and happily. So, very excited to share some of the progress we’ve made.

Today I will be talking about in-basket reductions specifically, and how I made some progress at our organization and also really to highlight the STEPS Forward® module, which shares really fantastic information on how you can do the same yourself. So a little brief introduction and then we’ll get right into the pragmatic aspects of the work and share a few tips and pearls, and we’ve left ample time for questions as well.

Just sort of start with the headline, which is, what did we do? This is Atrius Health data, and this is the in-basket message volume for PCPs specifically spanning over 2016 up until most of 2022. We normed it for clinical FTE as well in the orange.

And what you can see is that over that time the total message volume normalized by FTE has actually reduced by about 25%, which is considerable. We’re not done yet, but this shows we’ve actually made some real progress.

So how did we actually get to this place of in-basket reduction? And we’ll dig a little deeper later in the webinar, but this is the work that we did that led to that 25% reduction. Specifically, we used a strategy of elimination, automation, delegation and collaboration to find a multitude of tactics to reduce the total volume. We really understood that there was no one single fantastic solution that would solve it all and that we had to tackle this piece by piece to get the results we wanted.

So I’ll expand on some of these examples, but you can see areas where we’ve actually eliminated messages—Media Manager and CC Chart, Hospital—I’ll talk a little bit about the automation that we employed in our in-basket to reduce prescriptions and normal labs, and talk about some of the delegation and collaboration work that we’re doing that is ongoing and a very important part of the in-basket.

So our process and what we think is a very sound way to start is this simple five-step process and the beginning is always understanding your current state. And I’ll talk a little bit more about that in the next slide. Getting your strategic approach, creating governance and work groups, who’s going to participate, how decisions will be made, establish and execute on those tactics and measure the impact, make sure that you’re actually making progress.

Membership Moves Medicine™

  • Free access to JAMA Network™ and CME
  • Save hundreds on insurance
  • Fight for physicians and patient rights

So as I reflect on this process, I think one thing that has become clear to me is how you start is very, very important and who you involve. And so I think of this in sort of two ways. The first is that, who are we involving in this work? And one might assume that we should just involve clinicians to get together, design what they think would be best, and ask someone to build it for them. I would actually argue that if you combine your clinical, your technical and your operational experts at the very beginning of your work, you’re going to find solutions you didn’t know were possible. The co-development of solutions between the technical experts and the clinical experts really led us to find new ideas.

As example, we didn’t know we could eliminate certain routing pathways. We didn’t know why some things were getting routed into our in-baskets. And so when we sat down together to examine what we wanted to do, we learned from our technical experts and they learned what was important to us, both the clinical leaders and the operational leaders. So we really blended new perspectives on what inbox design could be by sitting at the table from the beginning.

The second important piece is, take some time to evaluate your current state. Don’t minimize that. We all know the in-basket is terrible, but I think you really want to explore quantitatively and qualitatively what’s happening in your in-basket at your institution. Certainly we start with the volume of messages. How do we measure it? How do we know what it is and what type of messages? How are they categorized? What are the relative proportions of each message type? That’s basic.

But then I think the next part, which was a little more qualitative, was what was the intent of that message? Did I need to do something or is that an FYI for me? Did I need to see it, or did somebody else need to see it? And that leads us into who is really the ideal team member to address that particular message type. Typically today most organizations route things to a physician and expect that they will then send it on to other folks on their team to take next steps. I don’t think that’s a durable solution and there is simply too much volume coming at us. We need to get it to the team first and they need to delegate up to you, the physician. So take some time in the beginning.

In the STEPS Forward® packet that I hope all of you have next to you, this is the same table, but expanded a little bit. And just to articulate that the important part of building a strategy in this is perhaps stop wishing there was one thing you could do and you’d be done. There’s going to be several things that you have to do and we think that if we can first look at ways to eliminate waste, that takes things away without having to change workflows for a lot of your clinical team and you. And then moving towards automation where you’ll have to do deeper work with your teams, and I’ll explain through some examples on how we can change workflows, how we can trust automation.

And in delegation, it’s incredibly important that you have a strong team structure and so you can’t start delegating in the in-basket if your clinical practice doesn’t have a good team structure to build off of.

Collaboration I think is such a key point. Most in-baskets were designed as if there’s a singular user and that’s all that’s needed. The reality is, is that the in-basket is clinical information, it’s clinical care and it belongs to the team. And so we need to figure out how we can have multiple people participate, how we can reach across departments and participate.

So this is a framework that we used and what I’d like to do now is actually dig into a few examples to really show you how we did it, because I know that that’s generally what people want to hear.

