Philip Greenspun’s Weblog

We meet at 7:00 am in the “Anesthesia Ready Room”, a small office with only three computers shared by the 50-person anesthesia private practice. I was excited to be assigned to follow Dr. D, who is widely respected by surgeons. He’s a pale 38-year-old sporting wide circular glasses. We go into OR 17 to set up for the case, a left total knee arthroplasty. He shows me how to pull medications by inserting a needle attached to a syringe into the rubber sealed glass vial. We go over how to pull doses of common anesthesia medications, e.g., succinylcholine, rocuronium, propofol, etomidate. “As a general rule, the right dose is usually half of the vial. That was true during my training, but people have gotten a tad larger now so maybe round up.” He adds, “The exceptions are these.” He points to a nondescript compartment in the anesthesia cart where epinephrine, phenylephrine, and ephedrine are stored. “I’d say the most common medication that I see get overdosed with serious consequences is phenylephrine (“Neo”). It’s meant to be made into a drip (“gtt”, latin for guttae or drops) by inserting the dose into a 100 mL bag for pressor support over time. A typical rate is 100 mcg/min. We use it by giving 200 mcg as a one-time dose for pressor support between induction and the operation beginning.”

“One of the hardest things to do as an anesthesiologist is to do nothing. The patient’s blood pressure drops when we induce the patient. A lot of people want to reflexively fix that with a pressor like ephedrine or phenylephrine,”. Dr. D continues. “Watch how much the blood pressure increases once we start intubating and cutting on the patient.”  We put the pulled medications (in syringes with attached needles) into a locked drawer on the ventilator workstation. He then asks if I have intubated before and with which laryngoscope blade. I respond that I’ve intubated twice, and have used only a “Mac” (MacIntosh laryngoscope features a curved blade versus the straight Miller laryngoscope). “I don’t understand people who use Miller. It’s forcing the pharynx to be a straight configuration when it’s clearly not.” He grabs a “7 French” endotracheal tube (ETT) and Mac 3. I ask, “Should I get a stylet?” (Stylets are a stiff malleable metal wire that is inserted into the ETT to help guide it through the cords.) He laughs, “Stylets are not necessary. It’s $11 that goes down the drain.”

We head to pre-op to consent a 58-year-old for general anesthesia and potential blood products. She is about to undergo a total knee arthroplasty (replacement) with general anesthesia. He asks the patient standard anesthesia questions: any dentures? Previous anesthesia, and if so, any problems? The patient answers that she has terrible post-operative nausea. He asks her to open her mouth wide, and touch her top lip with her bottom teeth (assess range of motion of jaw). She signs the forms, and then the anesthesiologist performs an adductor canal nerve block under ultrasound guidance. A pre-op nurse is dedicated to assist with these nerve blocks. Looking for a good lifestyle job in healthcare? The nerve block nurse’s only job is to wheel an ultrasound cart over and hand the anesthesiologist a pre-op needle, which means she’s busy only about 30 percent of the time and then goes home at 2 pm. We head back to the OR.

After a few minutes waiting in the OR, the nurse wheels the patient in. I clumsily attach the pulse oximeter, three electrode lead stickers, and blood pressure cuff. I mistakenly attach the “left leg” lead, supposed to go under the left breast, on a layer of gown. “Honey, this isn’t my breast!” I grab another sticker and place it below her breast and attach the electrode. He supervises me closely as I preoxygenate the patient with a mask delivering 100 percent oxygen. He then pushes the propofol into the IV and we wait a few seconds. He instructs, “gently touch her eyebrows to assess if she’s asleep.” The first time her eyelid twitches but after five more seconds she is fully induced. 

“Let’s see how you mask ventilate. This is probably the most important skill in anesthesia. If you can mask a patient you can relax.” I adjust the pressure in the circuit with a knob as I squeeze a 2 L bag with my right hand and use a “C” grip on her mask and jaw to elevate the jaw. “Really squeeze at the angle of her mandible. Get in that crease.” I am squeezing the mask with my thumb and index finger in a “C” shape to rotate it against the jaw that is being pulled up and out to move the tongue forward and open the airway. The first breaths I hear a leak as I squeeze the bag with my other hand, but adjust my grip. I look for chest rise, fog in the mask on exhalation, and finally check the end tidal CO2 mass spectrometer waveform on the anesthesia workstation.

