week 6

It is the first day of your emergency medicine elective rotation and it is a very busy evening. The attending, Dr. Roberts, suggests that you shadow for the first few hours to become oriented to how the emergency department works. She is about to see a patient and is speaking with the nurse outside of the door. The nurse says, "Kayla is a 16-year-old female who was seen just a couple days ago. Looks like she was diagnosed with pelvic inflammatory disease (PID). Her abdominal pain isn't getting better, and now she's vomiting."

Dr. Roberts says, "I've seen lots of cases like this before—the doxycycline can cause a lot of stomach upset, and they can't keep it down. I'm guessing she's still in pain because she's not been adequately treated. Teenagers can be so tough to treat, even for simple things."

You and Dr. Roberts walk into the room together and introduce yourselves.

"Hi Kayla, I’m Dr. Roberts. I have a medical student with me today, is it okay if they listen while we’re talking?" (Kayla continues moaning uncomfortably)

(While moaning) "Yeah, whatever."

"Wow, it looks like you’re not feeling any better at all."

"Uh-uh"

"Can you tell me what’s been going on?"

"Oh, the pain, it’s just worse… It’s a lot worse. Whatever they did before didn’t work."

"So tell me what’s been going on over the past day or so?"

"Oh, the pain, it’s just, uuuuhhh! I can’t stand it anymore!"

"Where is the pain?"

"Ugh, down here." (points to L pelvic area)

"Okay, so still on that left side?"

"Mmm-hmmm."

"So the way it feels, does it feel any different or just worse?" (As Kayla is answering, Dr. Roberts’ pager goes off. She pulls it out and looks at it, then puts it back on her waist while Kayla is answering).

"Just worse… Ugh, it comes and goes, but it’s so much worse!"

(A little distracted, getting back into the moment). "So it’s really bad huh?"

"Yeah, it gets really, really, really bad and I have to throw up."

"Okay. How much have you been throwing up today?"

"Uuuuh, probably, um … a couple times."

"Anything bloody in there, or dark green? (Kayla shaking head, saying 'uh-uh,' while Dr. Roberts is talking) Okay. So the pain, on a scale of 1 to 10, how bad is it?"

"When it’s really bad? A ten!!! OH, GOSH!"

"Does anything make it worse?"

(kind of irritated) "When I move! It’s just … it’s so bad I can’t even think straight!"

"Okay, have you taken any medicines for it?"

"Um, I think my mom gave me something, like ibuprofen, I don’t know, but it didn’t work."

(The nurse pops her head in.)

Nurse: "Just a reminder, you still need to put an order in for room 3 before I can give the medications."

(Turns to nurse) "Okay, yes, sorry I’ll be right there" (turns back to Kayla, getting more terse and direct in her questions, though still has empathetic body language, seems somewhat distracted). "Ah, when you left the hospital last time they gave you a prescription for some antibiotics. Have you gotten those?"

"Um, I think so, yeah. But I’ve been throwing up so I missed my dose today."

"Okay. Just a couple more questions. Any fevers?"

"No, I don’t know. We don’t have a thermometer."

"Peeing and pooping okay?"

"Yes."

"And then just a couple questions I ask everyone your age. Are you sexually active?"

"Yes, but my boyfriend said he’s been tested."

"Do you use condoms?"

"Sometimes…" (moaning)

"Having any vaginal discharge?"

(Frustrated) "The same as before! I already told you guys this … Can I please have something for this pain."

It can be helpful to think about the decision-making processes we use to make medical decisions.

You log in and pull up Kayla's electronic medical record (EMR). You see that her gonorrhea and chlamydia tests are still pending, and then navigate to the note from her ED visit two days ago, when she was seen by Dr. Santos, to gather more information.