So I’m going to start with a place that we started, which was CC chart, which is that carbon copy chart or the note that is sent to you by a specialist and urgent care colleague or perhaps one of your cross coverings and they’re sending you their note—and it comes in CC chart. And when we looked at this back in 2017, it was 16% of our total message volume. At that time, when we looked at the total message volume for a full-time PCP at Atrius Health, they received 100 messages per business day, 500 a week. So this was 16 a day. The volume was too big.

But then we actually started looking at the messages themselves and when we looked at them, we saw that many of them held low clinical value, meaning it was a normal dermatology check, “we’ll see them next year.” Or “saw for a sore throat, did a strep test, gave him antibiotics.” Things that you don’t actually need to take an action on. And when there was something valuable, such as a specialist making a new diagnosis of significance in your patient, the CC chart was sent and you had to read the whole thing to find that. So we looked at that and we saw this high variation in clinical value.

We also sat with our Epic partners who showed us why and how things were routed to the PCP. We had no idea that long ago people had made decisions about which department routes to which PCP office and why. And so we had this very messy, highly variable routing framework that didn’t make a lot of sense and was based on decisions made many years ago. And what we first did is we just got rid of that.

We said no more automatic routing. If you’re going to send me a message, you have to push it.

You have to actively send it.

We then sat with our colleagues outside of primary care and came up with practice agreements. What should you send me and why? And if you send it, I’d like you to put a little header so I can see—new diagnosis, change in prognosis, something needs to be done—so I’ll know when I look at my CC charts, there’s something I actually need to do.

So we ended the automatic routing. We set up the practice agreements advising our colleagues on what to send us. And then we purged. We found that a lot of folks had kept CC charts in their in-baskets for years, even. We purged any CC chart out of an in-basket that was over 60 days. Mind you, these are all notes that are going to be in the chart. You’re not eliminating data but you’re taking it out of the in-basket. We quickly achieved about a 40% reduction, which persisted, and we found that people were saying, "Now when I see a CC chart, I know it’s more likely to have something valuable for me." So that was an elimination of low-value clinical work that came out of the in-basket.

After we did that, we moved on to Media Manager. Now, Media Manager is that folder for us that has all those scanned documents, anything that comes in and has to be scanned and sent in that way, and it usually gets scanned in a central medical records department and then routed up to the PCP who would have to read it and “done” it, and then it was filed in the chart.

It was a small volume across the board for us, but when we looked at some of our sites who are closer to outside facilities, outside hospitals, they had a higher volume. So it was plaguing certain sites more than others, and most importantly, uniformly, people said, “I can’t find that … there’s something valuable in here, but it’s not that frequent and I have to read a lot of stuff to find it.”

So we sat down, we created a governance group, and this governance group had PCPs. Some of my chiefs of my departments are practicing PCPs. It had our Epic colleagues, and it had the leader of our information health systems, the one who did the categorizing. And they set up a grid and they categorized things as what these documents were. "Do you need to send it to the PCP or can it be what we call 'filed violently?'" Meaning it goes in the chart, but it doesn’t have to be checked off by a PCP, just goes right into the chart.

We also found things that were misfiled or miscategorized, and they were valuable. One of the things that was most common was a discharge summary from a psychiatric hospital. That’s something we want to have. And so we recategorized it to a folder where you could find it if you were looking for it. So we made new assignments, we eliminated a lot of things that you don’t need to see, and in the end we had a 98% reduction in that folder, which continues to persist.

The other thing I’d say in this space is that sometimes we found workflows that were creating more waste in here. Specifically, someone would send a consult note on paper to a PCP, the PCP would read it, sign it, send it to medical records. It would get scanned and sent back to you to read it again in the in-basket. So we sat down and we really worked on those workflows to make sure that we weren’t sending things to people twice. We found waste that we created ourselves.

We got pretty excited by that and we went to the next folder where we thought we could find some waste. And this is actually a very important folder. We have a folder of emergency room and hospital notifications. We, working in Boston, work with about 14 or 15 hospitals on a routine basis and they send automatic feeds, they’re called ADP feeds—or admission, discharge and transfer. These are automated feeds that will send a message to our organization that goes straight into the in-basket and tells you if your patient is in the ER, in the hospital, getting discharged, going to a skilled nursing facility, going home. Unfortunately, what we found is because of the automation, there was a lot of duplication in here. And for every admission, there were about six notifications, and that can get kind of annoying.

