I was so moved by this remarkable presentation that I took several hours to transcribe it. The presentation – delivered without notes by Larry Weed, MD at Emory University’s medical grand rounds in 1971 while referring to a chart he borrowed from its clinic– is as valid now, maybe more so, than it was 50-plus years ago. Thanks to VisualDx and its CEO Art Papier, MD, who was a medical student of Weed’s, for posting the video on YouTube.
Lawrence Weed, MD was a professor of medicine and pharmacology at Yale in the 1950s when he developed the concept of the problem-oriented medical record and the universally used SOAP note (subjective, objective, assessment, plan) for documenting patient care, which he incorporated into an electronic system. His dry humor comes through as he addressed his white-coated audience in Atlanta, much as it did in a JAMIA oral history where he described telling a surgeon who bristled at the idea of a computer adding value to his experience and intuition, “I’m not saying that you don’t have intuitive feelings. What I’m suggesting is that they may be worthless.”
Larry Weed died in 2017 at 93.
This is a long transcript of the nearly hour-long presentation, but I guarantee it’s worth reading.
It’s good for me to be here. I don’t know whether when I’m done you’ll think it’s good that I was here. But I can be a real hit-and-run driver. I don’t have to stay here, and if I’m lucky, I won’t get sick in Atlanta, because once you antagonize somebody and then you get sick in front of them, you’re never quite sure whether you’re safe or not. But from what TV says, you shouldn’t get sick in America anyway.
What I’d like to do is go at this problem, not from the point of view of the record. We really aren’t taking care of records — we take care of people. We’re trying to get across the idea that this record cannot be separated from the caring of that patient. This is not the practice of medicine over here and and the record over here. This is the practice of medicine. It’s intertwined with it. It determines what you do in the long run. You’re a victim of it, or you’re a triumph because of it. The human mind simply cannot carry all the information about all the patients in the practice without error. And so the record becomes part of your practice.
When you hear someone say, “I know lots of good practitioners [unintelligible] surgeon. He doesn’t keep any records at all.” How do you know he’s so good? Because he talks fast? Because he operates fast? Then we have to see the record. How many gastrectomies has he done? How many haven’t had infections? What do the wounds look like? How many dumping syndromes? How many have died? How many got infections?
Without a record, we’re not going to believe him. It’s like asking somebody his golf score three years ago without it written down. It’ll always come out better than it actually was.
We’ve got to look at the record. If this represents practice, you’d walk into a place like this and say, “I’d like to see how medicine is doing here. People say it’s a very good place or a very poor place.” How do you know?
Let’s see a patient’s problems. I picked this [chart] out of your clinic this morning. If you tell me what the problem is and I have a few minutes, I’ll either know myself or I’ll go to books or I’ll go to experts. We’ll say, “That’s pretty good standards for that problem.” But I’d have to know the problem well.
The first sheet is a little sheet here that says Oakland the Avenue or something. Then, phenobarbital addiction. Impression: probable addiction. Then a scribble here. Then there’s a blue sheet. I keep leafing through all this stuff. Then there’s an extra report — normal brain scan. Now they really didn’t do that for phenobarbital addiction, I don’t think. I’m leafing through this and I say, geez, I’d like to know where the problems are. You’d say, “Come on now. Dr. Weed. Pull yourself together. Let’s not try to make a big thing out of this record business just because you happen to be interested in records.”
You know, I’m interested in nucleic acid chemistry. I’ve been a biochemist a lot longer than I’ve been fussing around in clinical medicine. It’s not that I’m so interested in records. I’m interested in medicine. I had to use these to find out what was going on, and it’s got me absolutely climbing the wall. I could set it aside like I used to and say, “Never mind the record. I’ll tell you all I know about pyelonephritis.” But that doesn’t have anything to do with her. That’s Grand Rounds on me. That isn’t what you’ve come for.
So I’d say, “I’d like to know the problems.” You say, “They’re at the end of the workup. Find the first workup and you’ll find the problems.” So I come to here and I read through this impression: CVA. Number Two, extreme anxiety neurosis. Was that all the problems? All right, that’s all the problems. We can see how you diagnosed it and what you did for it. We’ll see if that’s good care for CVA.
I’m combing through here and it says blood pressure 180 over 98. Thorazine. They’re giving the Thorazine for a stroke? No, they’re giving that for the anxiety, maybe, I’m not quite sure. Then what’s all this SSKI? Then here is LE preps times three. For anxiety, or a stroke? Then x-rays of the left hip and the pelvis. Now you might say, “Don’t get excited. She probably fell out of bed.” Did she? I don’t know. Urinary tract infection. Honest to God, now they’re x-raying the left shoulder and the left hand next. Next impression, same patient — chronic obstructive lung disease. Personality disorder.