HISTORY
Chief Concern:	Pelvic pain
History of Present Illness:	16 y/o F with left lower and mid pelvic pain, moderate, started this a.m. Came on suddenly, sharp, some intermittent relief but no clear relieving or exacerbating factors. Tried ibuprofen and heat packs, no change. Non-bilious non-bloody vomiting x 2. +Vaginal discharge, white, no pruritis. No prior episodes. No known prior sexually transmitted infections. No sick contacts.
Review of Systems:	Negative except as per HPI. Reports no dysuria, hematuria, flank pain, fevers/chills, diarrhea, constipation. LMP: periods irregular since Nexplanon placed 6 mos ago.
Past Medical History:	Asthma
Medications:	Albuterol PRN, Nexplanon
Allergies:	NKDA
Family History:	Non-contributory
Social History:	Sexually active, 4 lifetime partners male and female, last intercourse 5 d ago with male partner, consensual, no condom, positive occasional EtOH and marijuana use, no other illicit drugs, no history of sexual abuse, no history of depressive symptoms. Lives w/ both parents and sister, 10th grade, does well in school.
PHYSICAL EXAM
Vitals:	T 37.9 C, P 85 bpm, BP 110/72 mmHg, RR 14 bpm, POx 99%RA, Wt 62kg.
General:	A&O, NAD, appears mildly uncomfortable, lying in bed
HEENT:	NC/AT, MMM
Cardiovascular:	RRR, no M/R/G, nl S1/S2
Respiratory:	CTAB
Abdomen:	Soft, TTP in suprapubic and left pelvic region otherwise NT elsewhere, +BS, non-distended, no hepatosplenomegaly, neg psoas, no guarding/rebound, neg Murphy's.
Pelvic:	Normal external Tanner 5 female, moderate thin white/yellow discharge in vaginal vault, no cervical discharge. There is discomfort with movement of cervix and during left bimanual adnexal exam, no pain on right during bimanual examination.
Extremities:	WWP, CR < 2 sec
Neurological:	Grossly normal
Skin:	No rashes
LABS	Negative HCG, negative wet mount, GC/chlamydia sent and pending, UA pH 5, SG 1.020, neg nitrites, neg LE, trace heme, trace protein, neg ketones, neg bili, neg glucose.
IMAGING	Abdominal radiograph read as normal loops of bowel, no air fluid levels, scant stool throughout colon, overall unremarkable.
ASSESSMENT PLAN	16y/o F with 12hrs left pelvic pain and vomiting, sexually active, with cervical motion tenderness and Left adnexal tenderness. Most likely PID. Negative UA rules out pyelo, negative HCG rules out ectopic pregnancy. Pain in LLQ, not RLQ, appendicitis unlikely. Pt expresses concern for severe pain but exam does not seem consistent with surgical process such as appy or torsion. KUB not consistent with constipation or with obstruction. Appears non-toxic and tolerating small amounts of oral fluids in the ER. Given 250mg ceftriaxone x1 in ER, Rx doxycycline 100mg PO BID x14d, advise f/u with PMD in 2–3 days or sooner if worsens or not tolerating PO. Call pt at 999-999-9999 confidential cell for f/u GC/chlam results.

Dr. Santos clearly considered a number of other items on his differential diagnosis besides PID, and considered why each diagnosis may or may not fit with Kayla's presentation. He has "rank ordered" his differential diagnosis to come to the conclusion that she likely has PID. This is an example of system 2 thinking: Dr. Santos consciously weighs multiple factors in making a decision; at the same time, however, his illness scripts for each of these diagnoses are influenced by his previous experiences.

You have never seen a patient with PID before and you decide that you want to create your own differential diagnosis using a very deliberate approach. Considering a number of alternatives, you feel like you need more information from the patient. You go in and examine Kayla and find that she has significant tenderness in her abdomen. As you're walking out of the room, Dr. Roberts approaches.

"Phew, what a night!" she says. "Back to Kayla... it sounds like she's going to need to get admitted for IV antibiotics since she can't tolerate the oral treatment for her PID. I already wrote the order to get her a bed. Let's go in and I'll do Kayla's exam before she goes up." Your history and physical examination have made you wonder if this might be something other than PID. But before you're able to get in a word, Dr. Roberts opens Kayla's door.

"Hi, Kayla, sorry for the interruption," Dr. Roberts says as she washes her hands. "We think you need to come in to the hospital so that we can give your antibiotics through an IV. Before they come move you to your room, we just need to do a quick exam." Dr. Roberts briefly listens to her heart and lungs. Dr. Roberts begins to reach for her abdomen and Kayla curls up her legs and retracts.

Dr. Roberts tells Kayla to relax, but Kayla keeps pulling up her knees and wailing when Dr. Roberts tries to touch her lower belly. You see the frustration in Dr. Roberts' eyes as she just tries to get through a cursory abdominal exam. She tells Kayla that the nurse will be in soon to place an IV and says kindly, "I'm sure you'll start feeling better once the IV antibiotics are started. I'll ask the nurse to give you some medicine to help with your pain, too," and leaves the room.

