To many ER CTs – not my imagination


Category : Medical Rants

The December issue of the Annals of Emergency Medicine reports on this problem Computed Tomography Use in the Adult Emergency Department of an Academic Urban Hospital From 2001 to 2007:

Editor’s Capsule Summary
What is already known on this topic
The use of computed tomographic (CT) imaging in the emergency department (ED) has increased in the past few years, but the increase for specific types of imaging has not been well described.
What question this study addressed
The authors determined local utilization rates for CT imaging of the face, head, neck, chest, and abdomen-pelvis for 2001 through 2007.
What this study adds to our knowledge
In this single, urban, academic emergency department, CT utilization increased steadily, from 51 per 1,000 visits to 106 per 1,000 visits during the study period (approximately 10 CTs per 1,000 visits per year). Increases were found in all types of imaging, with the greatest increases in chest and neck imaging.
How this is relevant to clinical practice
Understanding CT utilization does not directly affect current practice but provides information that can guide future imaging strategies.

As I talk with hospitalists around the country, this report confirms our observation.  This study does not discuss why this occurs, but I will be so bold as to speculate.

Emergency Departments are besieged.  The law that commands EDs to provide care without regard for payment status leads to an increasing patient load.  The primary care shortage drives patients to the ED.  Emergency physicians practice in the "fog of war".  The Wikipedia definition of the phrase includes this wonderful quote: The great uncertainty of all data in war is a peculiar difficulty, because all action must, to a certain extent, be planned in a mere twilight, which in addition not infrequently — like the effect of a fog or moonshine — gives to things exaggerated dimensions and unnatural appearance." Carl von Clausewitz (Prussian military analyst).

It appears that too often CT scanning takes the place of a careful history and physical examination.  This can occur when the emergency physician is drowning in patients.  But we should care!  CT scans when accumulated likely cause cancers.  Unnecessary CT scanning contributes to increased health care costs, not just from the cost of the scans, but from the chasing down incidentalomas that often follows a CT scan.

What solutions should we consider?  The obvious first solution is to create an appropriate outpatient infrastructure in this country.  We need to pay outpatient physicians better and allow them to spend time with patients.  The next radical suggestion is that we should modify emergency medicine training.  In the late 70s I spent a couple of years working in emergency rooms.  My internal medicine training was highly worthwhile.  I believe that emergency physicians need more inpatient experiences to better understand the natural history of disease. 

This is a crisis of convenience.  It has great financial and indirect costs.  We should not blame emergency physicians, but rather work with them to correct this problem. 

Comments (30)

I would be more amenable to your suggestion were it not so patently smug.  By the way, the name of the journal is not "Annals of Emergency Room". Starting with a dismissive attitude toward Emergency Medicine, which is your habit, isn't going to open the conversation you want.

The premise is we do too many CT scans in the ER.
Then we beg the question.
I am all for good thinking, good histories and physicals, and against the danger of excessive radiation from CT scans. 
There is no question we are doing more CT scans than we used to.
But I would pull back and ask what is too many and how do we know we are doing too many?
What is the definition to the first question and what are the data to answering the second?
Ask yourself, how do we know we are not doing enough CT scans (I am not saying that is the case, only positing that is an equally valid hypothesis in the absence of concrete in formation).  
Just think about this- we are doing far fewer V/Q scans for PE. IS that good or not? Probably form a diagnostic standpoint, yes. From other standpoints who knows, but it is one of the effects of more CT scans. I am sure there are others, good and bad.