As I mask the patient monitoring the end-tidal CO2, tidal volumes, and peak pressures (keep under 20 mmHg to prevent air from going into the stomach), the attending asks me, “We know we can mask the patient, so we can relax. Why do we pre-oxygenate?” I answer with a snarky response, not understanding the full significance of the question. “Umm, to get more oxygen in them.” He responds with a smile, “Yes, but why does it work?” He explains that at the functional residual capacity (FRC, the amount of air in the lungs when exhaled such as when paralyzed) there is 2 L of air, and therefore 400 mL of oxygen (20 percent at 1 atm). Your body uses about 35 mL oxygen per kg per min, which varies depending on the health of the individual and food intake.” For a typical 70 kg person, that is 35 mL oxygen per second so I would only have about 10 seconds before the patient starts to desaturate if we intubate without preoxygenation. He summarizes, “Pre-oxygenation is really denitrogenation of the air.” Instead of only having 400 mL oxygen, by masking the patient with 100 percent FiO2, the patient now has 2 L of oxygen, allowing for about a minute before the patient begins to desaturate. In reality once you preoxygenate, you have several minutes before you need to worry.” After another few breaths, he states, “You’re ready. Give it a try… Wait. Take your badge off, you don’t want to cause a corneal abrasion. Those hurt more than a kidney stone!” He pushes the Rocuronium, a paralytic agent that will prevent the patient from moving any somatic (voluntary) muscle.

After I transfer my badge, notebook, and pens from the front scrubs pocket into my pants, I grab the Mac blade, scissor the upper and lower teeth with my thumb and index finger to open the jaw, and place the blade in horizontally, avoiding the teeth. My goal is to first locate the right tonsillar pillars, then sweep the tongue with the blade over to the left, and insert the blade deeper until I identify the epiglottis. I struggle to find the epiglottis in this patient, and hand it over to the attending who quickly inserts the tube. “Good technique, that was a hard airway. Better luck next time.” He adds, “Careful when removing the blade. A chipped tooth comes right out of my paycheck. $10,000, no question.” 

I auscultate both lungs (listen with a stethoscope) to confirm good airflow, look for fog in the tube, and confirm end-tidal CO2 for a few breaths. He laughs as he’s already given a few bag breaths and turned on the ventilator setting to pressure control (ventilator mode that delivers volume up to a set amount of airway pressure). “It’s good to do that as a habit, but I saw the tube go through the vocal cords, I see good fog in the tube, and consistent end-tidal CO2. Confidence in medicine is key, especially in surgery. Confidence comes from experience. Now, relax. You’re welcome to grab some breakfast or coffee. We’ll just be sitting here until the case ends.” (In search of a stooI, I wander the OR hallways until an OR nurse guides me to the stash. I lug the stool through the maze of gadgets in the OR suite while trying not to touch anything sterile.)

Dr. D has a dedicated chair. We sit together and look at our phones (no need to wear gloves unless touching the patient) for two hours while occasionally talking about subjects he thinks will be helpful, e.g., his ABCDEFGHI mnemonic for taking over a patient.  Airway (confirm placement of airway), Breathing (look at end-tidal CO2, peak pressures, pulse oxygen saturation, arterial blood gas, if applicable), Circulation (blood pressure and heart rate, EKG), Drips (ensure medications are in, including pain, nausea, etc.), Effluent and Fluids (ensure good urine output and adjust IV fluids accordingly), Heat (don’t forget the Bair Hugger), and Injury (make sure the neck is in a neutral position, the eyes are not in contact with anything, the arms are not under pressure and pronated; he explains to me that the ulnar nerve is the most common injury during anesthesia. A lot of time the surgeon is not present when positioning the patient, he explains this is a bad habit because the surgeon is just as liable for any positioning injury as the anesthesiologist). The surgeon tells us that they are starting to close as he unscrubs and steps out of the room to allow his PA to suture the skin and dress the patient.  

[Editor: “Doctor Says a Device He Invented Poses Risks” (New York Times, December 24, 2010), “Two decades ago, Dr. Augustine, an anesthesiologist in Minnesota, helped pioneer the idea of keeping a patient warm during surgery. Doing so, studies have shown, produces benefits like less bleeding and a faster recovery. Dr. Augustine’s invention, the Bair Hugger, changed surgical practices and made him a fortune. The device, which works like a forced-air heater, carries warmed air through a hose to a special blanket that is draped over a patient. These days, Dr. Augustine asserts that his invention is a danger to surgical patients receiving implant devices like artificial heart valves and joints. The forced air, he says, can spread bacteria associated with hospital-acquired infections. Coincidentally, Dr. Augustine, who no longer has a financial stake in the Bair Hugger, also says he has a safer alternative, a warming device that works more like an electric blanket and does not use forced air.” A twin-size electric blanket is $25 on Amazon in 2020. Why did 3M pay $810 million for a company making this hot air system? Why wouldn’t hospitals just buy blankets from Amazon and throw them out after each surgery? Ordinary blankets are used in the OR, washed, and reused.]