We also have our nurses look at daily registry reports of who’s being discharged from the hospital so we can do the appropriate outreach and make sure we’re caring for our patients. So we could take this away from the physicians without impacting that. When we considered all our options, this is one where we completely removed it from the in-basket and our IT team built a dashboard that you could pull.

They took this information and created a PCP-centric dashboard that you pull whenever you’d like to. It lists all your patients who are admitted or discharged in whatever time frame you want to look at, but it was also designed to give you the information you need as a clinician. Where’s the discharge summary? Has someone called them? Do they have an appointment with me? Those are the most important things.

So we made sure that was highly visible. So this was a 100% reduction, and putting information in a much more user-friendly format for the PCP.

Then we moved to automation and this was actually a very deep and involved project that really took our entire organization over a couple years and has been absolutely fantastic. We know in primary care and in many specialties there are many routine reliable tasks that happen again and again and don’t necessarily need you, the physician, to sign off on them each time. Prescription renewals is one of these.

It’s not unusual that beyond our patients requesting renewals―pharmacies, insurance companies―there are multiple ways that people request renewals from us. We tried very hard to get people to give longer refills―90 days and 4 refills―which we think is the best practice to eliminate waste. But we still had a very high volume, in fact, 16 per day per full-time PCP back in 2017.

We thought about what are we doing when we renew a prescription? And we’re typically checking the chart, making sure the patient’s up to date with their care needs, deciding if we’re going to renew it or if we’re going to ask them to come in, or something to check. Or if we just need to remind them to book a follow-up in months to come.

We wanted to find a solution to automate this and we wanted to move quickly. So we decided to work with an outside vendor that had a mature product that was essentially an automation platform that we could put right into our electronic health record. We are on Epic at Atrius, but this same platform could be put in a multitude of other EHRs.

What the platform is doing is when the request comes in, the automation is scrubbing the chart. It’s looking to see, do they have the appropriate blood pressure readings to get their antihypertensives? Do they have the appropriate visits set up in the future and the past? Is there any lab work or any other care gap that I as the physician would be looking for when I renew a medication?

It worked well―we did do some piloting and we made some adjustments so that we could align clinical operations with how we worked, but we also made sure that we had physician governance ensuring that these protocols had great oversight and were endorsed by a body of physicians, nurses and clinical pharmacists. We reviewed every protocol to make sure it matched the Atrius Health standards.

After we finished our pilot of two sites, we spread it to 21 and in that work ... so in an adult primary care space, we reduced the prescription renewals that go to the PCP in-basket by 50%. We do not require a physician to sign off on them, they are automated under your name.

It continues to work well, and we centralized our team over time to take advantage of some level loading and some efficiencies. We saw a big backlog every Monday. We wanted to apply some resources so that we would be caught up on Monday and not wait until Thursday to get caught up on the renewals. We made sure to have some metrics that looked at patient experience, specifically, how quickly are we getting the renewal to them? What’s our service standard?

So there were a lot of pieces of that to make sure that patients were having a good experience while we did this. And it continues to be an incredibly successful automation that we’ve enjoyed, and we’ve been spreading it to other service lines beyond adult primary care.

So, emboldened by the ability to put automation in clinical practice, which doesn’t happen a lot in primary care, we thought about another automation that we might do. And so this was looking at normal lab results. We know in total the results folder of most EHRs is almost 20% of it. It’s quite a bit and we do a lot of preventative medicine at Atrius Health and so we certainly are checking hemoglobin, A1Cs and lipid profiles and so on. And we were sort of shocked to see that across the board we had a pretty high number of normal results. We looked at current state, almost 70% of my adult primary care practice, the results were actually normal.

We talked about the idea and we dedicated a physician leader who had a multidisciplinary group working with them, including lab leadership, including several physicians, nurses and others. And they set about the task of saying which of these normal labs could we not tell the PCP about or the ordering provider—simply go straight to the patient?

Now we are all working in an environment with greater transparency of clinical information. Our patients have access to their data much faster. So this is very much aligned with the patient’s ability to see things on the portal in real time. But we really discussed and figured out what’s normal-normal, meaning it’s normal in any clinical circumstance. There’s no circumstance in which it could be abnormal. So things we couldn’t work on are things like a thyroid-stimulating hormone, because if a woman is pregnant or there’s other conditions going on, you can’t rely on the reference range to say it’s normal. But we were able to identify a fair amount of things that are normal in all clinical circumstances.

We vetted this extensively with our physician leadership. We have leadership at all of our 21 sites and all of those chiefs were involved in thinking about the professional change to not see the labs you ordered. We thought about when you’d want to see the lab, even if it was normal because it might spur you to do something else. And so we gave a little check box that will say "CC to me no matter what" in case you really want to know, even if it’s normal.