Having experience on both sides of the coin – rural GS practice and rural ED coverage –
1. I concur heartily with the issues regarding marked improvements in availability for OP medicine – absolutely; BUT
2. IP medicine training is the LAST thing EM docs need – might have been helpful prior to 1980, but now – no one sees the integration of natural history with medical care anymore – way too fragmented in time and space – mobility, hyper-specialization…you know the story, and have written extensively about same…
3. Fog of war + 7/11 mentality = immediate gratification demands…"What is wrong,,,now" is the demand at all levels, particularly in the ED, and I see no end to same…a function of our attention spans, media, etc…
Hence – I must rule out what will/could kill you, what must be operated on, what will require upstream triage and transport to BCH (big city hospital…),,,and I am not allowed to wait and see anymore. In my office when I counsel same, the followup visit is missed or canceled within the next 48 hours and records requested..OR I hear in church of the complicated disaster we are asked to pray for at BCH…etc.
Having started in the game when it was heads only, then grainy fuzz in the abdomen, to what we have now…more may be better – rare negative appendectomies, unexpected findings, earlier and planned interventions.

The #1 cause of excess CT: litigation.  Having been is several depositions, there is always the question, "Doctor, why did you not do XXX?"  If XXX=CT, and there is the remote chance that the CT would have altered the outcome (and there is always a witness for hire who will state unequivocally that it would have), you can just start writing a check.  Juries have a faith in technology that is not justified, but that's the world we live and practice in.  The defendant physician can have a dozen experts on his side.  The plaintiff needs only a wheelchair.

If what you say is true, andI have no reason to doubt you, then doesn't the cost-benefit argue for a CT scan?
IF the CT scan costs the hospital, I'm talking cost not ridiculous charges, say $500, and failure to do it lead so to  one settlement a year of $500,000, the finding would have to be so remote as less than 1/1000- in which case I seriously doubt you would lose many suits (not to say anything of helping the person you diagnose). 
OF course there are other things to be considered, radiation, other tests to be done (and tests avoided) but that argument doesn't make any sense when you examine it – too many CT scans are being done because it costs too much money in litigation not to do them. If that's true, do them- it will be cheaper in the long run
Mind you- it may not be true. But no one knows – everything is anecdotal.
In every other business though, litigation costs are part of any equation. Why should medicine be different? 

The goals of internal medicine aren't the same as emergency medicine.  Not only is emergency medicine required to work up what is the most likely diagnosis, but it must also rule out emergencies.  It's the business of stuff that could kill you soon, and a good history and physical can only take you so far.  (
I would like to echo Cory.  What is the proper role of CT in the diagnosis of an acute disease?  No one knows, but EM as a speciality is working on it and cares deeply about it.  On the other hand, what's the acceptable miss rate for subarachnoid hemorrhage?  The number is zero.  At least in the case of subarachnoid hemorrhage it's clear that prior to the invention of CT, EM was under utilizing CT in the work up.  
Balance the acceptable miss rate of disease versus the unknown "proper use rate" of CT, add a dash of tort, and within there you'll see how we got to where we are today.

It appears that too often CT scanning takes the place of a careful history and physical examination. 
Dr. Centor, have you even set foot in an emergency department recently? I don't know what kind of button-pushing monkeys you think we are, but I am very deliberate and cautious with my imaging. I'd like to argue three things:
*Hindsight's 20/20. It was an "overutilization CT" once you know it's normal.
* CTs frequently change the disposition on patients, who go from "admit" to "discharge," when they would have otherwise been admitted for observation, serial abdominal exams, etc. (who hospitalists don't even ever see, you only see the N of admitted patients) and
* I find that it's the consultants and inpatient teams who don't want to hear the consult or admission without imaging done and the patient all tucked away:
Surgeons with patients with rigid abdomens but no free air, for "pre-op planning;"
Medicine who disagrees with the ED about PE risk and won't accept a patient without a CT;
Urology with patients with kidney stones;
Trauma Surgery with patients with penetrating neck wounds (the alternative is a bronch, barium swallow, EGD and angiogram);
Neurosurgery with an altered patient with a blown pupil;
Neurology in a patient with a stroke;
Ortho with a tibial plateau fracture for "pre-op planning."
Many more patients are coming thru the Emergency Department than used to, and we're ordering the scans — but try admitting a patient without one. You can't.
Here's who I've imaged in the past two days with CT:
Sudden onset headache 10/10 worst headache of life and neck pain after chiropractic manipulation;
3rd visit bounceback back pain radiating to the RLQ and tender with 2 negative workups so far, no prior CT done;
Ataxic 29 year-old with dysdiadokinesia starting 1 day ago with headache and vomiting;
90 year-old on coumadin with two syncopal episodes and visible head trauma;
Left facial droop with forehead sparing sudden onset 1 hour ago;
Seizing AIDS patient;
Chest and upper back pain, hypertensive, unequal pulses, diaphoretic
I went to the same medical school as my colleagues in internal medicine, and learned just as well how to take a history and physical. If you would happily accept the above patients on your medicine service without imaging, then please, come work at my hospital. No one else will touch them with a 10-foot pole!
* I did not image the 24 year-old with headaches for the past two months who wanted a "full body MRI just to be sure," or the hypotensive guy with epigastric abdominal pain but no tenderness on exam, nor the guy with a classic story for kidney stone. I could have. But instead, I took an adequate history and physical.