Then I go to the lab sheets, and you know what I see? PBIs, BUNs, serum sodiums.There’s a whole bunch of electrolytes in a row. Now you don’t do serum sodiums and all those electrolytes every day for a stroke. You don’t do them for anxiety. There must be another problem.
Now I can’t audit it. I don’t know. I don’t know whether you’re giving good care to the problem. I don’t even know whether you’re finding all the problems. If Problem Number One is hypertension and Problem Number Five is depression and you’re giving amitriptyline for Problem Number Five, the depression, that’s all right in itself. But that antagonizes guanethidine, and if you stop the amitriptyline and then they up the dose of the guanethidine, she has shock, hits her head on the bathtub, and she comes in here for a subdural hematoma, that’s your fault. But I can’t find it because it’s too hard to interpret this.
You can’t look at the management of a single problem without knowing the context. What are all the problems? Yes, she should have the hip pin, but not today, because Number One is heart failure. Yes, she should have fluids restricted, but Number Three is azotemia. Yes, she should have lots of fluids. It’s tough. It’s tough, and you shouldn’t have to spend a second finding what are all the problems.
Now what kind of a record have we got here? We’ve got a source-oriented record. It’s not a problem-oriented record. What I mean by that is this. In a source-oriented record, you put all the lab data together. You put all the x-rays together. You put all the temperature sheets together. You’ve got all the nurse’s notes together. You have all the doctor’s notes together. I say, “I’d like to know what you’re doing for the lady’s ear.” Well, there’s the temperature. Then I read the notes about the doctors and the progress notes. If you read some of these progress notes, doing well, home tomorrow. Phenobarb. Acute arthritis. Shoulder swollen. ECS 600 milligrams. RTL. ABC XYZ . It flows. It’s a single paragraph. The elbow, the urinary tract. It’s a series of things.
That’s the doctor, then you have the white counts and the serum sodiums and the urines together. Then you have the x-ray of the ear with the x-ray of the chest with the x-ray of the hip. I say, what are you doing that for? Why do you put the x-ray of the ear with the x-ray of the hip? What’s the ear got to do with the hip? Well, nothing, but we like to put those together. Why did you have all those ear cultures with those urine cultures? Is she urinating in her ear? What are you doing that for?
When people source-orient data, you appear very unscientific. But it’s even worse than the appearance. It’s the very essence of the practice of medicine. This is not an idle discussion of little technical bookkeeping details. The practice of medicine is the way you handle data and think with it. The way you handle it determines the way you think. Once you get over a period of time with multiple variables, the very structure of the data determines the quality of the output. This is what’s so hard for medicine to accept. They can’t say things like, “I know lots of good doctors don’t keep good records.” They can’t be separated that way.
You might say, you could figure out what’s wrong with that ear if you wanted to. You could sit down and read the whole record. With 200 million people, to get quality, I might spend three hours. But even then, I couldn’t do. It it would be impossible because I would be guessing. I’d come to the order sheet. This order sheet has all these orders, and I’d see penicillin on it. I wouldn’t be sure whether that’s the urinary tract infection, the pneumonia, or the ear. It doesn’t say. I see brain scan. I could guess that maybe that’s for the stroke, or I don’t know, or maybe he’s worried about a subdural or maybe he’s worried about a tumor. I don’t know, I’m guessing.
Then you might say, why don’t you call up the doctor to ask him? I’ll say, but that note was two months ago. You can’t be serious. You don’t say to a teller in a bank, “Do you remember Mr. Jones who came in two weeks ago? How much money did he put on the shelf?” If she answered you, you’d think she were crazy. If I had a technician read the spectrophotometer on Wednesday and write the numbers in the notebook on Thursday, I would fire her and I would get complete support in a basic science faculty. But in a medical faculty, writing discharge summaries three weeks later? Operative notes, preoperative notes after the operation? Somebody writes all his progress notes on Sunday morning? That’s fiction, it’s not science. Better that you not write anything at all than something that’s not absolutely reliable.
A source-oriented record is essentially useless from the point of view of a rigorous audit. And mind you, if you can’t audit a thing for quality, it means you do not have the means by which to produce quality. They’re inextricably entwined. If you can’t evaluate what you’re doing, then there’s a very serious possibility that you do not know what you’re doing, and that you’ve never defined your goal. That’s true in medicine. We have not.
You hear clinicians say, “Good clinicians always problem-orient their records.” Oh no they haven’t. No, they haven’t.
Let’s look at the four phases of medical action. This is the database. That’s the first phase of medical action. After you get some information, you’re going to formulate the problems, so you’re going to make a problem list. After you’ve got a problem list, you’re going to have a plan for each. Then you should follow each, and those should be titled and numbered progress notes.
