User login
“We really aren’t taking care of records — we’re taking care of people.” — Dr. Lawrence Weed
What is the purpose of a progress note? Anyone? Yes, you there. “Insurance billing?” Yes, that’s a good one. Anyone else? “To remember what you did?” Excellent. Another? Yes, that’s right, for others to follow along in your care. These are all good reasons for a progress note to exist. But they aren’t the whole story. Let’s start at the beginning.
Charts were once a collection of paper sheets with handwritten notes. Sometimes illegible, sometimes beautiful, always efficient. A progress note back then could be just 10 characters, AK, LN2, X,X,X,X,X (with X’s marking nitrogen sprays). Then came the healthcare K-Pg event: the conversion to EMRs. Those doctors who survived evolved into computer programmers, creating blocks of text from a few keystrokes. But like toddler-sized Legos, the blocks made it impossible to build a note that is nuanced or precise. Worse yet, many notes consisting of blocks from one note added awkwardly to a new note, creating grotesque structures unrecognizable as anything that should exist in nature. Words and numbers, but no information.
Thanks to the eternity of EMR, these creations live on, hideous and useless. They waste not only the server’s energy but also our time. Few things are more maddening than scrolling to reach the bottom of another physician’s note only to find there is nothing there.
Whose fault is this? Anyone? Yes, that’s right, insurers. As there are probably no payers in this audience, let’s blame them. I agree, the crushing burden of documentation-to-get-reimbursed has forced us to create “notes” that add no value to us but add up points for us to get paid for them. CMS, payers, prior authorizations, and now even patients, it seems we are documenting for lots of people except for us. There isn’t time to satisfy all and this significant burden for every encounter is a proximate cause for doctors despair. Until now.
A fully formed, comprehensive, sometimes pretty note that satisfies all audiences. Dr. Larry Weed must be dancing in heaven. It was Dr. Weed who led us from the nicotine-stained logs of the 1950s to the powerful problem-based notes we use today, an innovation that rivals the stethoscope in its impact.
Professor Weed also predicted that computers would be important to capture and make sense of patient data, helping us make accurate diagnoses and efficient plans. Again, he was right. He would surely be advocating to take advantage of AI scribes’ marvelous ability to capture salient data and present it in the form of a problem-oriented medical record.
AI scribes will be ubiquitous soon; I’m fast and even for me they save time. They also allow, for the first time in a decade, to turn from the glow of a screen to actually face the patient – we no longer have to scribe and care simultaneously. Hallelujah. And yet, lest I disappoint you without a twist, it seems with AI scribes, like EMRs we lose a little something too.
Like self-driving cars or ChatGPT-generated letters, they remove cognitive loads. They are lovely when you have to multitask or are trying to recall a visit from hours (days) ago. Using them, you’ll feel faster, lighter, freer, happier. But what’s missing is the thinking. At the end, you have an exquisite note, but you didn’t write it. It has the salient points, but none of the mental work to create it. AI scribes subvert the valuable work of synthesis. That was the critical part of Dr. Weed’s discovery: writing problem-oriented notes helped us think better.
Writing allows for the friction that helps us process what is going on with a patient. It allows for the discovery of diagnoses and prompts plans. When I was an intern, one of my attendings would hand write notes, succinctly showing what he had observed and was thinking. He’d sketch diagrams in the chart, for example, to help illustrate how we’d work though the toxic, metabolic, and infectious etiologies of acute liver failure. Sublime.
The act of writing also helps remind us there is a person attached to these words. Like a handwritten sympathy card, it is intimate, human. Even using our EMR, I’d still often type sentences that help tell the patient’s story. “Her sister just died. Utterly devastated. I’ll forward chart to Bob (her PCP) to check in on her.” Or: “Scratch golfer wants to know why he is getting so many SCCs now. ‘Like bankruptcy, gradually then suddenly,’ I explained. I think I broke through.”
Since we’ve concluded the purpose of a note is mostly to capture data, AI scribes are a godsend. They do so with remarkable quality and efficiency. We’ll just have to remember if the diagnosis is unclear, then it might help to write the note out yourself. And even when done by the AI machine, we might add human touches now and again lest there be no art left in what we do.
“For sale. Sun hat. Never worn.”
Dr. Benabio is director of Healthcare Transformation and chief of dermatology at Kaiser Permanente San Diego. The opinions expressed in this column are his own and do not represent those of Kaiser Permanente. Dr. Benabio is @Dermdoc on X. Write to him at [email protected].
