Podcast: Play in new window | Download

Subscribe: RSS

In this episode, Barbara discusses:

 

  • Barbara Hales names the “invisible second shift”—the unpaid hours physicians spend at home on charts, portal messages, and paperwork—and how it steals time from family and rest.
  • Barbara Hales shares seven practical ways to use AI as an assistant, not a replacement, including ambient scribing, drafting portal responses, summarizing research, and streamlining insurance and administrative tasks.
  • Barbara Hales contrasts two physicians—one using AI and one doing everything manually—to show how saving just 30–60 minutes a day can compound into 120–240 hours a year.
  • Barbara Hales explains how AI can multiply a physician’s impact by turning everyday patient questions and explanations into reusable education materials and marketing content that grow the practice without adding more work hours.
  • Barbara Hales insists that doctors must “delegate the task, never delegate the judgment,” emphasizing privacy, safety, and clinical oversight so AI reduces burnout while preserving trust and the humanity of medicine.

 

Key Takeaways:

“AI’s real power in medicine isn’t replacing physicians—it’s reclaiming the invisible hours lost to documentation and administration so doctors can be fully present with their patients and in their own lives again.”

 

Connect with Barbara Hales: 

 

TRANSCRIPTION

Chapter 1 — Why Physicians Need Their Time Back

 

Dr. Barbara Hales 00:04

Seven ways doctors can use AI to get their time back. Hello, everyone. Welcome back to Marketing Tips for Doctors. I’m your host, Dr. Barbara Hales. I want to start today with a question. When was the last time you finished seeing your last patient, and you were actually finished for the day? Not finished seeing patients, finished. No charts waiting for you. No patient portal messages. No paperwork you promised yourself you’d finished after dinner. No laptop sitting on the kitchen counter staring at you while your family is doing something without you. If you’re laughing right now, I understand because for many physicians, the workday doesn’t end when the office closes; it just changes location. You go home, have dinner, maybe spend some time with your family, and then out comes the laptop, and somewhere along the way, something happened to medicine. We became doctors because we wanted to take care of people. We did not go to medical school because we had a burning desire to become world-class data entry specialists. We didn’t dream about answering portal messages. We didn’t fantasize about prior authorizations, and I’m fairly certain that when you wrote your medical school application essay, you didn’t say, “My lifelong ambition is to spend more evenings completing documentation, yet here we are, and that is why I want to talk about artificial intelligence today. Not because AI is fashionable, not because everybody is talking about ChatGPT, and certainly not because I think technology should replace physicians. Quite the opposite. I think we need to ask a completely different question: Can technology give physicians enough time back that we can actually be physicians again? Apparently, doctors are beginning to think the answer may be yes. According to the American Medical Association’s 2026 Physician Survey on Augmented Intelligence, 81% of physicians now report using AI professionally. That’s more than double the percentage in 2023. Think about that. In only a few years, AI went from something many doctors viewed with suspicion to something being used in medical practices every day, and here’s another number that really caught my attention: 70% of physicians surveyed by the AMA said they see AI as a tool that can automate tasks contributing to work-related burnout. So today, I’m going to show you seven practical ways physicians can use AI to get some of their time back. But before we get to those seven, I want to talk about what we’re really trying to solve, because the problem is not technology. The problem is time. The invisible second shift. Picture a physician we’ll call Dr. Sarah. Sarah has a busy primary care practice. Her last patient leaves at 515. She gets home around Her husband has dinner waiting. Her daughter sits at the table, telling everyone about something that happened at school. Sarah is listening, sort of, because part of her brain is thinking about the unfinished charts. She knows there are 11 of them, so she eats dinner. She asks her daughter a few questions. She helps clear the table. And then she says the sentence that physician families know all too well. I just have to finish a few charts. She opens the laptop. 8 o’clock becomes nine. Nine becomes 10. Her daughter goes to bed. Sarah finally closes the computer at 1037, and here’s the sad part: nothing particularly unusual happened that day. There wasn’t an emergency. Nobody crashed. There wasn’t a complicated admission. It was simply Tuesday. That story is a composite, but I’ll bet many physicians listening recognize themselves in it. This is what I call the invisible second shift of medicine, and this is precisely where I believe physicians should begin thinking about AI, not with robots, not with futuristic diagnostic machines. Start with what I do every day that doesn’t actually require me to be a physician, because that is where your time is hiding. Let’s begin with the obvious one: documentation. AI-powered ambient scribes can listen to the physician-patient conversation with appropriate systems, safeguards, and consent procedures, and generate a draft clinical note. You review it, you correct it, you approve it. The physician remains responsible, but you’re no longer starting with a blank page. And this isn’t just theoretical. In one medical group study involving 79 providers and more than 25,000 AI-generated notes across 23 specialties, high users of an ambient AI scribe had a 21% decrease in time spent on notes per day.

