Saturday, May 24, 2014

Gone for a Week!

Hi all,

Blogs have been sporadic because I am getting married this weekend! I'm very, very excited. It has been a really eye-opening process planning such an event, and I am so lucky to have a perfect partner. I will be gone for a week from this blog. I hope you enjoy a lovely Memorial day weekend. It'll be beautiful.


Friday, May 23, 2014


The last set of patients in the CVICU are the post-transplant patients. We do a good number of heart and lung transplants, and their clinical course can be quite variable. Single lung transplants, especially if they are done off bypass, tend to recover quickly. Without the physiologic insult of bypass, these patients have fewer bleeding problems, wake up quickly, need fewer vasoactive drips, and leave the unit quickly. If they have an epidural for pain management, they end up being the least complex patients we have. The ICU course of a double lung transplant, in contrast, can vary quite a bit. Depending on the quality of the transplant lungs, the length of bypass time, and the other comorbidities, these patients can have a quick extubation and smooth course or be intubated for weeks heading for a tracheostomy. The fluid and pressor management can be tricky, and sometimes we sacrifice other organ systems like the kidneys to keep the lungs alive. Similarly, our heart transplant patients have quite individual clinical courses. Some come out on high dose pressors with an intra-aortic balloon pump, and weaning the cardiovascular support can be really hard. With a long bypass time, patients can have complications with bleeding, delirium, and renal failure, so a lot of it depends on the skill of the surgeon and technical difficulty of the transplant. The variable clinical course really engages our medical assessment, decision-making, and treatment plan.

Monday, May 19, 2014


Along with the "regular" cardiac surgery service patients, we also care for even more specialized patients. The device service is a group of cardiac surgeons and cardiologists that manage patients with complex mechanical heart devices. Most commonly, we have patients with end stage congestive heart failure who get LVADs or left ventricular assist devices. These machines replace the main pumping function of the heart; the surgeons implant a conduit between the left ventricle and the aorta and the machine does all the work ejecting blood from the heart since the heart cannot do it by itself. This is interesting in itself; the LVADs eject blood continuously so these patients have no pulse and sometimes their blood pressure is simply one number (if the heart is not doing any work, then there is no separate systolic and diastolic value). Pre-operatively, these patients often have multi-organ system failure, so when they come out of surgery, they can keep us up all night. Although the left heart is supported, their right heart is not, and that's usually what kills these patients post-operatively. The device can pump blood from the left heart to the body, but if the right heart cannot fill the left side, the device is useless. So we spend a lot of care titrating epinephrine, inhaled nitric oxide or epoprostenol, milrinone, vasopressin, and other agents to support the right side. Meanwhile, the liver, kidneys, and brain are quite tenuous as they recover from surgery. These patients really engage the gamut of critical care knowledge.

Rarely, surgeons will implant assist devices to both sides of the heart: an LVAD and an RVAD. These are bigger surgeries with greater bleeding and infection risk so even though both sides of the heart are supported, patients don't always do well. Although we have done total artificial hearts, I didn't see any during my rotation. The goal of nearly all these procedures is a bridge to transplant; most of these devices are not designed for long-term support, and their ongoing risks are high, so patients undergo these operations in hopes of eventually getting a new heart.

Saturday, May 17, 2014


Immediately following my reigonal anesthesia rotation, I went to the cardiovascular ICU. Known as one of the toughest rotations for anesthesia residents, I had a little bit of dread. What scared me the most was the acuity of the patients in the unit. Although we have the usual "straightforward" postoperative cardiac patients after a bypass surgery or simple valve replacement, we also get extremely sick patients after redo-sternotomies, aortic root surgeries, heart and lung transplants, and other complex surgeries. It's not the simple patients recovering from surgery that scare me but that minority of patients who are actively dying.

We have about 20 patients on the census at any one time. The majority of them are post-op from cardiac surgery. They include some of the more tertiary-center surgeries like minimally invasive valve replacements, transcutaneous transfemoral valve replacements, and large aortic surgeries. These patients either do really well or really poorly. While minimally invasive or transcutaneous surgeries are smaller traumas to the body, the patients we get have a lot of comorbidities. Their heart may do great, but their severe COPD, pre-operative renal insufficiency, chronic pain, or altered mental status hamper the recovery. The aortic surgeries are the opposite; surgeries are a huge ordeal with many hours on bypass with extreme cooling of the body. But these patients with Marfan syndrome tend to be young and healthy. Some of them recover in a few days and others have a long intensive care unit course. The cardiac surgery service also accepts aortic dissections from the community who come to the intensive care unit for blood pressure management and determination of a therapeutic plan. These vary a lot in acuity; some are sent for a mild descending Type B dissection which is non-operative so we put them on blood pressure medications and get them out of the ICU quickly. Others have to go emergently to the operating room without medical optimization; the post-operative critical care course is always rocky. The wide range of patients, surgeries, and acuity pose a constant challenge for the cardiothoracic ICU team.