We piloted it. We had a little pandemic disruption, so it was a longer pilot than expected. But when we were done with the pilot we were able to expand pretty seamlessly across 21 sites and it took 30% of the lab results out of the PCP in-basket. So a good reduction there for the normal labs.

I think delegation and collaboration are such important tactics and this is where these tactics start to involve your team relationships, the design of your team and how you change culture and how we all need to take care of our patients as a team and not an individual. And what I found over the years is that the in-basket rarely reflected a good team-based primary care model. We do it in our clinic but then we go to the in-basket and we practice solo.

So when we thought about delegation, we realized we really need to embed it in the in-basket routing structure. Instead of sending me an abnormal labs―one of our current pilots is abnormal vitamin D. Instead of sending it to me and saying, “Jane you can send it to a nurse and they can, by protocol, manage it.” Why don’t we send it straight to the nurse? We have a protocol for them to manage it. They certainly can ask a question if something’s not clear, but they can do this independently by protocol―all the patient review, the results, figure out the next steps, follow-up and so on. So it’s really more top-of-license work for nursing and although the volume of abnormal vitamin Ds as you can see is not that big, the important part is this is a message that takes some time to manage.

You have to find the patient and make the plan and so it is actually removing some time management, but it is also ensuring that we have a consistent approach to abnormal vitamin Ds and how we manage them. So we feel like the success of this will lead us to do many more like this where we are going to use the in-basket to enforce delegation.

Biggest problem across the country―PMAR, or patient medical advice request. We too also had 100% doubling of the medical advice requests. Now we have long had our EHR set up so that when the patient is writing a medical advice request email, they can check a box saying this is for renewal, this is for a referral, this is a medical question, non-urgent. And so they’re already sort of self-triaging. So our 100% increase was in the ones where it was a medical question, not a renewal or anything else. And we had again, well, the doctor will just send it on to somebody else. So let’s reverse that. Let’s send it to somebody else and let them triage it up to the doctor.

A lot of people are doing this across the country. I think it’s incredibly important work. What we did perhaps a little differently is we experimented with the difference between sending it to a nurse first or to a clerically trained person—we call them PSRs—first. And we actually found the PSRs were able to eliminate a fair chunk, and then delegate up to a nurse if they needed, and up to a physician or advanced practice clinician if they needed to. And when they do this, the sites where this is active, it’s about a 40% reduction.

It is definitely a more involved project, you have to do a lot of work training people and setting up what’s appropriate for you to manage and what needs to be delegated up to someone with a licensure or clinical training. But as you can see, there’s already low lying, 40% of those messages didn’t even need to go to a nurse. So quite a bit we can do there and quite a bit more that we can all do.

This was another example of measurement that was helpful for us, which was that we are able to track who’s touching the message in the top one. Which category of team members? So in the gray is RNs and you can see when they started touching it first. And then we said, “Nope, we’re going to use the clerical in blue.” And they took over and the RNs went down.

And we kept looking to make sure that the physician was still having a downward slope and they were. So we looked at who’s touching and who’s “done-ing” or completing a message. And this kind of measurement’s really helpful when you’re trying to make sure that your team is consistently doing the routing and the responsibilities that you’ve designed.

So finally, I always think it’s important to talk about the non-technical, the collaboration aspects. I was inspired by a group―a Kaiser group―who had a clinical coverage department that covered in-baskets for any leave or departure over a week, and we developed our own department like this in 2021.

We currently cover extended leaves, such as a medical leave. We cover when a physician or APC leaves and we need some help at the site to cover their in-basket. But we also jump in and help cover when we have a struggling clinician who is really falling behind in their work, showing signs of burnout. And they need some time to get new habits, such as a scribe or other things that they can do. They need some time to figure out how they’re going to develop their own efficiencies or what we might do to support improving their workload.

And so we can come in and sort of help cover their in-basket temporarily while the local team can really work with that clinician and figure out what is the best path for them. It’s called the Clinical Coverage Department and we’ve had a lot of retired PCPs who like to do some part-time work because they can do a lot of this virtually and a couple of APCs, and it’s been one of our more successful endeavors in the last few years. As all of you know, when your colleague is out or away and you’re covering their in-basket, it is undoable and very stressful for everybody. So this has been a huge help.

