Reflexive testing algorithms

[Enkidu]

---

Members don't see this ad.

I was reading an article on the CAP website about reflexive testing, and it kind of piqued my interest. From what I gathered, at MGH when a patient has a prolonged PTT, the clinician simply checks a ''prolonged PTT'' box on his lab requisition form and all appropriate tests are performed automatically and he receives a report with a pathologists interpretation.

Is this a common way to run a lab? I was under the impression that clinicians just had to know every lab test to run for prolonged PTT and order them all and interpret the data on their own. Probably a reflexive method makes more sense, and it makes ordering lab tests more properly like a pathology consult than an all-you-can eat buffet.

if interested, here's the link to the article: http://www.cap.org/apps/cap.portal?...ation.html&_state=maximized&_pageLabel=cntvwr

 

[pathstudent]

That's an awesome idea and an awesome way to make money off CP. Assuming there is a -26 code for interpreting coag and other lab tests.

 

[Enkidu]

It seems like this type of system is the best all around. It should decrease unnecessary testing and lead to better interpretation of lab tests.

I kind of wonder why this isn't the standard of care. I can see a clinician being found liable for not checking the ''prolonged PTT'' block and then going on to order inappropriate tests.

 

[Gene_]

There is a lot of random test ordering by clinicians in the area of coag. An algorithmic approach, even with pathologist compensation, is probably cheaper than shotgunning every coag test available - and then getting the Hematology consult.

 

[Enkidu]

What other situations would this type of approach also make sense? Would it make sense for there to be an ''anemia'' block on the requisition form, or is that too simple?

Also, it seems like there are a lot of weird antibodies associated with lupus, each with their own specificity and implications for prognosis. Do clinicians just order these up, or do they tend to use just one or two? Do pathologists impact which antibodies are ordered and how they're interpreted? Maybe it's not really that complex, but I recall it seeming that way when I studied it.

 

[sirenomelia]

This paradigm works well in molecular- ie you start with a most likely scenario- a major BCR-ABL PCR and then you reflex to the minor if the major isnt there.

 

[ScotchBonnet]

We do HPLC for variant hemoglobins and send the ones that can't be fully parsed out to Mayo. They do a cascade beginning with electrophoresis and culminating in DNA sequencing if needed.

http://www.mayomedicallaboratories.com/test-catalog/Overview/81626

 

[Enkidu]

How common is it for these reflex testing panels to be given to the clinician with a narrative interpretation by the pathologist? That article cites a study from MGH in which the clinicians said that the narrative interpretation of the coag tests significantly benefitted them in something like 80% of cases.

Also, which types of lab tests would pathologist interpretation likely be beneficial in this same way? It kind of seems like pathologist interpretation of tests should be the standard of care, even though clinicians also interpret the tests. After all, clinicians also try to read their own imaging, but they still have the radiologist sign off on it.

Of course, I don't mean every lab test should require pathologist interpretation, but at least a significant portion of them. Isn't all of this stuff part of CP boards?

 

[solumanculver]

Here's an interesting popular article about clinical pathology providing narrative interpretations

http://www.mc.vanderbilt.edu/reporter/index.html?ID=9211

A quote from the article: “Everybody would think it's crazy if the internist got back a CAT scan and the radiologist just said, ‘Call me if you have a question,’” Laposata explained. “But across the country in the clinical lab, we just give doctors complicated numbers and say, ‘Good luck.’”

Do you guys think that analogy is sound?

 

[KCShaw]

Yes and no.

Yes in that the lab, like radiology, creates something new which has to be interpreted. No in that many, many results can and should be interpreted by clinicians without complicating the matter with a canned path interpretation (which may unnecessarily conflict with the management course). Although clinicians are taught to interpret basic radiology, most are more aware of their own limitations when reading them, and make a habit of checking the final report. Imaging is also not usually ordered in multiple batteries of tests several times a day, unlike some CP tests.

I certainly think there is a place for more reflex testing and interpretive assistance. In most cases, interpretive assistance currently comes from clinician specialty consults rather than from the lab -- this seems logical since clinical specialists are attuned not only to diagnosis but also to management, where pathologists will generally be at a disadvantage. But clinicians/PCP's/FamP's are also encouraged to begin a workup prior to arranging a consult, and combined with residency training seeming to encourage shotgun ordering, we end up with a poor-use-of-resources problem.

Those institutions, consult services, and/or pathology labs who are paying attention, however, can and probably should identify the most common mis-ordering or interpretive problems they face and make an effort to address it -- whether by reflex testing, interpretive assistance, lectures/conferences, or whatever makes the most sense. Unfortunately they may not be financially inclined to do so if they're being paid to work more than they otherwise should.

 

[Enkidu]

many results can and should be interpreted by clinicians without complicating the matter with a canned path interpretation (which may unnecessarily conflict with the management course).

So the clinician's algorithm is: 1) Order some labs 2) Consult specialist 3) Specialist orders appropriate labs that were neglected 4) Specialist diagnoses patient

It seems like it would be fool-proof to run things the way that MGH does with coags. Check ''prolonged PTT'' box, all appropriate labs are ordered and pathologists sends an interpretation. It could be that a specialist consult is not even needed. According to this Laposta character, this is the model that clinicians prefer (whether that's true or not isn't clear, of course).

What do you mean when you say that the path interpretation may conflict with the management course? Does that critique apply to radiology as well, that the radiologist may produce an interpretation that conflicts with management?