As you leave the room, Dr. Roberts says, almost to herself, "The abdominal exam was so tough, I'm sure she won't tolerate a pelvic exam. We'd have to move her to a different bed. Besides, she just had one a couple of days ago, I don't think I'd find anything new."

You start to discuss what you found in the chart and some of the thoughts you had, but Dr. Roberts interrupts, "I have to get back to the trauma bay. Let's try to talk about Kayla later when things slow down." But things never slow down and the shift ends without further discussion.

As you walk towards the ED three days later for your next shift, you think about Kayla and wonder how she did. Your rotation requires that you review charts of the patients you've cared for and you start with Kayla. You learn that Kayla was started on IV antibiotics, and two days later when her severe pain and vomiting continued, the gynecologists were consulted.

After seeing Kayla, the Gynecology team discussed with the Pediatrics team that maybe PID wasn't the correct diagnosis after all. They thought that ovarian torsion was possible and decided to take Kayla to the operating room for a laparoscopy. During the operation, they found that her ovary was necrotic and could not be salvaged and needed to be removed.

When you have a little break and go to grab a coffee with Dr. Matthews, the Pediatric Emergency Medicine fellow you're working with, he asks you how the rotation is going. You tell him about the case and say, "I don't know how we could have missed that! It seems so obvious now. I just wonder what we could have done differently."

Dr. Matthews responds, "These things are always easier to see in hindsight. The important thing is to learn from them so that we don't make the same kind of mistake again. What do you think went wrong?"

Dr. Matthews responds, "I totally get it. It can be hard to talk about these things without feeling like you're assigning judgment or blame to the providers involved—especially when you're a student talking to your residents and attendings. It is normal to feel guilty and frustrated after an error occurs. Many studies have shown that providers suffer significant negative consequences when they are involved in an error."

A constructive response to diagnostic error is fundamental. Here are examples of some potential responses as well as some modifications that could make the responses more constructive.

Original Response	More Constructive Response
Dr. Roberts was too busy. She barely got to take a history, we glossed over the physical exam and I never got to tell her what I found.	It was busy, people were stressed, and a trauma had just arrived. The team had to be efficient. I don't think we were aware of how many shortcuts we were taking, but looking back, we were not as thorough as we should have been.
No one was listening to Kayla when she talked about the severity of her pain.	Dr. Roberts has seen many sexually active female teenagers with pelvic pain before. This influenced her history taking, thought process and decision making. In the end, though, there were things that made her different from the other patients she'd seen in the past.
I feel like our care was sub-optimal. We didn't even do a pelvic exam.	The ED is not designed to make pelvic exams easy to perform on any patient. This combined with a busy night and a recent pelvic exam influenced Dr. Roberts' decision to defer the exam.
I knew there was something going on; I didn't speak up because Dr. Roberts is the attending. I'm just a med student, and I was only shadowing. I should have said something.	It is intimidating to start a new rotation as a medical student. I need to remember this is a teaching hospital and attendings are used to interruptions and questions. Dr. Roberts was using System 1 thinking, which I have not developed yet. Everyone can play an important part in patient care and asking questions can be helpful.

Dr. Matthews continues, "You know, I have something that I think might help you as you think about this case and for approaching future ones. It was given to me by my mentor when I first started fellowship. I'll make you a copy. I know that as I reflect on my errors, I've missed diagnoses at times because another diagnosis was easier to make. For example, I missed a diagnosis of appendicitis once because the patient had a long-standing history of constipation."

Cognitive Load

Cognitive load is a concept that is frequently applied to medicine when discussing diagnostic error. It is human nature to minimize cognitive load when possible because mental processing is labor intensive and also highly valuable, therefore we automatically reserve processing power for when it is needed most. This tendency is instinctive and cannot be turned off. We rely on heuristics we have formed--mental shortcuts, or pattern recognition--to ease the cognitive burden and conserve energy. We automatically utilize heuristic-based processes unless there is an intentional, conscious overriding of this tendency.

An example might be when you are considering two different conditions in a patient: one that is a common, simple disease and another that is more rare and more complicated. Your brain will likely automatically consider the common, simple disease before the other one due to the fact that it is easier to think about.

It is important to recognize this concept so you can learn ways to overcome it when diagnosing patients' problems.

The cognitive load is highly connected with heuristics, since heuristics are mental shortcuts our brains use to make decisions that require little energy.