The other initiative that we’re doing which takes time to build, but I think is a really important part of primary care going forward, which is the idea of having an advanced practice clinician and a primary care clinician sharing a panel. And having dedicated continuity so that for years I worked with a nurse practitioner, she and I shared a panel of patients together. The patients knew they had two of us, not one of us. They never lost me, they just had two people. And our goal is to stop saying “cover me” and start saying “let’s share the care of our panel, including the in-basket work.” So this is work that’s ongoing and I think important for all of us.

And before we open for questions, I wanted to just add a little of tips and pearls that I’ve gained over the years, and you’ll find many, many things in the STEPS Forward® module as well.

I think again, the co-development from the very beginning will enhance your success.

Engage your chief medical information officer, engage your EHR leaders. They need to be at the table with you because much of this is technical and much of this is team, and you need to be together to make those decisions and find your opportunities.

A second important piece is about decision-making. I find that a lot of us don’t know who made the decision that somebody’s sending me this document and I don’t know, do I need to see it? Who decided that it was sent to me? I hear that from PCPs every day. Why am I getting this?

You can make decisions about what you and your team should see. And if you employ clinical governance and you think about the clinical value, about safety, about all those parameters, I think you will find there are a lot of decisions you can make about things you actually don’t need to see or could be seen by somebody else.

And sometimes you need to standardize a certain workflow to achieve something such as automation. We had to have certain standards in our refill management in order to automate refills for 350 providers. We couldn’t customize it to each individual provider. So I think it’s an important thing to think about standardization in the right spots where it allows you to drive forward something that supports everybody. And that’s why making decisions is really important and having that clinical government.

A good in-basket is as good as a good team. Think about your team roles and how your MAs, nurses and everybody else on your team works today. I also think I’ve learned it’s always important to reevaluate certain clinical workflows, and again in our STEPS Forward® you’ll find that a really important one is prescription refills—90 and 4 refills will really save a lot of back and forth between you and patients. You’re still going to see them next year. But it’ll save you from having to refill it again and again throughout the year.

Pre-visit labs—we started this years ago. We had finally gotten to 75% of us using pre-visit labs; took a lot of professional change management. But when the patient comes in, they see their labs with you when they’re there in the visit. It saves you time that would have been a letter, an email, and it’s actually really nice for the patients to be able to discuss those labs with you. So those kind of clinical workflows are ripe opportunities to improve efficiencies in many ways, including your in-basket.

I also think that ongoing EHR trainings to support personal efficiency skills complement this work. I never want anyone to believe that it’s just about your own efficiency. It’s not. It’s a messy system and very few people can be efficient in it, but even if we designed it perfectly—I didn’t learn this in medical school, not all of us are born with great efficiency skills to look at a large volume of work and try to sort out how to work smoothly and efficiently for you and your team. So I think that that’s a really important part of what we all need to do in our institutions.

And I know there’s some questions about things like AI and other things I think I’ll just lightly sort of leave you with. I think that there’s a lot of work across the nation to go beyond measuring volume and to understand actual time. There are limitations in some of our measurements and we need to better understand the time that it takes us to do this work and the cognitive burden of messages. Not all messages are created equal. So although mine, my work had showed a lot of volume reduction, I know that much of that volume was this, the simpler messages and that what I really want to focus on going forward is what are the more complex messages that require more time?

I think it’s important for us to all just recognize that portal utilization, our patients are using it. Epic Garden, Cerner and all your EHRs are with us, and we’re just going to have to figure out how to better design health care to meet the needs of our patients who like to use the portal for a variety of things.

Understanding individual preferences, think we’re not quite there yet. A hundred messages for me is not the same as it is for somebody else. We have to figure out what works and how people know what the best and optimal work case and work environment is for them.

And finally, I think I know some of you have some questions. I think that there is some role for AI and automation for better access opportunities for patients to find us that we could certainly improve upon. But we cannot lose the humanistic and relational care that we give. And so it’s really that balance of where it can help our teams be more efficient without chat-botting me and all the rest of you into a different future. So those are my future thoughts. And I’m going to hand it over to Chris to start with questions.

Dr. Sinsky: Well, thanks so much, Dr. Fogg. What an excellent presentation.

I’d like to first start by reminding the audience that we do have resources available. So we do have the Inbox Reduction Toolkit that Dr. Fogg mentioned. We also have an Inbox Reduction Checklist (PDF), a brief synopsis of some of the things that organizations can do, and then you can go to the toolkit and find the details about how you might begin an initiative to reduce inbox.

Dr. Fogg also wrote an article for the New England Journal of Medicine Catalyst that was published just in the last month or so on inbox reduction. So several resources specifically to systems approaches to managing the inbox.