I tire of reading that the primary reason for overuse of expensive diagnostic modalities is fear of lawsuits. Malpractice lawsuits have been studied extensively. The studies show that physicians who order too many imaging studies and other diagnostic tests are more likely to be sued and more likely to lose the lawsuits. The reason is obvious: by ordering too many diagnostic tests with little justification, it is easy for an attorney to paint the picture of a floundering incompetent who had no clue about the patient's condition and ordered diagnostic tests seemingly at random.
I spent over four years as lab director at a VA medical center where the risks of lawsuits were very low. Physicians still overutilized imaging studies and lab tests. In non-emergency situations they did so for two main reasons: laziness or a compulsion for diagnostic certainty.
In the VA ER, the main reason for the overuse of imaging studies and clinical lab tests was that the physician either could not or would not take the time to perform a more detailed history or a more intensive physical exam. The rough equation was that two dozen lab tests or a CT scan equalled 15 minutes of physician time. In a swamped ER, such trade-offs might be appropriate, since 15 minutes of physician time can save a life or two. However, the trade-off is inappropriate during less hectic times. I believe that some physicians failed to change their work patterns during less hectic times and continued to substitute imaging studies and lab tests for bedside time.
However, I believe that the reason for the rapid doubling of the rate of ER-ordered CT scans is the greater availability of CT scanners, and not greater avoidance of bedside time by ER physicians. Modern CT scanners are both faster and cheaper than older ones, so it is no surprise that they are used more.

The simplest explanation is the most likely explanation.  We order too many scans because we can.

Next question.

I forgot to add, there is a reason doctors in  central Africa  don't order too many scans.  It's because they can't. 
Now.  Please disperse.  There is nothing else to see here. 

If you want to cut down the number of CTs ordered in the ER, then require a prior authorization process and make CT scanners less available. My two local hospitals have scanners within 20 feet of the ERs that run 24 hours a day to get 20-30 CTs per scanner daily. If I see a patient in the office with RLQ pain, it easier for me to send the patient to the ER than to justify a stat CT. The prior auth includes requiring fever, elevated WBC, and acute onset of RLQ abdominal pain. Even then, it may require a physician to physician prior auth to get it ordered stat as an outpatient. The stat CT charge at my local hospitals range from $2,000-$5,000 each, depending on the contrast or no contrast.