We turn off the anesthesia gas and reverse the rocuronium with sugammadex. The anesthesiologist explains as we wait for the patient to wake up that rocuronium has become the standard paralytic because of this new reversal agent. “It was a brilliant move by the pharmaceutical company. If you don’t use the reversal agent, the patient could feel short of breath and weak for several days.” He adds, “You also don’t want to get that call from a lawyer when the post-op patient has respiratory distress.  ‘Why didn’t you reverse her?'” When the patient starts to move her arms, and cough, I deflate the ETT balloon cuff, and pull the tube out. The anesthesiologist pushes the bed into “steer” mode and carts her off to the post-op recovery room, jumping on the frame and riding the bed around every turn as he waves to the OR staff. I struggle to keep up with him. He backs the patient into the post-op room, and “gives report” to the nurse. We head to pre-op (right next door) to get ready for the next

I ask to focus on abdominal CT during this final week of radiology, which turns me into an expert on finding steatosis (fatty liver, caused by alcohol, viral hepatitis, or obesity). It’s easy to identify because it’s on every abdominal and pelvic CT! Prasanna’s partner explains how to differentiate steatosis from fibrosis/cirrhosis by the liver morphology. CT can differentiate interfaces between air (-1000 Holmzfeld units), fat (-100), water (0), soft tissue (30), and bone (1000). The liver and spleen should be similar densities. As fat infiltrates the hepatocytes, however, the liver density begins to drop below 60. NASH (non-alcoholic steatohepatitis) is expected to surpass alcohol-associated liver failure to become the number one cause of liver transplantation. The radiologist explains, “The liver edge should be smooth. Once fibrosis occurs, it becomes nodular. Cirrhosis is also associated with enlargement of the caudate and left lobe. You can do a ratio, but just eye-ball it.”

The abdominal work list is exploding. The attending yells, “Six scans come off one scanner in one hour. Six abdominal scans. Why couldn’t they have interweaved some lumbar and head CT? Come on! I can do maybe 5 CTs per hour. Six from one scanner, and this seat covers five other scanners. We are just getting slammed.” As soon as he is done venting and has refocused on another case, his phone rings. “Come on!”

We overhear Prasanna yell, “God Dammit.” We walk over to investigate. Prasanna waves me in. “This is the MRI from the hip arthrogram we did earlier today. What do you see?” Based on irregular T2 signals with T1 replacement (bone marrow destruction) and articular cartilage flattening, I answer, “O-N.” Osteonecrosis is bone death, typically due to reduced blood flow. He tries to pull up the X-ray. “They didn’t get a f***ing X-ray. This is why you always get an X-ray first. This could have been diagnosed weeks ago instead of waiting for an MRI. He’s going to lose both hips.” I ask what caused this? “O-N can be caused by lots of things. Osteomyelitis (or infection of the bone) is one, but I don’t think both his joints are infected. It could be from long-term steroid use, inflammatory conditions, congenital abnormalities, and trauma. There’s a whole differential. Sometimes it’s just idiopathic [unknown cause].”

We do a leg bone length study on a 13-year-old. Children who suffer a broken leg can end up with one leg growing dramatically longer than the other. We measured from the top of the femoral head to the top of the talus. “The truth is orthopedists do their own measurements, so I don’t get too technical. Each has his or her own favorite method. Some old school private practice orthopaedists keep their radiographs in-house. I do all this for our health system billing and CYA. They need our help for MRIs and CTs.” Prasanna asks, “What do you think the most common lawsuit is for orthopaedists that keep radiographs in house? … Missing lung cancer on a shoulder X-ray.”

On Friday, I work with a guy who finished radiology training only three years ago. “This seat [MSK] is so boring that it erodes my soul.” He drones “Normal” into PowerScribe after every X-ray, which allows ample time to discuss the coronavirus: “I don’t think people realize what is coming. The virus is reported to have almost a 20 percent infection rate. On the cruise ship, one asymptomatic person infected 600 people. Our health system covers about 1 million people. We have 54 ICU beds. The numbers just don’t work.”