We also have a STEPS Forward® toolkit on prescription renewals so you can learn more about the 90 plus 4 approach.

And we have one on pre-visit planning and another toolkit on pre-visit labs, so if you wanted to go upstream as part of your inbox reduction strategy and go upstream and reduce some of those incoming requests for prescription reauthorization, and reduce the need to review labs twice, you can do that by setting up pre-visit lab, pre-visit planning 90 plus 4.

So we have a lot of questions in the chat, both those that have come in live and those that came in ahead of time. And so I want to start with one that was around the 90 + 4 since that’s something we’ve been talking about at the conclusion of your talk. And this person asks really how do we encourage and force necessary follow-up, clinical follow-up, if we no longer are using an expired prescription as the hook, as the enforcer for bringing people in for their appointments? So I’d love to have you speak with us about that.

Dr. Fogg: Great question. When we were looking at our refill automation years ago, it sparked a lot of great conversations in this area and one thing that really struck me then that I hadn’t thought of before is, we’re relying on the refill, the renewal, to remind ourselves that a patient might need care. But what about all the patients that don’t call us for renewals—that don’t take their medications? What are we doing to remind ourselves about them? So it’s a really uneven and highly variable way to keep track of your patients. You’re just keeping track of the people that are actually taking the medicine and asking you for it. So it kind of shifted my thinking to our outreach mechanisms to ensure that patients are getting great care should not rely on renewals. That’s sort of the first premise.

So what can you do? And I think this gets into the area of there are a lot of emerging systems, even automation systems coming out of Epic and others where you can do reminders for people who need to come in for annual exams, need to come in for certain screenings, that is outside of the renewal system.

So I’d encourage you to look at what’s available to you not using the renewal system, because I think you’re then only focusing on a subpopulation.

Dr. Sinsky So I really like that and I often think about how expensive it is to use the prescription renewal as the hook, the enforcer for prescriptions. And another strategy is to simply reappoint people at the end of today’s visit for the next appointment, and to match that with the prescription renewal volume that you give. So maybe others have had the same experience I’ve had―“Come back and see me in a year; here’s a six-month refill for your, whatever.” Well, if that’s your approach to managing patients, you will automatically burden your staff with additional work to be done between visits. If you feel that the patient needs to be seen in six months and needs lab, then schedule that, reappoint that at the end of today’s visit.

So another question that’s come up is around unintended consequences. What are some of the unintended consequences that you might have seen for some of your inbox reduction efforts? Did you have any unintended consequences about having normal-normals go off into the portal, or unintended consequences with respect to any of your other strategies?

Dr. Fogg: It’s interesting to me that the lab automation of normal labs, we didn’t have unexpected consequences and actually people sort of doubted that it would be acceptable to physicians and patients. But it turns out we’ve been scanning for―we use Press Ganey―we’ve been looking at the comments, we’ve been asking people—nothing. And I think patients are somewhat … you know, we think about medicine, we’re a little farther behind many other industries where there’s a lot you can do as a consumer and an electronic interface to manage whatever you are, your bank, whatever it may be. And so I think our patients have been looking at their labs probably sooner than they got a letter for a good while. So that did not have unintended.

I would say though that the trickiest thing is the patient medical advice request. And I think that there are many patients who still believe that they are just emailing you, and you are sitting there ready to email them back, and you’re not in a room seeing, you know, 25 patients. And so one of the challenges we found is some patients being upset or surprised that another team member got back to them.

Another is really and I’m sure you’re all experiencing this, this emerging sort of sense from our patients that they should be able to get care via these medical advice portal emails that really you should bring into the office or at least do a video visit.

And so I think I’ve learned in that, it’s a combination of we have to figure out how to change our patients’ expectations and culture around what’s appropriate for sort of in-basket health care. We also need to figure out better ways to train our staff because, as you know, especially those of you in adult primary care, you get a lot of messages that are two or three things mixed together, and it’s very hard to parse that all out as a staff member. But I think you can do it, we just need to invest more time in that. So that was probably one of the unexpected consequences, is performance complaints in the patient medical advice world.

Dr. Sinsky: Yeah. Great. Well, there was one question live and two in the earlier questions that people had submitted around finding those urgent messages within this whole stack of more non-urgent messages. And so did you have an approach to either direct patient questions that were urgent away from the portal altogether, or to float those to the top—within the portal? How are you dealing with making sure you’re not missing urgent requests that come in through the portal or that attempt to come in through the portal?