Thursday, May 15, 2014

Regional Anesthesia

By the end of my regional anesthesia rotation, I felt very confident doing most of the standard blocks. I understood the anatomy of the brachial plexus and the lower extremity nerves, became facile with the ultrasound, and developed reasonable skill for finding a nerve, putting a needle next to it, and numbing it up. It's actually quite fun and satisfying as the results are immediate, and the benefit for patients significant. By providing good post-operative analgesia and in some cases avoiding general anesthesia, we could really improve the patient's experience with orthopedic surgeries.

The problem is that unless I continue to practice nerve blocks, it is something that fades with time. Everything in medicine is like this; if you were to ask me the current guidelines for cholesterol management or how to deliver a baby or how to read a CT scan, I'd be quite rusty, though they are all things that I've learned in the past. This is the challenge of regional anesthesia; we either have to make it something we do regularly or we have to live with the fact that we will never be experts.

Rest assured, it's not a hard skill. After spending several years in medicine, I've realized that putting a needle into a target is not all that difficult. But doing so efficiently and understanding the nuances and risks are a little more tricky. I don't envision myself blocking people day after day, but this rotation has given me the confidence to broach regional anesthesia in the future if I feel that it is best for the patient. For example, it is useful to have in the toolkit in international medicine where general anesthesia is riskier and pain medications more scarce.

The rotation as a whole was a steep but wonderful learning experience. I did upwards to a hundred blocks in four weeks, and learned not only the technical skills but also the systems issues regarding regional anesthesia. I was thrilled by how my patients did. I was frustrated by the paperwork. I came to understand a whole new facet of anesthesiology.

Tuesday, May 13, 2014


Our regional rotation is one of the few anesthesia rotations where we do a lot of patient followup. For a regular anesthetic, I don't usually visit a patient post-operatively to see how everything went. I'll often scan a chart to make sure no untoward events happened, but it simply isn't built into our day to visit all our post-ops. (I think there is a lot of value in doing so and am a proponent of seeing inpatients after their anesthetic, but to work, time has to be set aside for it). In the case of a big surgery or anesthetic, I'll often use some of my lunch hour or stay late at the hospital visiting a patient, something I mostly do for cardiac surgeries, liver transplants, or cases that had a little deviation from normal.

But on our regional anesthesia rotation, we make a big point to see every single patient we block. We check to make sure the nerve block worked for the surgery, provided some pain relief, and resolved without an issue. We make sure questions are answered, patients are satisfied, and surgeons give their feedback about a block. Even for outpatients, we will call and check in on them.

The process of followup is pretty satisfying; we see the result of our handiwork, feel good about the blocks that went swimmingly, and get a sense of the blocks that failed and why. We also get to do some PR, teaching patients about anesthesiology, educating them about nerve blocks, and explaining why we do what we do. However, the system we use for documenting followup is a monstrosity. Most of my evenings were spent struggling with a database that didn't load right, asked the wrong questions, and couldn't be edited. They are working on updating it, but as with all things in medicine, change is slow, and the paperwork is much worse than the clinical care of patients.

Sunday, May 11, 2014

Happy Mother's Day!

Happy mother's day to all the mothers out there! Like an aging wine, my gratitude to my mother (and my father) takes on new flavors and characteristics each year. They have supported me, taught me, nurtured me, and loved me, and it seems so inadequate to acknowledge it only once a year. But here it is. Happy mother's day. As more of my friends become parents, this is taking on a different feeling for me. I'm so proud of them as they explore this new stage of life. I also think of the women laboring and delivering today. I spent nearly the entire day in the hospital, but not on L&D. So many ways mothers should be honored.

Friday, May 09, 2014

Sick / Not Sick

The truth is, most doctors have shown up to work at least once when they were sick. Probably most people have shown up to work with a cold or sniffles or a cough. For physicians, there are unusual pressures to do so. The training environment, hierarchical structure, and role models of medicine encourage powering through illnesses. Taking a day off means someone else has to cover your patients, which incurs more risk to those patients who don't know their doctor. Taking a day off means someone else has to cover your call, working overnight or on a weekend. Taking a day off means you're wracked with guilt. Taking a day off means you're weak. Medicine, at least historically, nurtures, comforts, and cares for patients, but rarely does so for the physician. We are seen as impregnable; we ought to know all about health and disease; and we should never get sick.

Clearly, this is the wrong attitude. And medicine is slowly changing to reflect that. There is less stigma to taking a sick day. There are backup plans in place to cover for physicians who are out. We watch out for each other and make sacrifices for each other. But most importantly, coming to work sick puts patients at risk. Especially with highly contagious diseases like the flu, infection can spread rampantly. Influenza to a healthy doctor is an annoyance; to an immunocompromised elderly patient, it's life-threatening. For our patients' sake, we should not come to work when we think we could get others sick. Being doctors, you would expect us to know this.

