High Prevalence of Alternative Diagnoses in Children and Adolescents with Suspected Long COVID—A Single Center Cohort Study 2023 Goretzki et al

Abstract

Background: Long COVID (LC) is a diagnosis that requires exclusion of alternative somatic and mental diseases. The aim of this study was to examine the prevalence of differential diagnoses in suspected pediatric LC patients and assess whether adult LC symptom clusters are applicable to pediatric patients.

Materials and Methods: Pediatric presentations at the Pediatric Infectious Diseases Department of the University Hospital Essen (Germany) were assessed retrospectively. The correlation of initial symptoms and final diagnoses (LC versus other diseases or unclarified) was assessed. The sensitivity, specificity, negative and positive predictive values of adult LC symptom clusters were calculated.

Results: Of 110 patients, 32 (29%) suffered from LC, 52 (47%) were diagnosed with alternative somatic/mental diseases, and 26 (23%) remained unclarified. Combined neurological and respiratory clusters displayed a sensitivity of 0.97 (95% CI 0.91–1.00) and a negative predictive value of 0.97 (0.92–1.00) for LC.

Discussion/Conclusions: The prevalence of alternative somatic and mental diseases in pediatric patients with suspected LC is high. The range of underlying diseases is wide, including chronic and potentially life-threatening conditions. Neurological and respiratory symptom clusters may help to identify patients that are unlikely to be suffering from LC.

Open access, https://www.mdpi.com/1999-4915/15/2/579

 

This is irrelevant to the findings, but is it normal for clinicians to treat “somatic” and “psychosomatic” as synonyms? My understanding was that “somatic” means the same as “biomedical”, as a strictly etymological interpretation would suggest.

 

I know the definition of vs common usage of the word 'somatic' is somewhat divisively complex but......

isn't this trying to reverse-logic?

SSD (somatic symptom disorder) I think can be happily criticised for basically being a 'next level' from the previous incarnation which required exclusion of other conditions that explain the symptoms.

So until DSM V 'somatic' diagnoses required exclusion of all other explanations - 'distressed about symptoms that have genuinely been looked into'

NOW they are trying to suggest a physical diagnosis of long covid is only possible if you exclude a somatic diagnosis - which used to only be possible by excluding the physical diagnoses and possible explanations but now got away with suggesting it's a 'positive rule-in' based on clinician assessing 'distress' (which in its extreme means someone with cancer distressed about the prognosis could be diagnosed with SSD for having distress, because that is an arbitrary 'absolute' - although not even that as it is by clinician's eye/assessment not a blood test - and not relative to circumstance)

 

To be fair I think they are using somatic as an alternative to biomedical because they do not want to say biomedical. It could easily read "biomedical and mental". As you noted, psychiatrists just call biomedical "somatic" so they can keep the discussion on their ground.

My problem is - what mental diseases, how were they validated, and how can you be sure? In cohort of such size I might suspect some brain (I prefer that to mental, its my alternative) diseases, not many, but some possibly not due to covid. On the other hand secondary brain dysfunction due to compromised circulation and potentially many other physical problems arising from covid, are likely.

My interpretation of this study is its just another way to push LC patient's problems into the too hard pile. Its a way to avoid real research and responsibility. When mental diseases are understood I suspect they are moved to another field, like neurology for example. The rest are not understood, but hey they have these theories.

 

Yes, sorry, on re-reading they are using the word properly. The red herring was the confusing use of “somatic/mental”.

 

It is set up for confusion as later on in the paper (in Discussion) they note that some of the other diagnoses found under the mental disorders section were 'somatisation disorders', so it seems they are including somaticisation, just not under 'somatic'

 

It is interesting to look at the table of diagnoses. The most common were:

RPIMS-TS 16 (31%) which is the inflammatory syndrome triggered by covid

Bronchial asthma 6 (12%)

Covid 19 5 (10%)

Lambert-Eaton Syndrome 5 (10%)

oncological/haemotological disease 4 (8%)

adjustment/somaticisation disorder 3 (6%)


The rest listed were 1 or 2 individuals (although I don't know whether 'groupings' or how specific something was makes a difference here ie could some of those have been 'grouped' to be bigger numbers etc)

Apart from adjustment/somaticisation disorder only depression (1 (2%)) was listed under 'mental disease'. Although the discussion note there were those who had concomittant mental illnesses.


I thought the following was interesting to have a think over (haven't managed to yet but always wary of finding what you look for etc and not knowing whether LC clinics in Germany also have a staffing that tends to always include respiratory, with the 'rest' being a bit more mixed):

I'm guessing this next bit 'came from' that finding, and is a suggested 'way forward' or method development, but haven't read in enough detail to see where this last bit came from?

As others have said the discussion noting the following key message is reassuring:

 

Yes the word "biomedical" must fill the psychiatrists with dread, another "customer" lost to the biomedical field.

 

Yet with antidepressants the spiel was/is that it is biomedical, hence why you need them etc. There are a number of others where these are treated pharmacologically (who knows what the actual theories mean re: that or how solid they have to be in that literature etc)

 

It's funny as a layperson because many at some point before they 'get wise' assumed that - whatever the illness - people with 'something wrong' were 'thoroughly investigated'. So in essence there would be files for people with ME/CFS that noted different blood test anomolies if they had them or different symptom profiles - and anything else they also had was discovered and treated.

We of course know that isn't the case in the UK - was it ever the case if you did have a decent GP or health system?

But that's perhaps where the awful lie of BPS infferring 'nothing wrong ever found' really gained its traction in that mythical world.

This in effect seems to be having to sell the idea of what should be occurring anyway - which is that whether they also end up with LC as well after all that people should be looked into by what is wrong with them in a sensible way. Because after all some poor so and sos could have ten things and others none wrong with them at the same time. And because how do you study LC if you don't note these things down - as some comorbidities might be telling you something about LC, and others if not picked up interact badly with it or so on.

But what has happened to medicine where this 'concept' is having to be sold ?

 

Its the non-biomedical psychiatry that has an issue with it. There is data showing that some drugs have good effect on many patients, including with depression, though I think many are less effective than they claim. This is because of deliberately poor experimental design. I have forgotten all of the examples I used to know, but one way this is done happened with one of the antidepressants, an SSRI I think. They had massive effect on one subgroup, very beneficial, with minimal effect on most patients, but they emphasized the average effect size, misleading a casual reader into thinking it helped most patients.