 

Chapter 2 — AI for Documentation, Inbox, and Patient Education

 

Dr. Barbara Hales  07:26

Even more interesting to me was what happened after work. Before the pilot, about 32% of participating providers reported spending eight or more hours per week documenting outside clinic hours during AI use, which dropped to about 8%. Think about it: 32% to 8%. That’s not just an efficiency statistic. That’s dinner. That’s exercise. That’s reading a book. That’s going to your child’s soccer game. That’s sitting on the couch and doing absolutely nothing and not feeling guilty about it-that is life. Imagine Sarah again. Only now her practice introduces a properly vetted ambient AI documentation system. At first, she’s skeptical. She checks every note carefully, and she should, but gradually something changes. Instead of writing the entire note after the visit, she reviews and corrects a draft. One evening, she gets home. She puts her purse down. She walks into the kitchen, and her daughter looks at her and says, “Mom, where’s your computer? Sarah says, “In my bag. Her daughter looks confused. “Aren’t you going to take it out? And Sarah says, “Nope, I’m done. Now that’s an illustrative story, but that is the outcome we’re talking about. We’re not trying to make physicians type faster. We’re trying to make it possible for physicians to stop typing. Next, the inbox. Ah, yes, the inbox-the place where a message that begins with “Quick question, doctor, occasionally contains 14 questions, three attachments, and a medical History going back to 1987. AI can help draft responses to routine patient communications. Notice the word I used: draft, not diagnose, not independently prescribe, not send medical advice without physician review. Draft. You provide the judgment. AI provides the first pass. The 2026 AMA survey found physicians are already using AI to generate draft responses to patient portal messages; think about the difference between writing 20 responses from scratch and reviewing 20 reasonable drafts. Even saving two or three minutes per message adds up, and there’s another benefit: AI can help you turn a rushed answer into a clear, patient-friendly explanation. You still decide what is medically appropriate, but AI can help you communicate it more efficiently. AI use number three: patient education. Here’s an area I think is enormously underused. How many times have you explained the same thing? What does pre-diabetes mean? Why do I need this medication? What should I expect after this procedure? What does this lab result mean? Why do I need to come back in six months? You explain it Monday, then Tuesday, then Wednesday, and Thursday morning. You’re explaining it again. What if you created a physician-approved library of patient education materials? AI can help you take your explanation and turn it into a one-page handout, a frequently asked questions sheet, a patient-friendly email, a short video script, or instructions written at a more accessible reading level. You review it for accuracy. Then you have something reusable. This is where AI becomes more than a time-saving tool; it becomes a way to multiply a physician’s knowledge without multiplying their hours. Doctor using AI versus doctors who aren’t. Now let’s compare two physicians. Same specialty, same number of physicians, same number of patients, same clinical ability. Doctor A is not using AI. Every note begins from scratch. Every patient handout is created manually, or the staff searches for one. Every portal response is typed individually. Every complicated article takes 20 minutes to read before Doctor A determines whether it matters. Every piece of administrative work begins with a blank page; Doctor B uses AI selectively, and I want to emphasize selectively. AI drafts the note. Doctor B reviews it. AI drafts the portal response. Doctor B corrects and approves it. AI turns the doctor’s explanation into patient education. Doctor B verifies it. AI summarizes a long research article. Doctor B goes to the source before making a clinical decision. Doctor B is not surrendering a medical judgment. Doctor B is eliminating unnecessary friction. That distinction matters because the physician who uses AI appropriately isn’t necessarily practicing medicine faster; they’re spending less time on everything surrounding medicine, and the difference compounds. 10 minutes here, 15. Minutes there, 20 minutes somewhere else. Pretty soon, we’re not talking about minutes anymore. We’re talking about hours. AI used number four, keeping up with medical literature. Here’s another struggle physicians know very well: the literature never stops. Studies, guidelines, drug updates, safety alerts, consensus statements, journal articles. Nobody can read everything.

