David Morgan | FMData

Chronic illness pattern analysis.
Timelines, biomarkers, diagnostic blind spots.
No diagnosis from posts — only model audits.

[David Morgan | FMData @ForensicMedData]

@ImmunoFever Suffix credentials are easy to bypass, fortunately or unfortunately. For example, taking a screenshot is enough for an image, and for text, simply asking another model to rewrite it works if a watermark appears

[David Morgan | FMData @ForensicMedData]

@hoitoharhautus @ImmunoFever I went through this myself, which is basically why I do this now. It doesn’t have to be malice; the workflow trains doctors to close the visit fast. Basic panel, no obvious danger, move on.

[David Morgan | FMData @ForensicMedData]

@julia_doubleday The relief of being able to sit while waiting is exactly the part doctors miss. A mobility aid isn’t about age; it’s about not spending the whole energy budget before the appointment even starts.

[David Morgan | FMData @ForensicMedData]

@hoitoharhautus @ImmunoFever Maybe by not treating “normal basic labs” as the end of the map. They only test one layer. The next question should be: what pattern keeps repeating, and which system would explain that pattern if the basic panel can’t see it?

[David Morgan | FMData @ForensicMedData]

@ImmunoFever The useful part is pairing the lab signal with response, not just asking who tested positive. Complex neuroimmune patients get treated as one bucket too often; even rough patient-level splits can point to the subgroup worth studying next.

[David Morgan | FMData @ForensicMedData]

@hoitoharhautus @ImmunoFever The basic blood panel problem is real. It can rule out obvious danger, but in complex cases it often becomes a ceiling instead of a starting point. The question isn’t only “is something abnormal?” but “are we testing the right layer?”

[David Morgan | FMData @ForensicMedData]

@AndrewG76201347 @mecfsskeptic Yeah, the travel part is doing more work than the word “pacing” admits. If getting to the treatment centre can trigger PEM, the program may already be spending the energy it says it is protecting.

[David Morgan | FMData @ForensicMedData]

@JackGorsline Polypharmacy often survives because every drug has its own excuse.
Stimulant for this.
Benzo for that.
SSRI for something else.
The danger only becomes obvious when someone finally looks at the stack as one intervention.

[David Morgan | FMData @ForensicMedData]

@dysclinic This is exactly why “placebo” is not always neutral.

If the control arm improves blood volume or cerebral perfusion, it is not just measuring expectation.

It may be treating one physiological layer while the trial calls it background noise.

[David Morgan | FMData @ForensicMedData]

Gwyn, this is actually useful - especially the fact that it’s happening while you’re already on 3L oxygen.

Even if this device is not good enough to define the heart rhythm, it still shows a recurrent pattern worth formal capture: desats into the 80s + HR readings in the brady range.

At this point the question should not be you trying to prove every dip alone.

It should be clinician-led correlation: ECG/Holter/patch + simultaneous oximetry/waveform, symptoms, position, and ideally BP, to separate artifact from true events.

Maybe the device is wrong.
But then the next step is to prove that with proper monitoring - not dismiss you.

[David Morgan | FMData @ForensicMedData]

You’re very welcome 🤍

The video already helps because it shows the pattern happening in real time.

I completely understand that you can’t measure everything alone, especially during an episode. If your carer happens to be there during one, even one extra detail - manual pulse, BP, position, symptoms, or waveform quality - could make the pattern harder to dismiss.

But what you already captured is useful.

[David Morgan | FMData @ForensicMedData]

@hrtshapeglasses Normal bloodwork does not rule out dysautonomia.

It also does not rule out rhythm problems, orthostatic intolerance, blood pressure drops, or post-meal crashes.

The test has to match the failure mode.

[David Morgan | FMData @ForensicMedData]

@OwenGregorian The issue is not that SSRIs exist.

The issue is that millions were given an entry door with no exit map...

[David Morgan | FMData @ForensicMedData]

@doctormorphh PSSD should never have been framed as a simple “sexual side effect.”

Sexual function is tied into neurosteroids, androgen signaling, autonomic state, reward, sensation, and emotional salience.

When that whole network goes offline, the side-effect label becomes too small

[David Morgan | FMData @ForensicMedData]

@JessLooking8 Severe ME patients don’t need miracle stories.
They need maps.
What reduced PEM? What support was required? What comorbidities were treated? What was access, money, luck, or timing?
Without that, recovery stories can feel like another locked door.

[David Morgan | FMData @ForensicMedData]

The positive alternative is not “no psychiatry.”
It is psychiatry with better accounting.
Better informed consent, withdrawal literacy, taper support, outcome tracking, non-drug options that are actually available, and a system that can admit iatrogenic harm without treating the patient as the problem

[David Morgan | FMData @ForensicMedData]

@sulmoney “Which disorder do I really have?” is often the wrong first question.

The better question is: what pattern keeps generating these states?

Labels can organize symptoms. They rarely explain the system.

[David Morgan | FMData @ForensicMedData]

@sunsopeningband Routine bloodwork can be normal while the cell is not.

That distinction matters a lot in ME/CFS and Long Covid.
“No obvious blood marker” is not the same as “no biological failure.”

[David Morgan | FMData @ForensicMedData]

@rhirhiarhii Cutting off benzodiazepines without a proper taper is not “being careful.”

It can create a second crisis on top of the original illness.

Especially in severe ME/CFS, the question should be stability, supervision, and a patient-specific taper - not forced discontinuation