How best to answer multidiagnoses questions what involve NEXT BEST STEP?

[SlaveOfTCMC]

---

Members don't see this ad.

Derived from a qbank:

A child with a VSD suddenly presents with anisocoria w/ dilated right pupil, headache, and left sided weakness.

The child also has an acute fever, and a new mitral heart murmur (not ascribed to the LSB holosystolic murmur attributed to the VSD)

So I am down to choosing between

CT Scan of the brain (brain abscess/septic emboli)

and

CBC and blood culture (endocarditis)


So in real life, multiple teams would be working on this poor child. But for Step 2, am I to answer this question based on 1) What is most life threatening? (both are if you ask me, but the more precise timeframe has the brain abscess more threatening)

or

2) What is causing the etiology?



The answer was CT scan of the brain, btw.

 

[turkeyjerky]

CBC is virtually always wrong, since it's really never useful (just as in real life).

For these questions, I would recommend thinking about the real world management and then picking the corresponding correct answer which is most critical to the patient's management. If a lab test is actually required for the diagnosis and will have a meaningful impact on the management, then that can be the right answer. But most of the time the question stem will give it to you clearly enough that you should opt to treat empirically. This is especially the case w/ time-sensitive conditions. Also, choosing the lab implies that you will wait for the results to come back before doing anything else.

This case, you can easily diagnose endocarditis. So you think about what you would do in real life. 4 things come immediately to mind: blood cultures, antibiotics, head CT (w/ contrast in this case) and echo. Choosing "cbc and blood cultures" implies that you would wait for the results to come back prior to treating--which you would not do in real life. Had the choice been "draw blood culture and administer empiric abx", then I would've chosen that.

Think about each choice in these questions as a fork in the road, which will determine subsequent management (if diagnostic choices) or as initial or definitive therapy (if therapeutic options). Diagnostic modalities that will not alter management are never the right choice.

 

[MedSchoolPhenom]

CBC is virtually always wrong, since it's really never useful (just as in real life).

You are a complete idiot and will be a terrible physician with thinking like that.



On a related non-scathing note....please wikipedia the following disease: Acute Leukemia.

 

[turkeyjerky]

You are a complete idiot and will be a terrible physician with thinking like that.



On a related non-scathing note....please wikipedia the following disease: Acute Leukemia.

Wow, love the hyperbole. What's the name of that sdn rule? Oh yeah, Goodwin's law.

I'm sorry you or your loved one had leukemia, but I recommend you come back here in a couple years after you've actually been in med school and know something about this test. Also, you might want to search wikipedia for "context".


P.S. If I was really concerned about leukemia, I'd start out w/ a peripheral smear--1/3 of ALL presents w/ normal WBC.

 

[b-real]

Maybe off-topic, but:
I would agree with TJ on this. Whenever you order a CBC (or any lab test/imaging modality), always ask yourself whether your assessment and plan will change dependent on the result. During third year, I would always order CBC. The best residents/attendings would question why I needed it. A lot of the time, I couldn't think of a valid reason, just that I wanted it.

 

[OveractiveBrain]

Derived from a qbank:
A child with a VSD suddenly presents with anisocoria w/ dilated right pupil, headache, and left sided weakness.
The child also has an acute fever, and a new mitral heart murmur (not ascribed to the LSB holosystolic murmur attributed to the VSD)

So I am down to choosing between
CT Scan of the brain (brain abscess/septic emboli)

and

CBC and blood culture (endocarditis)
So in real life, multiple teams would be working on this poor child. But for Step 2, am I to answer this question based on 1) What is most life threatening? (both are if you ask me, but the more precise timeframe has the brain abscess more threatening)

or

2) What is causing the etiology?
The answer was CT scan of the brain, btw.

So if you used the SEARCH function for "next best step" the first thread would have led you to the following quote, by me. The [brackets] indicate what I changed to answer your question exactly. Here is the link provided, so you can see what other people have already said in answer to this question.

By the way, you were ruling out a stroke, an emergent condition.

http://forums.studentdoctor.net/showthread.php?t=865189

Think of the "next step" like this

1. Is it an emergency? If yes, choose emergent intervention
2. Is the diagnosis certain? Therapy
3. Is more information needed? Diagnosis

Therapy
1. Real life. There are guidelines as to when to start something versus somethign else. Like a BP of 142/84 gets HCTZ whereas a BP of 170/101 is going to get a combo of therapies based on comorbidities, even though both are "hypertension." This is just stuff you have to know

Diagnosis
1. Choose the next step. Not the best test, but the next step. You think its an MI. You dont go to cath. You get Troponins and Twelve leads. If they were negative, you get a stress. If that was positive, you go to cath. You have to know the flow of diagnosis.

Usually they ask you to choose WITHIN one category, not between them. As in, you will be asked either to intervene OR to diagnose. Then its up to you decide how much has been done and what the next step needs to be, always keeping in the back of your mind "is this an emergency?" If it is, emergency trumps everything else and you go for definitive treatment, [or at least rule out the emergency] (which is why your [Uworld] answer [was CT scan].

 

[AndyRSC]

Overactive, did you derive this approach from personal observation, or did you read this somewhere? It certainly sounds common-sense; only reason I ask is because I'd like to read the source, if there is one, for such algorithms.