Dr. Fogg: One important component of how we’re having our team look at the portal message first is for them to be empowered to see certain things and say, "that is a visit, and that’s a visit today." You know, "we’ll get them in today." And so we really, we found pretty uniformly, our PCPs were crying out for their team to say, “Just get them in, let’s not go back and forth. Just get them in.” And so that’s not unusual on the sites that are doing that where the clerical person will call the patient and say, “I can get you in today.” And the nice thing about having video visits still is that for patients that didn’t want to come in, well, let’s at least get a video visit if that’s appropriate. So I think that we do empower our staff to do that.

The other thing that our radiology department did, which was very helpful a couple years ago is they separated out a folder of significantly abnormal radiologic exams from radiologic exams. So you’re not plowing through the mammograms to find the one mammogram where there’s actually a very suspicious finding. And so what I hear is our PCPs say go to that folder first for the really abnormal, make sure I’ve done that, and I go to the other one where, you know, might be an incidental finding are completely normal. So that was a real help from our radiology department.

Dr. Sinsky: Wonderful. And that shows how having multidisciplinary members of your team and your planning team can make a difference.

So there were several questions around leadership buy-in and change management. Specifically, what kind of language did you find was useful and engaging C-suite leaders to lend their support to this? And I know your CEO was very, very supportive. How did you deal with staff or administration resistant to change, and particularly how did you get your MAs on board? So, I think we’re looking at both levels. How did you get leadership to support this, and how did you get MAs and nurses who were on the clinical teams to recognize that this was an important contribution that they could make and it was really part of their job responsibility?

Dr. Fogg: So starting with leadership, what I think is most important is you need to find where is your alignment of your C-suite’s priorities and yours? And I think they’re often matched, but it may not be initially apparent to everybody. And so, working with my C-suite, and I had a very supportive CEO, we are a growing organization. We need to retain and recruit more PCPs, and so we can get aligned around that. Well, what is the reason why PCPs are dropping off or not wanting to join? Well, the work. And what’s changed about the work?

And so part of what I had to do is help illustrate what the work is in primary care. And this is true in any specialty as well—specifically that the days of you see patients in the office and you take call at night and that’s it—are gone. What we’re really doing is we’re seeing patients in the office, we’re managing a lot of care through documentation and through in-basket, and that balance of face-to-face time with in-basket care has really tipped heavily and many, many people have done great work measuring how much EHR time is happening in relation to how much face-to-face. So demonstrating that through data but also through national studies was very helpful.

I even had one of my clinical informaticist colleagues went into the C-suite and she opened up Epic and she showed them how to document an annual exam. And they―they didn’t know. And so I think when they had an understanding that our work has shifted in this space, then I had a little more going to say, “If I’m going to attract PCPs and retain the great group that I have, I have to help them with this body of work. And we all want to recruit more and hire more; we want our PCPs to accept patients, we want them to be accessible. It’s not going to happen unless I can get that in-basket in a better state.” That’s where the alignment came in. And so I think that that got leadership excited by this.

Working with our teams back in 2016 and ’17, I remember some uncomfortable conversations. “That’s your email, that’s your in-basket. I’ll wait for you to look at it and you send it to me.” I think what engaged us is first as a leadership group, my nurse leader and I would make sure we were aligned before we went out together to talk to our teams. And sometimes I would speak on behalf of nurses and she would speak on behalf of doctors. We showed that we are looking across, we are sharing this work, and these are our patients, and that we need to have mutual respect for each other and mutual accountability. We spent some time―that qualitative look―at current state.

One of my leaders stood up at a meeting and read the first 10 message types in her in-basket and I just remember people’s mouths opening and saying, “I didn’t know that was in there.” And so I think we really worked on … we don’t like to call it in-basket—it’s clinical information, it’s care. And so this is our care. That opened up that conversation, it’s really important to reward your teams when they take steps forward on new tasks.

We have one site that did a really amazing job engaging their medical assistants in prepping and scrubbing their schedules for the next day but also working on in-basket and I think it was really a little bit of leveling of our hierarchies and saying, “You are important to the team as a medical assistant. I need your help with this. Thank you for getting that done.” So that culture of empathy and accountability really goes a long way and you’ve got to take some steps to build that.

Dr. Sinsky: Great. I love that example of the physician and the nurse working together and each being empathetic to their particular role types situation and speaking across those role types. I think that’s just a wonderful way to message that we are actually all growing in the same direction for our patients’ good.

So there’s a question around legal and compliance review of your processes, and I know I had sent you a question about that early on, shortly after I’d learned about your work. Is it kosher with various standard-setting organizations or CMS regulations, for example, to send ADTs directly to a dashboard rather than not? And so I’d love for you to address that and we’ll use that as one of our last questions for this webinar.