Hospitalist- I ask you  how do you know we order too many? Do you have actual information or does it just seem like too many?
It's like in baseball when the manager said so and so is such a good shortstop he saves us three runs a game with his fielding. Until someone actually looked at the numbers and analyzed them and found out that no one saves three runs a game with their fielding and there were 10 guys in the league better thna this guy anyway.  
solofp – you haven't answered the question of why we would want to make CTs less available. Are these hospitals really just doing them to make money or do they actually benefit people?
Maybe the thing to do is bring down those outrageous charges.  
I don't know the answer to these questions but I haven't seen any convincing answers given other than it just seems like…

To add one more thing:
I don't think the issue is that emergency physicians are too lazy, too incompetent, or too unskilled to do a good H&P.  I still think the issue is less tolerance for diagnostic uncertainty under reasonable time constraints.  If someone comes into the ER with RLQ abdominal pain, there's no such thing as watchful waiting for a few days to determine the clinical course.  That patient needs a diagnosis and a disposition in a matter of hours.
I would bet that you could take the most skilled, internal-medicine trained, grey haired, anecdote and time riddled physician, put them in the ER, and their practice would slowly approach that of an EM-residency trained physician.  Albeit very slowly, as that IM physician reinvents the wheel and rediscovers the past 30 years of EM medical progress for themselves.  The diagnostic demands and diagnostic uncertainty of emergency patients is intrinsic to the pathology of the ER; in only a small part does it have anything to do with physician ineptitude.

Good discussion here.
It seems that a lot of the comment here has to do with an impression that if ED physicians just spent a little more time with their patients, CT imaging would reduce dramatically.  I don't believe that — my experience doing emergency medicine is that a lot of CT scanning is ordered either to rule out conditions where we are expected to have 100% certainty not to miss (subarachnoid hemorrhages are one example already mentioned), or where the demand for certainty is being driven by the patient or our inpatient colleagues.
I request a number of  CT scans simply because our the surgeon, internist, or whoever is going to admit the patient won't do so until it is done.  There is a drive to have admissions  "packaged" and complete, with a certain diagnosis before they leave the ED.  .When was the last time you got a patient with nonspecific abdominal pain or diverticulitis admitted to the surgeon?  Sure, they "might" have an abscess, but a lot seem to get better with just symptomatic management.  Why not have them imaged the next morning from the ward if they are not better? 

Dr Centor is right! We *do* do too many CTs! I implore you all, lazy careless ED physicians to return to the True Path: Pneumoencephalograms and Peritoneal Lavage for trauma, VQ scans for PE, and a solid history and physical exam for Thoracic Aortic Dissection. (Did YOU check for pulsus paradoxus before ordering that CT? Lazy!)

A brief comment on my ER colleagues, who I happen to agree with generally on this point.
Why should the focus of the internist or surgeon be any different than yours? Shouldn't they want the most timely expeditious diagnosis also? (and doesn't everyone want to avoid the misdiagnosis?)
Don't be judgmental, persuade them that your approach is the best from any and all concerns. And personally I think most of the really good ones would agree with that. 

Cory, when I said "we", I didn't mean ER.  We is all doctors.  We all order too many because we can.  It's just too easy not to. 

I know it's beside the main point but VQ vs CT has been brought up at least twice now in this thread.  If you go by the evidence, not the popularity, CT is no better than VQ as the initial test for PE.  CT over VQ is trend based medicine.

[…] H&P When You Can Just CT? Dr. Centor is at it again, with more bashing of the Emergency Department, this time because we order too many CTs. He […]

Dr T-
Sorry if you’re tired of reading that fear of lawsuits is a driver of increased use of imaging.  Is it possible that physicians who are in specialties that tend to get sued a lot are the ones who image more, and that drives the standard of care in that specialty towards more imaging?  You suggest an alternative- that increased use of imaging is the mark of a bad doc, and that is why they get sued.  I take allegations from a lab manager that a physician may be a “floundering incompetent who had no clue…” with a grain of salt.  Get out here into the world of clinical medicine for a few years, and then criticize your medical colleagues.
You criticize physicians who have a “compulsion for diagnostic certainty.”  That’s actually the doctor I want.  And it’s the doctor I try to be.  Knowing the correct diagnosis can be useful when treating patients, and treating according to an incorrect diagnosis can lead to harm.  I would not have thought that it would be necessary to make this point to another physician in a medical blog, but there it is.
Every study that indicates overuse of this or that in medicine is done in retrospect, and is therefore both accurate and useless.  I do not think that there are no overused modalities, only that we keep that in mind before we criticize others.  That goes double for physicians in other specialties- just because I as a surgeon occasionally visit the ED to care for a patient that has been worked up by the ED physician does not make me an expert in Emergency Medicine.  And if the Emergency Physician has had to evaluate the patient for other diagnoses before arriving at the correct one and calling me, it does not mean that he has wasted resources in (compulsively) reaching the correct diagnosis.  He has done the job that the patient has metaphorically hired him to do- figure out what is wrong and fix it, or find someone else who can fix it.