 

[KCShaw]

I'm thinking primarily of common abnormalities, such as a slightly elevated creatinine, slightly elevated potassium, etc. A canned path comment about these things is unlikely to encompass all of the culprits, but in the case of civil litigation means the physician now has to have an explanation for everything in the canned comment. And a canned comment is probably not necessary because it's something most clinicians probably face fairly regularly.

Radiologists do occasionally produce a report which conflicts with management, but in my limited experience this is because of a difference in interpretation combined with some clinical correlation. I.e., a radiologist says something like "consistent with acute fracture" or "likely represents osteomyelitis" while the orthopedic surgeon thinks otherwise and treats differently. Fortunately most radiologists, as well as pathologists, tag on a line to somewhat address that, such as "..clinical correlation is advised."

I just think interpretive comments need to be tailored to your patient & clinician population -- not shotgunned like path ordering, heh. Many esoteric test orders should probably lead to a phone call to the ordering clinician, and certain things lend themselves well to reflex testing, as already mentioned. And you make a good point that it could limit the necessity of a specialty consult -- often there are good treatment algorithms readily available, while diagnostic algorithms may be harder for an inexperienced clinician to find or feel comfortable with.

 

[Enkidu]

I'm thinking primarily of common abnormalities, such as a slightly elevated creatinine, slightly elevated potassium, etc. A canned path comment about these things is unlikely to encompass all of the culprits, but in the case of civil litigation means the physician now has to have an explanation for everything in the canned comment. And a canned comment is probably not necessary because it's something most clinicians probably face fairly regularly.

Radiologists do occasionally produce a report which conflicts with management, but in my limited experience this is because of a difference in interpretation combined with some clinical correlation. I.e., a radiologist says something like "consistent with acute fracture" or "likely represents osteomyelitis" while the orthopedic surgeon thinks otherwise and treats differently. Fortunately most radiologists, as well as pathologists, tag on a line to somewhat address that, such as "..clinical correlation is advised."

I just think interpretive comments need to be tailored to your patient & clinician population -- not shotgunned like path ordering, heh. Many esoteric test orders should probably lead to a phone call to the ordering clinician, and certain things lend themselves well to reflex testing, as already mentioned. And you make a good point that it could limit the necessity of a specialty consult -- often there are good treatment algorithms readily available, while diagnostic algorithms may be harder for an inexperienced clinician to find or feel comfortable with.

Yeah, I see what you mean. I didn't really intend to suggest that pathologists should be interpreting chem 10 type labs. Those are certainly for the clinicians to sort through. It seems like certain conditions are diagnosed completely based on laboratory tests, though. And it would seem more efficient for the clinical pathologist to simply send an interpretation with the lab values. At least, it should minimize the room for diagnostic error and cut back on unnecessary clinical consults.

Based on Laposta's research at MGH, this is actually a service that clinicians want from the lab, and that they claim decreases diagnostic error. If that's the case, it seems like labs should provide it and pathologists be compensated for it.

Also, I really like reading lab tests and diagnosing patients, and I like to think that there is a clear role for pathologists in this, who are meant to be the authorities on lab tests and the final word on diagnosis.

 

[deleted314957]

compensation is the issue. books were written about how to get money from clin path consults shortly after TEFRA in the 80's. All payors have really cracked down on the hoops you must jump thru to get a pittance for clin path consults. They don't pay well and for the time and effort my time is better spent cranking out 88305's.

 

[Enkidu]

But if there was data showing an improvement in quality and a decrease in unnecessary test ordering (costs), I suspect that insurance companies would reimburse for it. Or is that not realistic?

 

[KCShaw]

If the data is good, I think it's realistic. But I agree that money is a major obstacle in making this sort of thing widespread. It's probably more likely that clinical specialists will provide algorithms, or that pathologists offer ordering algorithms, in the LIS ordering system. Still, depends on whether enough institutions can show an overall monetary benefit. I'm sure there are a -few- clinical pathologists who are glad to stay busy by providing this kind of service without solid compensation, but without competition or other impetus, there's not a lot of extra motivation. And certainly not much extra motivation to go into the field.

Although, I admit one major turnoff for me for CP in general was the lack of..well..doing anything. I would have been more interested if techs weren't doing essentially all the troubleshooting, making most of the decisions, determining validity of results, viewing smears, managing send-outs, etc. In -some- labs -some- pathologists get more involved in one or more of those processes, but in my experience the majority of the time it's managed in the tech hierarchy.

 

[Enkidu]

Although, I admit one major turnoff for me for CP in general was the lack of..well..doing anything. I would have been more interested if techs weren't doing essentially all the troubleshooting, making most of the decisions, determining validity of results, viewing smears, managing send-outs, etc. In -some- labs -some- pathologists get more involved in one or more of those processes, but in my experience the majority of the time it's managed in the tech hierarchy.

Yeah, this is my concern. I'm not fully convinced that I should do CP, but the study I linked to made me think that it might be interesting after all. I like to diagnose patients; if clinical pathologists don't do that, then I don't see the point in doing the residency, particularly if it is not required for the academic career that I'm interested in.

 

[KCShaw]

I think you would have to search out the right program to give you good experience in doing those sorts of things. In a lot of programs you may spend a little time with a CP pathologist, then get pawned off to the techs -- where you'll learn some things, but have to keep in mind the techs aren't really getting paid to train you and don't have a deeper medical background -- or you'll be told to go read, or do someone else's pet project largely useless to you.

However, as an attending you can sorta write your own ticket in CP -- if you want to get involved in doing this kind of thing there probably isn't going to be much stopping you. Some CP'ers do this, dabbling in their areas of particular interest. There just may still be no compensation for working harder than is generally "expected."