“We really aren’t taking care of records — we’re taking care of people.” — Dr. Lawrence Weed
What is the purpose of a progress note? Anyone? Yes, you there. “Insurance billing?” Yes, that’s a good one. Anyone else? “To remember what you did?” Excellent. Another? Yes, that’s right, for others to follow along in your care. These are all good reasons for a progress note to exist. But they aren’t the whole story. Let’s start at the beginning.
Charts were once a collection of paper sheets with handwritten notes. Sometimes illegible, sometimes beautiful, always efficient. A progress note back then could be just 10 characters, AK, LN2, X,X,X,X,X (with X’s marking nitrogen sprays). Then came the healthcare K-Pg event: the conversion to EMRs. Those doctors who survived evolved into computer programmers, creating blocks of text from a few keystrokes. But like toddler-sized Legos, the blocks made it impossible to build a note that is nuanced or precise. Worse yet, many notes consisting of blocks from one note added awkwardly to a new note, creating grotesque structures unrecognizable as anything that should exist in nature. Words and numbers, but no information.
Thanks to the eternity of EMR, these creations live on, hideous and useless. They waste not only the server’s energy but also our time. Few things are more maddening than scrolling to reach the bottom of another physician’s note only to find there is nothing there.
Whose fault is this? Anyone? Yes, that’s right, insurers. As there are probably no payers in this audience, let’s blame them. I agree, the crushing burden of documentation-to-get-reimbursed has forced us to create “notes” that add no value to us but add up points for us to get paid for them. CMS, payers, prior authorizations, and now even patients, it seems we are documenting for lots of people except for us. There isn’t time to satisfy all and this significant burden for every encounter is a proximate cause for doctors despair. Until now.
A fully formed, comprehensive, sometimes pretty note that satisfies all audiences. Dr. Larry Weed must be dancing in heaven. It was Dr. Weed who led us from the nicotine-stained logs of the 1950s to the powerful problem-based notes we use today, an innovation that rivals the stethoscope in its impact.
Professor Weed also predicted that computers would be important to capture and make sense of patient data, helping us make accurate diagnoses and efficient plans. Again, he was right. He would surely be advocating to take advantage of AI scribes’ marvelous ability to capture salient data and present it in the form of a problem-oriented medical record.
AI scribes will be ubiquitous soon; I’m fast and even for me they save time. They also allow, for the first time in a decade, to turn from the glow of a screen to actually face the patient – we no longer have to scribe and care simultaneously. Hallelujah. And yet, lest I disappoint you without a twist, it seems with AI scribes, like EMRs we lose a little something too.
Like self-driving cars or ChatGPT-generated letters, they remove cognitive loads. They are lovely when you have to multitask or are trying to recall a visit from hours (days) ago. Using them, you’ll feel faster, lighter, freer, happier. But what’s missing is the thinking. At the end, you have an exquisite note, but you didn’t write it. It has the salient points, but none of the mental work to create it. AI scribes subvert the valuable work of synthesis. That was the critical part of Dr. Weed’s discovery: writing problem-oriented notes helped us think better.
Writing allows for the friction that helps us process what is going on with a patient. It allows for the discovery of diagnoses and prompts plans. When I was an intern, one of my attendings would hand write notes, succinctly showing what he had observed and was thinking. He’d sketch diagrams in the chart, for example, to help illustrate how we’d work though the toxic, metabolic, and infectious etiologies of acute liver failure. Sublime.
The act of writing also helps remind us there is a person attached to these words. Like a handwritten sympathy card, it is intimate, human. Even using our EMR, I’d still often type sentences that help tell the patient’s story. “Her sister just died. Utterly devastated. I’ll forward chart to Bob (her PCP) to check in on her.” Or: “Scratch golfer wants to know why he is getting so many SCCs now. ‘Like bankruptcy, gradually then suddenly,’ I explained. I think I broke through.”
Since we’ve concluded the purpose of a note is mostly to capture data, AI scribes are a godsend. They do so with remarkable quality and efficiency. We’ll just have to remember if the diagnosis is unclear, then it might help to write the note out yourself. And even when done by the AI machine, we might add human touches now and again lest there be no art left in what we do.
“For sale. Sun hat. Never worn.”
Dr. Benabio is director of Healthcare Transformation and chief of dermatology at Kaiser Permanente San Diego. The opinions expressed in this column are his own and do not represent those of Kaiser Permanente. Dr. Benabio is @Dermdoc on X. Write to him at [email protected].