 

Chapter 3 — AI as a Research Assistant, Not a Doctor

 

Dr. Barbara Hales 15:40

The AMA’s 2026 survey actually found that summarizing medical research and standards of care was the most commonly reported physician use of AI among the specific use cases measured. That tells us something important. Doctors aren’t only looking for help with paperwork; they’re using AI to help manage information overload. Suppose there are six recent papers on a subject relevant to your specialty. Instead of reading every word of every paper, AI can help you identify the research questions, study size, primary outcomes, where the findings agreed, and where they conflicted. What limitations should you pay attention to? Then, and this is critical, you go back to the original evidence when the information will affect patient care. AI becomes the map. It does not become the destination, and that brings us to something every physician listening needs to hear. AI is an assistant, not the doctor. AI can be wrong; it can misunderstand context. It can omit information. It can confidently produce something that sounds wonderful and is completely incorrect. And physicians know something that technology does not possess: clinical judgment. We know the difference between the textbook patient and the human being sitting in front of us. We notice hesitation. We notice fear. We notice when someone says, “I’m fine, while everything about them tells us they’re not fine. Medicine is not simply information. Medicine is interpretation. Medicine is judgment. Medicine is trust. So the goal is not how much medicine can I turn over to AI. The better question is, how much non-physician work can I take away from the physician? That is a very different conversation. In part two, I’m going to show you three more ways to use AI, including one that can help you create weeks of marketing content from something you’re already doing every single day, and I’ll tell you another story. This one is about a doctor who realizes the greatest thing AI gave him wasn’t efficiency. It was something much more valuable. It gave him back a Saturday. So here are seven ways doctors can use AI to get their time back. What would you do with the hours AI gave back? At the end of part one, I told you about a doctor who got something unexpected from AI-a Saturday. Let me explain. The Saturday morning that came back, we’ll call him Dr. Michael again. This is a composite story based on the kinds of workflow problems physicians experience. Michael owns a small specialty practice, and every Saturday morning he has a routine: coffee, laptop, kitchen table. He tells some. He tells himself he’ll work for an hour. It’s almost never an hour. Charts to finish, letters to dictate, patient messages, administrative correspondence, and things he didn’t get to during the week. His wife has learned not to ask when he’ll be finished. One Saturday, after his practice has begun using AI to help with documentation and routine administrative drafting, Michael wakes up. He makes coffee. He walks toward the kitchen table, and there’s no laptop sitting there. He had finished his work Friday. His wife says, “Do you want to go get breakfast? And for the first time in longer than he can remember, he does not say, “Maybe later. He says, “Sure. Now you could look at that and say, “AI saved him three hours. I don’t think that’s what happened. AI gave him back Saturday morning with his wife. We have to stop measuring physician efficiency only in minutes. We need to start measuring it in life. AI used number five insurance correspondence and administrative drafting. Now let’s talk about one of everyone’s favorite subjects: insurance. I can practically hear you groaning. Prior authorizations, appeal letters, medical necessity letters, supporting documentation, repetitive administrative correspondence. This is precisely the kind of structured, repetitive work where AI can assist. A physician or staff member can provide the relevant, privacy-safe information in an approved system and have AI help draft a first version. Then someone qualified reviews it. This isn’t about fabricating an argument. It isn’t about inserting facts that aren’t in the chart.

 

Chapter 4 — Administrative Work, Practice Operations, and Content

 

Dr. Barbara Hales 22:44

It is about organizing legitimate clinical information into the format needed for the task, and this matters because physicians did not train for all those years so that their most limited resource, their time, could be consumed fighting with forums; AI may not make the insurance company say yes. Wouldn’t that be nice? But it may reduce the amount of physician time required to ask the question. AI use number six: practical operations. Here’s one physicians often overlook: AI doesn’t have to touch clinical medicine to be useful. Look around your practice. How many repetitive processes exist? Staff training, frequently asked questions, office policies, meeting summaries, job descriptions, phone scripts, patient reminders, internal checklists, responses to online reviews, ideas for improving scheduling, drafting standard operating procedures. You don’t need a medical AI system to ask, “Help me turn this messy process into a checklist. You don’t need AI making clinical decisions to say, “Here are the five questions patients ask my receptionist every day. Help me draft clear answers. And here’s one I particularly like. At the end of your staff meeting, take your approved, non-sensitive meeting notes and ask AI to organize them into: what did we decide? Who is responsible? What’s the deadline? What needs follow-up? Suddenly, a meeting produces action instead of another meeting. Now we’re talking. AI use number seven. Mark. And patient communication. This one is particularly close to my heart. Doctors tell me, Barbara, I know I should create content. I just don’t have time, and I understand. But here’s what I want you to realize: you are already creating content all day long; you’re just not capturing it. Every patient question is potential educational content. Think about today: how many questions did patients ask you? Should I be worried about this? Is this hereditary? When should I get screened? Is there anything I can do to prevent this? What’s the difference between these two treatments? Those questions can become a short video, a blog post, a patient email, a special media post, a podcast topic, or an FAQ on your website-you provide the expertise. AI helps you repurpose it. For example, you might say, “Patients ask me this question constantly. Here is the explanation I usually give them. Then dictate your explanation and ask AI. Turn this into a 62nd video script written in my conversational style. Then give me three possible titles. Then turn the same information into a patient email. Then create five short social posts from it. One idea, multiple assets. That is where AI can become an extraordinary multiplier for a small medical practice, not because AI becomes the physician’s voice, but because it helps the physician’s real voice travel further. Here’s a story. The patient who finally felt heard. Let me give you one last story. We’ll call the physician Dr. Patel. Dr. Patel has always been the kind of doctor who looks directly at patients when they are speaking. At least she used to be. Then came increasingly complicated documentation requirements, and gradually she found herself doing what so many physicians do: patient talking, doctor typing, patient talking, doctor clicking, patient asking something important, doctor saying, “Give me one second. While finishing a field on the screen, then her practice begins using an ambient documentation tool. She’s still responsible for the note. She still reviews it, but during the encounter, something changes. She turns away from the computer. She looks at her patient, an older woman who has been seeing her for years, and suddenly stops in the middle of the conversation. Doctor Patel asks, “Is there something wrong? The patient smiles and says, “No, I was just thinking how nice it is to see your face again. That story is illustrative, but think about what it represents. We keep talking about AI as though technology will make medicine less human. Used badly, perhaps it could, but what if we use it to do exactly the opposite? What if technology handles more of the machinery of medicine, so physicians have more time for the humanity of medicine? That’s the opportunity that interests me. Now let’s turn to our two physicians, Doctor A and Doctor B. Remember, they started in essentially the same place, same specialty, same patient volume, same clinical competence. Doctor A continues doing everything manually. Doctor B gradually delegates appropriate repetitive work to carefully selected AI tools, not clinical judgment work.