[…] Dr. Centor is at it again, with more bashing of the Emergency Department, this time because we order too many CTs. He cites this great study, by my friends/colleagues Jarone and Jonathan at my own institution, showing that CT imaging has risen in our own ED over time. And why has use of CT gone up? According to Dr. Centor, an academic hospitalist, it’s due to (his words, not mine): […]

Nothing advances the care of a patient like a diagnosis. Ruling out unlikely but potentially lifethreatening problems is probably just behind this.   Maybe rather than castigating doctors who are striving to accomplish the above, we should come up with protocols to limit the radiation exposure when CT's are needed. Modern CT scanners take far more slices and deliver more radiation than the scanners of the past. Perhaps for certain conditions the scanner could be programmed to take fewer slices. I'm not enough of a radiology guru to know how feasible this is, but it's a thought. I am old enough to remember looking for pineal calcification shifts to help diagnose subdurals and agonizing over whether I was missing an atypical appendicitis on people with abdominal pain and am really thankful we have this technology available (not to say it shouldn't be used with discretion, but God is it helpful!)

Don't forget patient satisfaction which is often measured in the ED. Scared patients feel more re-assured after a doctor orders "a lot of tests," than when he or she simply examines them and says, "I don't think that there is anything wrong with you."

plenty of consultants demanding advanced imaging before they would be willing to assume care of the patient. just sayin'.

The idea that ordering a CT saves time versus a detailed H&P underscores the mistaken assumptions made by non-emergency physicians about how we care for our patients. Any emergency physician worth her salt knows understands that obviating a CT scan by adding an extra 20-30 minutes interviewing and examining a patient saves much more time & work than simply ordering a scan.
Patients who wait for CT scans take up much more time and many more resources than those who don't. If we can discharge a patient clinically (maybe because of a low Centor score) than they free up all of those resources (i.a. nursing). This is essentially the same reasons why we love to admit those pesky low-risk chest pain patients BEFORE the first troponin comes back — boarding is bad for our patients.

jb: You are misreading my comment. I said that the malpractice lawsuit data shows that over-ordering diagnostic tests is more likely to result in a lost malpractice suit. (I made no statements or judgments about whether losing the suits is fair or appropriate.) Physicians who, when asked, state that they order unneeded diagnostic tests to avoid lawsuits are either working from a false premise or are lying to themselves or to the questioners about their reasons.

I work in an academic ED and know this blog author.  It is interesting that many of the residents he teaches in IM, routinely want CT scans on the admitted patients before they come to the floor.  Many times, I have not ordered the CT because I do not feel it is indicated.  My residents want to help their colleagues and orders their scan for the patient to get "on the way up".  But still counts as a scan in the ED.  Maybe the ENTIRE house of medicine should look at how we are training all of our housestaff….

Here's my take on ER physicians and CT scans.   I took more heat from this post than any other.

[…] scrutiny on the perils of unnecessary radiation, the blogosphere was abuzz about this topic (see Bob Centor and Bob Wachter among others).  Coincidentally, our grand rounds speaker last week was Dr. Bruce […]

[…] seem to be finally dying down after a contentious online debate sparked by Dr. Robert Centor, who blogged recently about a study showing increased use of CT scanning in hospital emergency departments, and suggested […]

Post a comment