You say, that’s what good clinicians have always done. They’ve done a history and a physical and lab work. That’s what the database is. Then they’ve made a list of impressions. We call it impressions, he called it problems, no difference there. Then we had a plan. We wrote in the order book, then we wrote progress notes. What’s so new about this?
Let’s take each one separately. Let’s take that that database. Was that ever defined? You know that that problem list is determined by that database. If all you know is her name, she doesn’t have any problems. If you know a name and her blood pressure, you may have one. If you know a name and her blood pressure and do a pelvic, you may have two. I would walk in here and I’d say, what’s the guaranteed database for these patients? The intern does a history and physical. You know what that means. Some ask five allergy questions, some ask 55. Some ask five if they have one admission, they ask one if they have five admissions, and they have none if they have seven admissions.
So the problem list is determined by where he trained, what he’s interested in, how many people came in last night, what the professor asked for, we have a rheumatologist for the attending. That’s not the way to run a shop. If you want a guaranteed list of problems and deal with the problems in context, you’d better define the database. You should get it every time. If you can’t get a complete database on people that have nails in their foot, you say fine, for these complaints like a nail in the foot, a broken arm, a person with a penile drip or gonorrhea, or something in his eye, we get what we call a mini-database. We do not get the whole database, but we give episodic care, but if you have a nail in the foot, we’ll ask these three questions, we’ll feel for your lymph nodes in your groin or something, we always give the TAP.
We do that much, but for comprehensive care, for someone with hypertension or weight loss or headaches, we do this. We have branching logic questions. We always do this on physical we get for this age groups. We get this lab work for the 40-year-old. We will get triglycerides and such and such. X-rays are done with this frequency. Paps are done in this frequency. People would say, we wouldn’t have time to do all that. Well, then get somebody else to do it. You’ve got to set the goal and then stick to it, and if you don’t have time to do it, get paramedical people to do it. They’ll do it better anyway and they’ll write it up so you can read it.
For instance in our clinics, as I was telling the the house officers this morning, “We don’t have time. It’s awful busy in the clinic. We don’t have time to get pelvics.” They have an excuse for everything. So what did I do? When they came in the front door, when they register, we have their age, they’re female, they’d go to the fifth floor, we’d trained nurses. They did the pelvic, the rectal, the belly exam, the breast exam, the thyroid exam. It was done in an organized way, checked off, they were checked out by the professor of OB/GYN.
You know as well as I do that they found much more than was being found in the medical clinic. In the medical clinic, they either didn’t get a pelvic, and half the time the fellows that did, they might as well been sticking their fingers out the window. You know that as well as I do. They were never cleared in an organized way on these problems, whereas when we taught the nurses, we ran them through 50, we checked them out in a systematic way – yes, she’s competent. We took no risks on box number one, getting a database. You use computers with branching logic questions. They can take them home. You use Mark-10 sheets. You can use interviewers, take these things and put them through a Selectric typewriter, through the business office computer, do whatever you have to do, but get the database and get it every time.
We found that with a questionnaire that had 32 questions, we got the vital signs, and we did it with paramedical personnel, it took between nine and 11 minutes. We found that the doctors were missing an average of 5.2 problems per patient, and some of them were quite serious. They were seeing in the patients what they wanted to see. They played Sherlock Holmes too early. They would ask one question and the next question was being determined by the first question, because that’s the way they were brought up in a CPC sort of an atmosphere. What do you think up next, doctor? Let’s put two men on the chess board. Move one and we will decide where to put the others on. Oh no you won’t, you’ll put them all on, we’ll look at the rules, and then we shall start to play.
It’s very arbitrary how much data you get before you start to think. Under pressure, if you let people get data in a Sherlock Holmes way, they get so they get less and less data, have more and more intuition, draw conclusions more and more prematurely, and get people into more and more trouble. Always saying they don’t have time.
In nine minutes, you can find out a fantastic amount of information if you will just do it. Just do it. Don’t think, just do it. People say, yeah, but it’s so arbitrary. I know it’s arbitrary, people. Everything is arbitrary. A football field is arbitrary. It could be 150 yards long, it could be 75, it could be 100. But if you do not draw the line, you will not play the game and you won’t how you’re doing.
Suppose I’m running down the field and I fall down on the 15-yard line. I get up and say, that’s a touchdown. By whose definition? My own — I’m tired today. That sounds absolutely ridiculous, but that’s the way we practice medicine. Best ENT man in town — that’s his database, here and here [gestures to ear and mouth]. You say, he never gets in trouble. Oh, maybe once out of 100 times. He took that lady up, took her larynx out, she happened to be in bad heart failure. She died of something, he didn’t know she was hypertensive. You realize, I’m sure, that the Lord and the chiropractors can get 85% of these people better. The only reason you run these fantastic establishments is to get that other 10%. The only reason you have a professor of medicine is to pick up that final 2%.
