I’ve been doing 12 hour shifts at a regional medical center that serves a predominantly blue collar community in Washington. My job as hospitalist involves showing up at 7AM on a Monday and meeting and taking care of about 16 new patients on the medical and surgical services as well as admitting a few more, usually from the emergency department. My shift ends at 7 PM or whenever I’m actually done. Mondays are big, but I do get to meets lots of people. The first two transitional days involve reviewing lots of data and I become gradually less dependent on my predecessor’s clinical judgement. It would be ideal for patient care if the physician who admitted the patient stuck around until discharge, but that would logistically mean that physicians worked every day, all day and all night. Nobody would sign up for that, and if they did, they would be toasted to a crisp very soon and patient care would suffer. Transitions are inevitable, but always involve increased patient risk. Sometimes it’s good that care transitions, and I pick up something that the previous doctor hadn’t noticed. Sometimes it takes awhile to understand the rationale behind treatment and sometimes I probably drop the ball without even noticing it. Eventually everything is rolling along nicely and then I go off service, usually 7 days after starting. Signout routines for the organization that employs me are good. Not only does the physician write assessments and plans in the patient’s chart, but a more casual and concise signout is part of the organization’s website and is kept up to date by routine to help the night physicians with care and for the change of service. All places are not like this. Some have terrible or non-existent signouts, and the places that are smaller and have 24 hour shifts (like my own community hospital) have face to face signouts that are really excellent. But that wasn’t actually what I wanted to talk about.
I have been doing more and more ultrasounds as part of my examination routine as a hospitalist, with the little GE Vscan ultrasound that I keep in my pocket. Using “point of care” ultrasound in the hospitalist setting is a slow starter. I read an article about training hospitalists to do ultrasound of hearts that was dated 2007, concluding that they weren’t as good as echo technicians after doing 35 ultrasounds. The authors were surprised. Why would they be surprised? Learning to do an echocardiogram well, obtaining pictures and interpreting them, takes more than 35 ultrasounds, and echo technicians do at least hundreds before considering themselves competent. Hospital physicians do use ultrasound pretty often for doing procedures involving placing needles in internal structures, but apparently not often for diagnosis. Many emergency physicians use ultrasound frequently and there is quite a bit of precedent for this, but once the patient is admitted to the hospital, bedside ultrasound for making diagnoses becomes less common. In the two hospitals I work in there are no other physicians other than intensivists routinely using ultrasound they do themselves for diagnosis. Which is too bad, because it is really fun and really good for the patients. I now ultrasound everybody I examine, unless I have already done it and there is no conceivable reason to think that they have changed.
Here are some examples of how my little ultrasound machine makes a huge difference in how effective (and happy) I am. Echocardiography is an obvious application. A few days ago I admitted a man who presented with fatigue and a fast heart rate. His EKG showed atrial fibrillation at about 150 beats per minute. His rate slowed somewhat with a diltiazem drip but then he began to feel worse. My ultrasound said that he had a severely weakened left ventricle, probably from prior heart attacks, and so diltiazem was a bad drug for him. I could adjust his medication appropriately knowing what his physiology was, not waiting for an official echo. He got his official echo eventually, but he was also treated appropriately right away. I could also see with a lung ultrasound that he was in congestive heart failure, which helped guide therapy as well, since I could follow that along day by day to look at how adequate treatment had been.
Then there was the 92 year old woman who came in with a hip fracture from a fall the previous evening. It was unclear how much hydration she would require before surgery, but my bedside echo showed that her cardiac function was normal and that her inferior vena cava was very decompressed and collapsible, consistent with dehydration, so she needed lots of fluid. And I could watch that vena cava with hydration to see when her tank was full. So nice to know rather than guess.
And there was the 88 year old man whose foley catheter I had had removed so he could go home. But he didn’t pee. He said he didn’t feel the need. My handy dandy ultrasound is also useful below the navel, and I could tell that his bladder was about the size of Texas, so he needed the catheter back. The nurses have a version of the ultrasound called a bladder scan, but I was right there and could show the patient the picture without delay, which was good for him since he could tell what I was worried about.
This was a big week for pancreatitis. Pancreatitis can be caused by gallstones, but in our population it is much more likely to be caused by drinking alcohol to excess. Still, you can be a drunk with gallstones. The 49 year old drinker with pancreatitis got his cardiac ultrasound to document how well he would tolerate vigorous hydration, plus I could see his gallbladder which was healthy looking and stone free. I could also see that he didn’t have free fluid in his abdomen which would have suggested more severe pancreatitis or coexistent cirrhosis of the liver. But even better than all that, I could show him his organs, which he loved, and made him much happier, more cooperative and appreciative. That was incredibly cool, since apparently appreciation wasn’t his usual MO.
And then there was the guy with diverticulitis who had had more pain the night before I came on service, then felt better by the time I saw him. He was bloated, with no bowel sounds, but not all that tender, not all that sick seeming, and with a normal white blood count. I ultrasounded his heart–couldn’t see a thing, not a single thing, all air. Then I checked his intestines, because diagnosis of small bowel obstruction by ultrasound is really not very difficult. Couldn’t see a thing. When you can’t see a thing what you are really seeing, barring that the machine is broken, is air. Air looks like a gray screen, like an old tv when the programming was all over. Not ever having seen absolutely nothing, I was stymied and ordered an x-ray, which confirmed that I had in fact been seeing air, inside the belly, that is to say free air from an intestinal perforation. Now I know. Air everywhere is free air and means a perforation (or recent laparoscopic surgery, but that would be obvious.) The patient had emergent surgery and is doing pretty well.
I have come to appreciate the larger hospital I am working in for various things, most importantly that they make absolutely no fuss about treating patients who are uninsured. They do over 14 million dollars in charity care per year, and the primary physicians and specialists all treat these patients without whining about not being paid. A hospital can’t run in the red for very long, though, so it is paying customers, in some way or another, who finance those who can’t pay. This hospital runs in the black, but is a not for profit organization, so money left at the end of the year gets funneled into either community programs or capital improvements. They donate generously to the local YMCA which has all sorts of popular wellness programs. I’m not sure what all else they do, but it is clear that promoting health in the community for the poorest people can impact the likelihood of those people needing expensive hospital care. This model seems like a good one for making natural consequences for a hospital favor good health for a community. This incentive only applies to patients who have no or low paying insurance, since hospitals will always thrive on the well insured who have repeated need for high cost and well reimbursed care. If hospitals were paid per capita to care for the insured, they would also have a good reason to attempt to keep these patients well, which would mean even more effort spent on prevention and health promoting programs and probably more aggressive research into high value and cost effective inpatient care.