And before I turn it over to you to address that, I want to let the audience know that there are two other resources that relate to this question of what’s truth and what’s myth around legal and compliance requirements. And so we have an initiative at the AMA called Debunking Regulatory Myths. We currently have roughly 16 different myths that are common among health care systems. And we state the myth and then we state our best understanding of the truth. And then we go to the source and we quote the source, whether it’s in the federal registry or whether it’s with the Joint Commission. And those are available and you can print those out as PDFs and use those to facilitate internal conversations with your colleagues in compliance and legal.

We also have a De-implementation Checklist (PDF), which is a brief list of some things you might want to consider as an organization de-implementing policies and practices that may have at one time made sense but have either outlived their usefulness or never actually lived up to the expectation in the first place. And we sent that list to the Joint Commission who reviewed that list, we made a few minor modifications, so we know that the recommendations for consideration of the implementation in that list are consistent with Joint Commission standards and also therefore consistent with CMS regulations.

So with that discussion of some of the available resources, Jane, I’m going to turn it back to you. What are some of the legal and compliance steps that you took to make sure that you were recommending and implementing defensible practices?

Dr. Fogg: Thank you. I’m so glad that we are addressing this because I often have found that regulations lag behind clinical practice. Regulations are often ambiguous, and if you sit down and read the nursing regulations in Massachusetts, as I have, you will learn that you could probably interpret it in different ways. And so over the years, I’ve made sure that I have had great access to our chief legal officer, who’s a former nurse, to discuss, is there a regulation that oversees this particular area or not or is this just what we’ve done? And who’s making the decision? And so that conversation offline with her has been very helpful.

We have a compliance committee—I am a member of the compliance committee—and so I’ve learned in those settings you need clinical voices in compliance because sometimes we’re discussing an issue or a problem, but you need the clinical context, you know, the clinical operations person there to inform that decision. If you don’t put people in those committees, you will get decisions that don’t work well for you.

So I worked with those two bodies, but I honestly learned there’s a lot we can decide for ourselves. We have great minds to think about what is safe, reliable, effective practice. You should be comfortable saying, “Why am I getting this? Is this delegation okay? Is there a scope of practice I need?” Talking to your legal team and just know that people do it in different ways. There’s no singular rule and I love the fact that the AMA is invested in this debunking of myths because I think we just didn’t question enough as clinical leaders that there are many things we can change and there are many things that nobody’s decided and we will do what’s best and safe for our patients.

So I work with them, but I do not pass everything through them. And if I do need to pass something through like delegating automated renewals and not signing off, I spend a good amount of time really aligning why are we doing this? How is it going to maintain safety? And make sure that we’re both on the same page. And in fact, we came to that agreement and could do that. So press on!

Dr. Sinski: Yeah, and I think much of the work that you’ve done is countering another myth. And that myth is that care is always safer if the doctor does every element of it. And we both know that that’s actually a recipe for safety hazards. That we create a very hazardous environment if we are anticipating that our physicians do everything, that every order is entered by the physician, every prescription is renewed by the physician, every inbox is managed by the physician. Then we have all this multitasking, which means task switching and we don’t actually give physicians time for the deep work of doctoring.

So Dr. Fogg, I want to thank you so much for sharing this incredible work that you have done over the last five years. Not only for the results that you’ve attained, but really for the insights that you have gleaned about how we can all help to reduce the burden of inbox for our health care professionals, for our physicians and for our teams.

And I’d like to end with a couple of, bringing together two of the comments that people put into the preregistration field. One physician said, “I just feel like I’m drowning. I consider myself responsible and responsive and I just can’t keep up. And I spend more and more time, extra time in order to keep up, so I’m drowning.”

And another who said, “Can you reduce my inbox burden enough to prevent me from quitting medicine before age 45?” And if we had a mission that really mattered, I think it’s that. To keep our physicians in the practice of medicine and actually doing high-value clinical work by reducing a lot of the low-value work that has turned out to constitute the majority of many physicians’ day.

So Dr. Fogg, you are leading the charge and I thank you so much for doing the work you did and then taking the time to share that with us today.

Dr. Fogg: You’re welcome. Thank you all for continuing this work.

Dr. Sinski: And thanks to all the participants who participated. And with that, we’ll sign off. Thank you so much.

Disclaimer: The viewpoints expressed in this video are those of the participants and/or do not necessarily reflect the views and policies of the AMA.