We know it’s arbitrary, but you must define it, and once it’s defined, once you realize when you fall down on the 15-yard line and say “that’s a touchdown,” I say no, you don’t get the score, you’ve got to over that line. Well geez, don’t I get some credit? I’ve been fighting all the way down this field for an hour, they battered me up tonight, I’m exhausted, and they were awful big guys that were on top. This guy down there the at the medical clinic, he doesn’t realize how big this is. I’m sorry, buddy, you don’t go over the line, you don’t get credit.
Once that’s very clear in your mind and the object is to get the data, you’ll figure out ways. You’ll learn to think of forward passes and you go home and study up new plays, because we’re not going to change the game just because you’re tired. You begin to improve this profession, but this profession truly is a cottage industry, everybody wandering around defining his own game. And when you’re allowed to define your own game, you’re a fool not to define it in a way that you’re always the victor. Of course the medical profession gets the appearance of being arrogant and independent. Anybody that’s been allowed to define his own game all his life, that’s conducive to arrogance. He never has a defeat. He’s always got a way out. She was too sick. She went sour. What does that mean?
That’s the first phase. Look at the second phase here. You’re going to get a problem list from the database. You say, we’ve always made a list of impressions. Did we ever do that any better than in the database? In the first place, we use the word “impression.” That was a terrible thing to do in the first place. If you use the word impression, or what you think, you then have to have the person who wrote the chart with you when you interpret the chart, because what he’s thinking is part of it. I’m not interested in what the impression is. I’m interested what you know to be the problem, and no ambiguity about it.
Occasionally people say, I don’t know, geez, Larry, everything’s black and white to you. You just put a number on things. I don’t know whether it’s rheumatic heart disease or a cardiomyopathy. I said, what do you know? What do you know? Do you know the diagnosis? No. Do you know a physiological finding, like heart failure? Yeah, I know she’s in heart failure. Then that’s your problem, Doctor, that’s your problem. If you knew it was rheumatic heart disease, put it, that’s your problem. We might say, if you want me to be absolutely honest, I don’t even know that. Well, what do you know for sure? Is it a symptom or a physical finding? Yes, I’m very sure she’s short of breath. I can guarantee you, Doctor, if I take you in the room there, you may not agree it’s rheumatic heart disease and you may not agree it’s heart failure, but you’ll have to admit she’s very short of breath. I don’ t know whether it’s chronic obstructive lung disease or cardiac failure. She may have piece of corn caught in her trachea, for all I know. Well, that’s the problem.
You might say, I picked up his chart he’s got, question mark, organic heart disease. I saw that in one of those clinic charts. I say, what’s the problem? He doesn’t remember, he had a lot of admissions last night – oh yeah, she’s that one with the funny cardiogram. Doctor, that was pretty risky. You never want to lose sight of the problem. If you hadn’t been here and I had to use that chart, I could spend an hour trying to find out what the problem was, and if I had to go through your laundry basket to find that EKG, that’s very risky. He says – this was a new intern – it seems logical to me, but you don’t honestly want me to put down Problem Number Four, funny looking EKG, now now do you? I haven’t yet had a course in cardiology. I just don’t know anything. I just don’t think we can be expected to know everything.
I said, yes, Doctor, if that’s the level at which you understand the problem, put it down that way. If that’s the level at which your care is being given, there’s nothing to be ashamed of about that. There’s no reason why you should know all about cardiology. An ophthalmologist doesn’t, and you don’t know all about ophthalmology, either. The neurologist doesn’t know all about endocrine disease. All you have to do is be honest. Then I’ll say to you, what’s funny about it? The dumbest person, instead of putting down “funny cardiogram,” will take one second say, what’s funny about it? Look at the reading — those are U waves that shouldn’t be there. So that problem is Number Three, U waves in cardiogram.
You can call up anybody in seconds in the middle of the night and say, “would you see this patient?” What’s the problem? You read off the problem say, “Number Three is U waves.” He’ll say, get a potassium, do this, I’ll be in. After 30 seconds, he’s with you. But if you say going out the door, “I got a lady that I think has a little heart disease, would you see her? I’m going to see if I can operate tomorrow.” You can go upstairs and pick up this record you can be glommering through it for half an hour and still not be sure what’s bothering you. What you usually do is throw it on the desk and say to the nurse, show me the patient. You wander down, look her over, and from your experience, you sort of think she’ll live through that operation. You hold your breath and clear her and that’s the end of it, usually. It should be much more precise than that. We haven’t defined problems. We’ve put down impressions.