However, this whole issue isn't so clear cut and obvious. We had a patient over the holidays with congenital heart disease who needed a bowel repair for duodenal atresia. The patient was four days old and the surgery was an emergency. But pediatric cardiac anesthesia is such a specialized field that our group of anesthesiologists for those cases is very small. And it turned out that the two anesthesiologists who were supposed to be in the hospital had the flu. One person who was not supposed to be working was already in the hospital doing in a pediatric heart transplant. All the others were out of town because it was Christmas. There was physically no one available with expertise in pediatric cardiac anesthesia to take care of this four day old baby. Now, a physician who works will sick puts his patients at risk. But here, being sick at home puts this patient at risk as well. The newborn ended up having his surgery the following day when one of the anesthesiologists felt better. It wasn't ideal, but it was the best compromise in solution.

Wednesday, May 07, 2014

Catheter Based Blocks

Talking about nerve blocks for post-operative pain management is fine and dandy, but one limitation is the duration of action. For a single injection, our longest acting local anesthetics only last for half a day. In fact, this can cause problems with patients who get a nerve block in the early afternoon; they go home happy, go to bed pain free, and then wake up in the middle of the night with excruciating discomfort when the block wears off. When I talk to patients about nerve blocks, I have to set up expectations on when the block will recede so that patients can start taking oral pain medications to smooth the transition.

However, in some cases, we can leave a small catheter near the nerve to deliver local anesthetic continuously. We no longer have to worry about the limited duration of a single injection. Placing the catheter is a little more challenging and a lot more time consuming than a quick injection, but when it is effective, it's very satisfying. For inpatients, we visit them every day to make adjustments on their local anesthetic infusion. We try to balance pain control with muscle weakness; the more numbing a patient gets, the greater the effect on the motor function. To optimize physical therapy, a patient needs to be able to move with minimum pain. For outpatients, we can send someone home with an automatic infusion system. The catheter is connected to a disposable bulb filled with medication. We adjust the rate of delivery before sending the patient home. When the medication runs out, the patient simply takes out the soft flexible plastic catheter and throws it all away. I think this has made outpatient pain management much easier as surgical discomfort can be managed outside the hospital.

Monday, May 05, 2014

Time Pressure and Regional Anesthesia

You learn very quickly in anesthesia residency that surgeons have little patience, especially for case delays. The key to keeping your surgeon happy is to keep the cases on schedule. We lament about this; no one complains if a surgeon takes longer than expected for a case, but if there's an anesthesia delay, we never hear the end of it. I actually think this is one of the biggest barriers to acceptance of regional anesthetic techniques. Surgeons are used to the speed of general anesthesia. A nerve block in the most skilled hands might take five extra minutes. A tougher block, more inexperienced practitioner, or a catheter technique can take ten or fifteen minutes.

We ought to do what is best for the patient, but sometimes surgeon preference trumps other priorities. I would hope that waiting five or ten minutes would not be a big deal, but somehow it is. And so if a surgeon is consistently delayed or if turnover time is longer than he expects, he may ask us to proceed with a general anesthetic rather than nerve block technique. I think we often acquiese though the best thing for the patient might be regional anesthesia.

Academic institutions try to address these issues with clever approaches. We have a nerve block service with a separate attending, fellow, and residents that perform all the nerve blocks needed. That way the regional team does the block rather than the anesthesiologist assigned to the case. Since we aren't providing anesthesia for ongoing surgeries, we can do the nerve block in advance, minimizing turnover delay. This also allows for a small cohort of regional attendings to ensure a high rate of success. However, the downside to this model is financial; the department loses money by paying for an attending to do nerve blocks rather than operating room anesthesia. This is why private practices rarely have a separate team, and instead rely on the assigned anesthesiologist to do an efficient, safe, and effective block.

But I don't think changing our delivery models is the only answer to this problem. We need to change the culture of the operating room so that it does not revolve around any of the providers, but instead focuses on the patient's needs. In pediatrics, for example, surgeons don't pressure us as much. They understand that separating a child from her parents, inducing anesthesia, placing an IV, and putting in a nerve block can take a variable amount of time. More importantly, they respect that what we do requires skill and isn't always straightforward. We don't feel as much time pressure to hand off the case. So in pediatric anesthesia, the assigned anesthesiologist performs the nerve block while everyone else waits. We do what we think is best for the patient.

This idea needs to trickle through the rest of the health care system. There may be other reasons to have a separate regional anesthesia team such as efficiency or cost. But an anesthesiologist should never choose an anesthetic technique simply because a surgeon won't wait for the block to set up.

Saturday, May 03, 2014

You Learn Something New Every Day

Image is from xkcd, drawn by Randall Munroe, shown under Creative Commons Attribution License.