 

Chapter 5 — Reclaiming Time, Using AI Safely, and Staying Human

 

Dr. Barbara Hales 29:57

Doctor B saves a few minutes. Documenting, a few minutes on correspondence, a few minutes researching, a few minutes creating patient materials, a few minutes handling administrative tasks. Let’s be conservative and imagine that adds up to only 30 minutes a day, five days a week. That’s two and a half hours, over roughly 48 working weeks. That’s 120 hours a year. Think of that-a 120 hours, three full 40-hour work weeks, and that’s from saving only 30 minutes per day. If the physician saves an hour, now we’re talking about approximately 240 hours, six work weeks. So ask yourself: What would you do with six weeks of your life? Would you see more patients? Would you leave the office earlier? Would you exercise? Would you finally take the vacation you’ve postponed? Would you have dinner with your family without your laptop? Would you sit with a patient for an extra three minutes instead of watching the clock? Would you grow your practice? Would you create another revenue stream? Would you simply sleep? There isn’t a correct answer. The point is, you get to decide. Start small. If you are listening to this and thinking, Barbara, this sounds great, but I don’t even know where to begin. Don’t begin with seven things. Begin with one. Tomorrow, notice the task that makes you say, “I can’t believe I’m wasting my time doing this again. Write it down. That’s your candidate. Then ask, “Could AI draft this? Could AI summarize this? Could AI organize this? Could AI turn this into a template? Could AI help my staff do this without involving me every single time? You don’t need to transform your entire practice this month. Choose one repetitive task. Get comfortable. Measure whether it actually saves time. Then choose another. That’s how change becomes sustainable. A word about safety. Because we’re physicians, I need to emphasize something very important: convenience does not outrank confidentiality. Don’t copy protected patient information into a public consumer AI tool. Your practice needs to evaluate privacy, security, HIPAA requirements, vendor agreements, accuracy, workflow integration, and appropriate human oversight, and every clinically relevant AI-generated output needs appropriate professional review. AI can generate extraordinarily convincing errors. So remember this: delegate the task, never delegate the judgment. That sentence may be the most important thing I say in this entire episode. If today’s episode made you look at your workday differently, do two things. First, choose one task this week that AI might be able to take partially off your plate. Don’t try to revolutionize your entire practice. Just reclaim 10 minutes, and second, subscribe to Marketing Tips for Doctors and share this episode with another physician who is still taking the laptop home every night, because every week on Marketing Tips for Doctors we talk about practical ways to build a stronger, more profitable medical practice without sacrificing the reason you became a physician in the first place. And remember, you’re one tweak away from greatness. AI won’t solve everything that’s wrong with medicine. It won’t eliminate insurance headaches. It won’t fix reimbursement overnight. It won’t replace the judgment that comes from years of training and experience. And I don’t want it to. What I want AI to replace is something very different: the unnecessary hour at the computer, the repetitive letter, the blank page, the administrative task that followed you home, the Saturday morning you were supposed to spend with someone you love, because perhaps the most exciting promise of artificial intelligence in medicine is not artificial intelligence at all. Perhaps it’s giving physicians the time to be More human, and if technology can help us do that, then I think it’s worth paying attention to. I’m Dr. Barbara Hales. Thank you for joining me on Marketing Tips for Doctors.

Till next time.