Now what else about the problem is it wasn’t kept up to date. I stumbled over the fact of arthritis, LE preps, hip x-rays, PBI, CVAs. This lady has at least 10 problems. What happens when I stumble over it in this way? I’m scared to death there’s some that I’m missing. So then when I find a couple of important ones, then I panic, I go back, and now I’ve got to read every word, because they’re scattered in the middle of pages and the end of lab sheets. Whenever a new problem appears, it should go on that problem list, and that problem should sit on the front of the record and it should be absolutely up to date. We have to be ruthless with the doctor, not who does the right or wrong thing for heart failure – only God is right or wrong for some of these problems, it’s very confusing — but you have to be ruthless with him if he does not keep the problem list up to date so that anyone in seconds can be in context and make intelligent decisions.
Over and over again, I didn’t know, postoperative, everything’s going wrong. Then we discover the old record that she’s had hypothyroidism and no one gave her the thyroid pills and we got mixed edema coma up on the operating table. That’s inexcusable, and it happens in every hospital in this country. Lymph nodes didn’t get cultured. I didn’t know. I didn’t know. I couldn’t get time to go to the operating room doctor. It can’t be you that takes care of a patient alone. When you see a head resident running around at night to see all the patients that came in, that’s fraudulent. He can’t possibly be the intern for that many people.
He either has a system he trusts or he’s going to lose. When he’s the most most ruthless were the people under him is when they violate the system, because the people are in the hands of the system.
When someone says “I take care of that patient. I’m her doctor.” that’s fraudulent. No one points to a Pontiac and says, “I made that car.” A system makes that car. And even in the pre-Flexner days, no one could take out cataracts, handle porphyria, diabetic acidosis, perforated ulcers, depression. No one ever did all those as well as they could be done. Of course we want specialization, and if you have a gall bladder problem that’s really tough, of course you want Cattell to sit up there and do it at the Lahey Clinic. He’s done thousands of them. He was magnificent. You want that if someone takes out your mother’s cataracts. You want the feeling that he’s done hundreds of them. He does them magnificently with minimal chance of failure. But yet if he’s that good and focuses, will he be able to encompass everything? No, he will not.
So you want people to be part of the system. You don’t want a family practice program where you teach them to be superficial. A system builds automobiles and it’s going to take care of people, and if we don’t recognize that simple fact, then there’s going to be an awful lot of people that are not going to get cared of and there will be a 100,000 people in the middle of Chicago floating around Cook County getting less than adequate care. It’s like Henry Ford saying, “I personally am going to make an automobile for everyone the population. I don’t believe in systems and assembly lines. I’d rather have my personal touch on it.” He’ll make an automobile for two or three people a year. The other 200 million will have none, and that is the basis for a revolution. This is the basis for a system. The record has to be it. You can write a check in Atlanta on a New York bank, why? Because there’s a system, and it’s known throughout. But you get a coronary in Atlanta and your records are in Chicago, just try to find things out in the first 24 hours, because even if you call up and they got it out of the record, the girl in the record room is in the same position I was. “What did you want to know, Doctor?” and there she is. You probably wouldn’t be able to get her on the phone, but if you did, you’re just wasting your money. She should be able to read off that problem list just like that.
What about the next phase of medical action, where we talk about the plans for each problem? What have we done there? We’ve taken an order sheet and just scribbled orders – penicillin, BUN, side rails, phenobarb, serum sodium daily, IVP tomorrow. You say, do you think that those sodiums every day are necessary? I don’t know, I don’t know what you did them for. I suppose if you had some fancy endocrine disorder, aldosteronism or something you were fooling with, that might be sensible. If it’s for flat feet, that’s a waste of money. No one should ever be able to write an order without coupling it with a problem.
When you write plans, you have to think of them in three phases. You put down Problem Number One, hypertension. You’ll find you get a whole new spirit out of your nurses once you start dealing with problem-oriented records. They become part of the team. They know everything they’re doing. They know how one doctor does it differently. They ask why. They become more and more sophisticated. Before, they were asked to go blindly. They went down and gave the penicillin. They didn’t know what it was for. They couldn’t see if you’re being consistent, whether you agree with other infectious disease people.
It was like walking into a room and people were throwing darts, and you say, where’s the target? Wherever the dart lands. This is the arrogance with with some doctors treat nurses. It’s no fun for the nurse. If you have a target up there and I come in here throwing darts, anyone will stop for a minute to see if you hit it. It’s a challenge. It’s tough on you if you miss it all the time with that many people watching, but that’s what education is all about. You learn to improve after a while. Either that your you’ll stop publicly taking care of people, and that’s that’s an advantage too in some cases.
Under any problem, under A, what are you going to do first? This is where you get more information. This is why you should always think about plans for problems. For more information, for why. This is where your rule-outs go. I’m going to rule out unilateral kidney disease. How? Right there, I want the rule-out. Are you going to do it with a timed IVP or renogram or what are you going to do? Aldosteronism, and if so, you go do one serum potassium, then do five, you’re going to do with the dividing line going to be 3.8, 4.3, 2.2? High-salt diet, low-salt diet. Let’s precisely decide now before we spend your money.