“We really aren’t taking care of records — we’re taking care of people.” — Dr. Lawrence Weed
What is the purpose of a progress note? Anyone? Yes, you there. “Insurance billing?” Yes, that’s a good one. Anyone else? “To remember what you did?” Excellent. Another? Yes, that’s right, for others to follow along in your care. These are all good reasons for a progress note to exist. But they aren’t the whole story. Let’s start at the beginning.
Charts were once a collection of paper sheets with handwritten notes. Sometimes illegible, sometimes beautiful, always efficient. A progress note back then could be just 10 characters, AK, LN2, X,X,X,X,X (with X’s marking nitrogen sprays). Then came the healthcare K-Pg event: the conversion to EMRs. Those doctors who survived evolved into computer programmers, creating blocks of text from a few keystrokes. But like toddler-sized Legos, the blocks made it impossible to build a note that is nuanced or precise. Worse yet, many notes consisting of blocks from one note added awkwardly to a new note, creating grotesque structures unrecognizable as anything that should exist in nature. Words and numbers, but no information.
Thanks to the eternity of EMR, these creations live on, hideous and useless. They waste not only the server’s energy but also our time. Few things are more maddening than scrolling to reach the bottom of another physician’s note only to find there is nothing there.
Whose fault is this? Anyone? Yes, that’s right, insurers. As there are probably no payers in this audience, let’s blame them. I agree, the crushing burden of documentation-to-get-reimbursed has forced us to create “notes” that add no value to us but add up points for us to get paid for them. CMS, payers, prior authorizations, and now even patients, it seems we are documenting for lots of people except for us. There isn’t time to satisfy all and this significant burden for every encounter is a proximate cause for doctors despair. Until now.
A fully formed, comprehensive, sometimes pretty note that satisfies all audiences. Dr. Larry Weed must be dancing in heaven. It was Dr. Weed who led us from the nicotine-stained logs of the 1950s to the powerful problem-based notes we use today, an innovation that rivals the stethoscope in its impact.
Professor Weed also predicted that computers would be important to capture and make sense of patient data, helping us make accurate diagnoses and efficient plans. Again, he was right. He would surely be advocating to take advantage of AI scribes’ marvelous ability to capture salient data and present it in the form of a problem-oriented medical record.
AI scribes will be ubiquitous soon; I’m fast and even for me they save time. They also allow, for the first time in a decade, to turn from the glow of a screen to actually face the patient – we no longer have to scribe and care simultaneously. Hallelujah. And yet, lest I disappoint you without a twist, it seems with AI scribes, like EMRs we lose a little something too.
Like self-driving cars or ChatGPT-generated letters, they remove cognitive loads. They are lovely when you have to multitask or are trying to recall a visit from hours (days) ago. Using them, you’ll feel faster, lighter, freer, happier. But what’s missing is the thinking. At the end, you have an exquisite note, but you didn’t write it. It has the salient points, but none of the mental work to create it. AI scribes subvert the valuable work of synthesis. That was the critical part of Dr. Weed’s discovery: writing problem-oriented notes helped us think better.
Writing allows for the friction that helps us process what is going on with a patient. It allows for the discovery of diagnoses and prompts plans. When I was an intern, one of my attendings would hand write notes, succinctly showing what he had observed and was thinking. He’d sketch diagrams in the chart, for example, to help illustrate how we’d work though the toxic, metabolic, and infectious etiologies of acute liver failure. Sublime.
The act of writing also helps remind us there is a person attached to these words. Like a handwritten sympathy card, it is intimate, human. Even using our EMR, I’d still often type sentences that help tell the patient’s story. “Her sister just died. Utterly devastated. I’ll forward chart to Bob (her PCP) to check in on her.” Or: “Scratch golfer wants to know why he is getting so many SCCs now. ‘Like bankruptcy, gradually then suddenly,’ I explained. I think I broke through.”
Since we’ve concluded the purpose of a note is mostly to capture data, AI scribes are a godsend. They do so with remarkable quality and efficiency. We’ll just have to remember if the diagnosis is unclear, then it might help to write the note out yourself. And even when done by the AI machine, we might add human touches now and again lest there be no art left in what we do.
“For sale. Sun hat. Never worn.”
Dr. Benabio is director of Healthcare Transformation and chief of dermatology at Kaiser Permanente San Diego. The opinions expressed in this column are his own and do not represent those of Kaiser Permanente. Dr. Benabio is @Dermdoc on X. Write to him at [email protected].