[Editor: This is late February 2020, about two months after the media began intensive coverage of COVID-19. As it happens, the hospital never did run out of ICU beds. The hospital filled up completely in January 2021, but mostly because patients couldn’t be discharged to their nursing homes so long as they tested positive for COVID-19. See “Our hero’s hospital is full (but not with patients who should be there)” in which I noted “Essentially, the hospital is packed because, even with nearly a year to prepare, state and local health departments that regulate hospitals and track hospital capacity couldn’t get organized to turn empty hotels into Covid-19 halfway houses.”]

The junior radiologist continues: “On top of this will be a supply crisis. Our health system reverts back to the medieval age when we don’t have common medications. Penicillin is not made in the US anymore. There is going to be a huge shortage of needles. China supplies everything, and they are shut down.” Is he stockpiling? “Oh yeah.” He grabs another coffee, his fifth today. “Let me get caught up.” He speeds through 10 radiographs in a few minutes, dictating with prefilled phrases. He turns to me. “The three fastest radiologists I have ever seen are all here. The fastest offered to do 1.5 lists and get paid at 1.5 FTE. I can see his point because he could handle it, but it would set a dangerous precedent if all you care about is speed. His offer was rejected, so he started the medical student clerkship. We’re not all as fast as him, so we fall behind when students are here.”

We review a pelvic CT. He laments, “Look at this! Hip pain. It doesn’t specify if the pain is in the hip joint, greater trochanter, or SI joint. No clinical history. I’m so used to it, but this lack of communication hurts the patient. Help me help you! The worst is when we get an abdominal/pelvic CT for ‘abdominal pain, unspecified’.” He continues, “Epic has made this communication crisis worse. The ED doc or PCP just clicks a worthless button and moves on. I can use Epic to read the doctor’s notes, but I shouldn’t need to do that. The MSK seat is not as bad as the abdominal seat as there are far fewer potential diagnoses.”

Statistics for the week… Study: 0 hours. Sleep: 9 hours/night; Fun: 2 nights. Example fun: Dog playdate at a local park followed by a dog-friendly brewery.

The rest of the book: http://fifthchance.com/MedicalSchool2020

A two-week elective in MSK (musculoskeletal) radiology. The private practice radiology group that staffs our health system’s department offers 4 hours per day for medical students. I’ll be working in a large newly constructed clinic building from 9:00 – 11:00 am and 1:00 – 3:00 pm each weekday.

A typical day is as follows: I meet the attending on the MSK seat at 9:00 am. Precise Prasanna, a 39-year-old MSK fellowship-trained radiologist, is walking on the treadmill and dictating a shoulder MRI. He stops the treadmill to chat for a few minutes. Having arrived at 8:00 am, he is caught up on the worklist and has “parked” five interesting cases from this morning. He leaves the room for five minutes to refill his water and chat with his colleague on the abdominal seat while I go through them on the diagnostic monitor screen. 

I report what I have observed: “There is a high T2 signal in the right superior hip labrum.” He replies, “Good, look at the cam deformity [enlarged femoral head knocks into the acetabulum] causing femoral acetabular impingement.” He continues, “FAI is now known to be the most common cause of early osteoarthrosis. We see this all the time in female soccer players. A 10-year-old presents for anterior hip pain from a labrum tear. When you ask them they sometimes report their hip ‘stopping’ but kids get used to the impingement feeling. Twenty years ago we would have forgotten about it; now we realize FAI causes early OA so we intervene before destruction of the cartilage.” FAI can be diagnosed on a simple AP pelvis radiograph.

(Arthritis includes the suffix “itis,” suggesting inflammation, but most arthritis is due to wear and therefore osteoarthrosis is the preferred term.)

He points out the interesting aspects of 20 X-rays and 6 MRIs before it is time for live patients. We do three arthrograms, in which contrast agent is injected into the joint space under X-ray guidance. Most commonly, this is with gadolinium contrast in preparation for an MRI to fully assess the hip or shoulder labrum. Sometimes, this is to get better information from a patient who is not a candidate for an MRI. For example, Prasanna performs a shoulder arthrogram on a 28-year-old female bicycle accident victim whose implanted hardware following a previous humeral head fracture (motor vehicle collision) would distort the signal from susceptibility artifact. He points to the leaking of contrast from the joint space into the subacromial/subdeltoid bursa (fluid-filled cushion underneath tendons), indicating a full-thickness tear of the supraspinatus tendon. 