When you see doctors on work rounds saying, “She had a little problem here, but you know, she was on a diuretic, I don’t think I’m going to worry too much about that,” that thinking should have been done before you drew it. When you see people thinking of what to do, how to do it, and how they’re going to interpret in 30 seconds at the end of a bed in a random fashion, it’s like a contractor saying, all these architect plans, let’s lay a few pipes here. The pipes of the john can’t come up in the fireplace. You can’t do that. Fortunately, as I say our house officers, you know now the sickest kidney is brighter than the brightest intern. I mean, it’ll it’ll sort your IVs no matter what you do.
But anyway, you’re going to put your rule-outs. You see I’ve taken those rule-outs away from the problem list. The problem should not have rule-outs, question marks, or probables. It should be a precise, reproducible statement of the problem at the level you can understand it and guarantee it, no matter how unsophisticated you have to get.
We’ve got to fix the system so that students are much more ashamed of being imprecise and dishonest than they are of being unsophisticated. They should never worry about whether they remember or whether they’re sophisticated. All they have to be ashamed of is that I miscommunicated. I overstated the case. I misstated the problem. As Bernanos says, the worst, the most corrupting of all lies, is to misstate the problem. Keep it pure, and then in your plans under more information, we’ll go your thinking and your logic. There’s your differential diagnosis. But don’t let it get mixed up with the problem until you can update your problem in a secure way.
I pick up charts and one I picked up today had infectious disease, question mark on a problem list. What does that mean? In our place, I picked one up the other day. It says Problem Number Five, rule out diabetes. I said, what’s the problem? He couldn’t remember whether it was the urinary tract stricture. Is she the one with polyneuropathy? He says, you don’t realize how busy this is. I said Doctor, never lose sight of the problems. I said, do you see what you would have done? You put “rule out diabetes.” You do glucose tolerance. The resident says, how’s that glucose tolerance? You say two hours was 115, fine, we cleared that up, she can go home.
Cleared it up? She never had it. You see, the problem is still vaginitis or neuropathy. Another plan was rule out diabetes, and when that’s normal, then what are you going to do next? You’ll find, if you do this rigorously, that over half the time, on half the problems, you will never resolve it. You’ve got to learn what Whitehead talks about, this capacity for a sustained muddle-headedness, a tolerance of ambiguity. Pavlov said you must teach a graduate student gradualness. He must never be forced to overstate his position, misstate his problem. Good medicine is a careful, rigorous inching your way towards a more and more secure position. A final diagnosis is a myth. There are never two cases of lupus the same. There are no absolute final criteria. You must define them, but recognize that it’s the evolution and the following of a patient that’s going to make the difference within these explicit definitions.
What’s B? What you’re going to do to treat?
Never mix what you’re going to do to treat with what you’re going to do to get more information now. You say, we never would do that, but yes you do. I’ll see in treatment when I pull a chart and separate and I see cholorothiazide here and they were getting urine sodiums for this aldosteronism. As I separate data, oh my God, you see that what they were doing is that they were getting more information on what they did than on what they had. We do that all the time. We do so much to a patient. We give them so many drugs, and so many procedures, and so much psychiatric confusion that when we do our tests, it’s really more information on what we’ve done to her than what the original problem was. The intern gets her in here, she’s got all this diarrhea, and the next thing you know he’s doing stool analysis on barium.
Let’s see under any plan what you’re going to do to tell the patient. In no place in American records do we have an organized approach to what we’re going to do tell the patient. Under that hypertension, did you tell her it was serious, or not serious? How you’re going to study it, or whether you’re not going to study?
All right, now let me quickly get from the plan to the progress notes. Never in American medicine have we had highly structured progress notes in a problem-oriented way, where we had a complete problem list and we numbered every problem progress note with respect to the problem. We’re in this box for now. Never write “doing well.” What does that mean? She’s got arthritis, heart failure, azotemia, broken hip, and ear infection. You put “doing well.”
What you mean is I’m a cardiologist, they asked me to look at it, I did, we said digitalize her, I came in today, I listened, the rales are going away, the edema is less, she lost a few pounds, gallop’s gone away, rhythm’s a little slower, rate’s a little slower, I think she’s doing well. I didn’t know she had glaucoma, these urinary tract infections annoy me, and I never worked up a broken hip in 20 years anyway. That’s what he means. Someone said to James Thurber, how’s your wife? In what way?
When you put the problem, you put 1. Hypertension, and then always write symptomatically and objectively your quick interpretation and your plan for the next step. Always give the patient’s point of view first, then what objective data you have, and where you going to go. You’re taking each problem in depth. Then you look back at the complete problem list and look at them in context. What does this mean? It means you can write a plan for azotemia, you can read Strauss’s book, or you can write one for heart failure [unintelligible], and you can know broken hips, but you can’t write a book for Mrs. Jones, who’s got this much heart failure, that much azotemia, and a broken hip. She’s absolutely unique. Eighty-eight keys on the piano and a million symphonies.
There is no absolute treatment for anything. You can lay out your plan explicitly, set up your flow sheet, and then look at it make a move, like in the chess game. Watch nature’s move, then make another move. You’re a guidance system. If you know those satellites, they get up there and they land here on Wednesday or Thursday by this battleship. What are they doing? They’re taking their position every instant with four computers on this system. They keep readjusting their course.The shape of the path is not precisely known until the input stops it. It doesn’t need to be known, but you have to take your parameters of guidance, how often to look, and you readjust.
When you go from Atlanta to Seattle, you never go the same way twice. Sometimes you go to the northern route, the southern route, and even if you try to go the same way twice, you could not. There’s a red light here, before there was a green light. The bridge is washed out in Chicago. You read detour signs, you go around. There’s a tornado in Montana. You meet a nice girl in, you know, Oregon. You call up your family, you’re going to be a couple days late. The car breaks down.
But you know all the principles — red lights, green lines, detour signs, automobile maps — and once you believe that boy has learned to go from Cleveland to Seattle, you don’t give them a special course now to go from Cleveland to Hong Kong and Cleveland to San Francisco. You expect that he’ll choose parameters, the same physiological ones over and over again, whether you having a fluid balance problem from heart failure or a bad burn or an intestinal obstruction or diabetic acidosis. You’re going to make a volume decision and a free water decision and an acid-base decision and a potassium decision. It’s the same heart and kidneys. The agent that threw you off is a little bit different, but the commonality of it all has got to be seen by the student or he’ll memorize and memorize and memorize and then collapse or just distort.
In biochemistry, we get so upset with those flip-flop circuits when we do the counting. You wouldn’t mind so much if they worked or didn’t work at all, but the trouble is they have worked, and they sort of stopped counting in the middle the night, but you wouldn’t know it and you’d report something that was absolutely ridiculous.
That’s the way interns and residents and doctors are. If they worked perfectly, it’d be one thing, or if they didn’t work at all. But we half work. We half guess. We half understand. We half know. But we’re never uncertain about going to the order book and writing a drug. No matter how busy a doctor is, I’ve yet to have a doctor say to me, I was so busy I didn’t have time to order anything. He’s always so busy he didn’t have time to find out anything, but he’s always got time to order something.
What I’m saying is that a doctor has to be a guidance system. He is not an oracle that knows answers. Once he accepts the concept of being a guidance system, then he knows that the data system is the basis from which all his work must take place. Then the record suddenly becomes an unbelievably important document in education, in care, and in research.
But as long as we were a profession that thought we could rest on the memory, and it what you know makes a difference instead of what you do, and as long as we thought of doctors as oracles that know answers instead of guidance systems in uncertain situations, we were willing to let the record in American medicine fall to this level. Now the computer people move in, and the Medicare people move in, and the non-medical people move, in they can hardly believe what they see. There is a crisis of major proportions.
The first hospital I went where we decided to stop source-oriented records for problem-oriented records, the first thing I did was we canceled all the conferences and all the rounding that wasn’t directed to specific problems. From 6:00 in the morning till midnight seven days a week, we got these things so the database was defined, the problem lists were up to date, the plans were in order, and there were progress notes at the right frequency.
It’s just like a fifth-year graduate student in biochemistry. When it gets down to the time and he’s got to get out, he stops going to hear the Nobel Prize winners, he stops all in these conferences, he stops running around, he isn’t going on with dinner with his wife, he isn’t chatting in the hallway. He’s doing one thing — he’s in that laboratory and he’s working on that thesis. This [the chart] is a doctor’s thesis. He may get an A in all his courses, but if his thesis is no good, it’s not creative, the work was no good and it’s unreliable, all those A’s in advanced calculus and biochemistry mean nothing.
The same way with us. If this is not interpreted or auditable and the quality is not good, then all the rounds, all the specialization, all the NIH, and all the hierarchy mean nothing. Mean nothing. You might say, but don’t you think we should have research? Yes, I think we should have research, but this is research. Every patient’s different. We don’t have to run away from the bedside to be profound and to be unbelievable scientists.
I worked a long time in nucleic acid chemistry and I can tell you from my own experience that there is nothing that taxes you intellectually or taxes your sense of science and philosophy the way this situation does. Basic scientists who’ve been allowed to focus on one or two problems and keep their data in a separate notebook and come in from eight to five and shut off the incubator when they want to go away, they have no concept of what it is to have five problems per patient, 30 patients on the ward, 150 problems running simultaneously.