After lunch with Jane and our new puppy, I return for the 1:00 pm session. The radiologist in the abdominal seat calls me over to look at a CT scan of the chest and abdomen. “What do you see?” I respond, “There is a clear hypointensity disrupting the bright signal of the right pulmonary artery. Is this a pulmonary embolism?” He answers, “Yeah, I just sent her to the hospital. I don’t see any right heart strain. That’s all!”

Prasanna dictates reports with PowerScribe, voice recognition software specific to radiology. Every word he uses serves to further delineate the pathologic process. He explains to me that the main goal of an musculoskeletal radiologist is to pick up subtle findings of a pathologic process, e.g., rheumatoid arthritis, psoriatic arthritis or severe meniscus tears, before it severely damages the articular (hyaline) cartilage. Once destroyed the joint is unsalvageable and must be replaced (arthroplasty). “For some diseases we can stop the inflammation with drugs or for some mechanical injuries an orthopedist can operate and prevent OA. If you see acute, non-traumatic, monoarticular arthritis, treat it as a septic joint until tapped [remove fluid with a needle].”

Every hour with Prasanna is an opportunity to learn more vocabulary, e.g., the Lisfranc ligament, named after the French surgeon who pioneered the “Lisfranc amputation” of the tarsal-metatarsal joint (mid foot) during the early 1800s. I learn names for common injuries from a shoulder dislocation, including the Hill-Sachs lesion (humeral head fracture as it strikes the glenoid) and the commonly accompanying Bankart fracture of the glenoid.

[Editor: read Madame Bovary for some insight into 19th century French foot surgery.]

School administrators had stressed that I was to work only with the MSK seat and stick to the 9-11, 1-3 schedule. On Thursday, however, I asked the two radiologists if it would be okay to work from 9-1, spending half the time with the abdominal seat, and having the whole afternoon free with the puppy. “Of course,” was the answer.

I watch Prasanna perform a hip arthrogram in prep for an MRI on a 59-year-old with worsening anterior hip pain and clicking for 3 months. He weighs at least 300 lbs. Once the needle is in the joint space, straw-color fluid slowly flows out of the catheter. This went on for a few minutes, until Prasanna aspirates a total of 50mL. “That feels so much better,” exclaims the patient. “I’m glad, the pain might come back a bit as I inject the contrast now.” After the procedure, he asks, “Do I need a hip replacement?” The radiologist explains, “We’ll know more once we get the MRI, but from just this X-ray, I see preservation of the joint space so my guess is no. You do have a large joint effusion and at least a labral tear so you might still need surgery, but not a joint replacement.”

The abdominal seat is reading a pelvic MRI on a 49-year-old female for rectal cancer staging. “The most important thing is if the tumor invades the sphincter complex.” The internal and external anal sphincter muscles are highlighted by the clear “intersphincteric fat pad” that is being pushed by the tumor on the posterior lateral side. The radiologist: “This is bad. She is going to probably have to get an APR (abdominal perineal resection, in which they remove the anus and create a colostomy). We’ll see what the rectal surgeons say at tumor board next week.”

We have a CT angiogram of an 86-year-old for an adrenal mass, her fourth in two years, due to an anomaly discovered on a CT scan after a fall. The abdominal radiologist says that the test should never have been ordered. “Leave this woman alone. Adrenal masses are statistically benign in the absence of metastatic disease (e.g., lung cancer). Teleradiologists never have the guts to ignore something out of fear of getting sued so she’s subject to never-ending imaging follow up.” He continues, “It’s weird to say, but I don’t always want the ordering provider to follow every finding in my report. Don’t treat the image, treat the patient. We balance this with the knowledge that this report will be forever cemented into the patient chart for litigation years in the future. We used to call up the ordering provider, or he would come down to us. With teleradiology, the doctors don’t collaborate and each one tries to defend against any possible lawsuit. It’s almost like we are in a game of tug-of-war on who bears legal ownership of a patient. Tag, you’re it! The result is that a patient who lives 4 hours from the nearest MRI machine will be doomed to perpetual follow-up on a statistically benign tumor.”

Statistics for the week… Study: 0 hours. Sleep: 8 hours/night; Fun: 2 nights. Example fun: Jane and I attend an engagement party for Outdoors Oswald, a mountain biker applying to emergency medicine at prestigious institutions. His fiance works for Epic, which allows her to work from home most days. She hopes to end up in New York City, even though “we’ll be broke.” They rented a private downstairs room, but did not order any food for the gathering. About half of the class was invited and consequently the open bar was used to the fullest extent. We left at 1:00 am with several classmates to grab a slice of pizza before Ubering home.

The rest of the book: http://fifthchance.com/MedicalSchool2020