They never taught a data system for that. But because it’s so difficult and because it’s just in its infancy of what we could do doesn’t mean it’s unscientific or that it’s unsophisticated, and when someone says, geez isn’t it more sophisticated to get new knowledge or go to the NIH and work with Nierenberg and find a new nucleotide or work out the code, those are pieces. They’re sophisticated, they’re worthwhile. I don’t begrudge him his Nobel Prize. I’m glad to have these things happen. I worked on these I got more more money and more job offers and more professorships offered from biochemistry than I ever did out of clinical medicine, and I don’t begrudge people. It was very useful to me at the time.
But to say that to sit up in the attic carving the chess men and writing the rules, as the universities have done for 2,000 years, to say that’s more sophisticated than playing the game with those men, that’s ridiculous. It’s unbelievably sophisticated to take those men and play the game. You don’t need to stop making the chess men. We don’t need to burn down the NIH or stop the research laboratories to go on to this more sophisticated state of playing. The students should never think of that clinic with hundreds of patients and all this confusion and how to get the database. That’s a very sophisticated problem in systems analysis, in efficiency. In order to decide if you’ve got 100 patients to see there this afternoon and you’ve got to get the database that will yield the most, which problems do you want? You’ve got to know a great deal about the science of medicine. If we’re after heart failure, is best to grab a venous pressure, add five questions on the history? What is the highest yield? If it’s hyperthyroidism, should we talk to her about diarrhea, weight loss? Should we grab PBIs? If there are 10 things you could do, which have the highest yield? What do they really know about hyperthyroidism anyway? Let’s turn it to play this game.
You’ll find that they haven’t really thought about it very rigorously. They just have the pieces. They’re going to put them together tomorrow, but tomorrow never came. Pusey said, but isn’t the university to discover new knowledge? Of course the university is to discover new knowledge. But the new knowledge we need now, and it’s most difficult and most sophisticated, is how to use knowledge. That’s that’s a very profound thing to do. This [the chart] is the physical representation of doing or not doing it. It runs head-on with society. It’s very easy to go down to the molecular level and work on trinucleotides or triphosphates or anything else, or to go out in outer space where only your methods can measure how badly off you are. In biochemistry, we used to say we have the microsomes and the mitochondria, That’s a pure amount of mitochondrial prep, we’d say. We thought it was until the methods got better and we could see it with an electron microscope and see it full of junk, microsomes, all sorts of junk in there.
In other words, the purity of these isolated systems is only good insofar as your capacity to find the faults, whereas when you work in that clinic, it’s at the macro level. It’s not so distant and so macro that you can’t see it. It’s not so micro that you can’t see it. It’s patients moving around. It’s like a big cell here. Instead of mitochondria, there’s patients and doctors and pharmacies. There’s a nucleus. It can be centrifuged and separated, put together and studied, but the reason we don’t like to do it is because your faults are so obvious. Your mistakes are so obvious. The lack of purity of your approach is so obvious. You can’t stand it, so you say it’s unimportant or it’s not scientific or that’s not why I came into medicine. We’re cowards. It’s perfectly clear that’s what the problem is. Society is unreasonable. It’s frustrating. It’s irrational. The cell was, too. The centrifuge was, too. Those mitochondria were, they weren’t pure prep. The only difference was is they couldn’t talk back and we couldn’t see it and we didn’t devise methods to see how badly off we were all right now.
Let me make one closing remark about what this has all got to do with the art of medicine. Where is the art of medicine going to go with all this if you if you have lists and numbers, for art is style, structure, form, discipline. It’s Andrew Wyeth making Jamie Wyeth do the painting 50 times until it’s right. Unbelievable discipline about technique. He made that boy tear up a painting 100 times. It’s George Szell, if you have ever watched him with that orchestra. The same passage 30 times until it was perfect, and no violinist stood up and said, this is interfering with my art. Nor did Bach say, three beats in every measure? That interferes with my creativity. No, art is Hemingway, three weeks on a single paragraph. It’s Bach recording in detail everything he did a couple hundred years ago so we can hear it today.
It’s not a scribble in the middle of the night. It’s not saying, “I took good care of her,” leaving absolutely no trail for us to ever find out whether you did or did not. We debase the word art itself when we call what we’ve been doing art. And it’s not science. We have to be extremely careful when we defend what we’re doing. We don’t reveal to others that we didn’t even get out of a liberal arts education, as Stravinsky says, that art is nothing more than placing limits and working against them rigorously, and if you refuse to place them and try to work within them but just flail about, you do not have art, you have chaos. That’s to a large extent what we’ve had.
I hear, and personally experience instances where the insurance company does not understand (or at least can explain to us…