Guide
Literature review for the medico-legal file
In clinical research, a literature review asks what the evidence suggests. In a medico-legal matter it asks something harder: will this survive cross-examination? A paper that "supports" a proposition in a review article may say much less when read line by line in front of an opposing expert.
MIA, Medical Intel, is a literature desk built for that standard. It works on the paper in front of you and writes a working that keeps three lines apart.
Where medico-legal literature goes wrong
Abstract-only citation. An expert cites a paper for a finding stated in the abstract. The full paper qualifies it, limits it to a subgroup, or reports it as exploratory.
Population mismatch. The study population differs from the claimant in age, comorbidity or setting, and the difference isn't addressed.
Association as causation. An observational association is presented as if it established cause in this case.
Standard of care confused with research. An investigational finding is treated as what a reasonable practitioner should have known or done at the time.
Out of time. Literature published after the events is used to judge conduct, without saying so.
The three lines MIA keeps apart
| Line | What it contains |
|---|---|
| What the file shows | Visible or audible findings only. No invented measurement |
| What may be drawn | Reasonable inference from the record, named as inference, not dressed up as a result |
| What it does not establish | A missing method, a thin sample, an abstract without the paper. The gap stays a gap |
A paper appraisal template you can copy
PAPER APPRAISAL: medico-legal
Citation: ___
Cited in report for: "___" (quote the sentence, page/para)
Full text on file? Yes / No (abstract only)
Design: RCT / cohort / case-control / case series / review
Population: ___ Matches the claimant? ___
Exposure / intervention: ___
Outcome measured: ___ Same as the claim? ___
Key result (with CI): ___
Primary or exploratory? ___
Published date vs events: ___
THE FILE SHOWS: ___
MAY BE DRAWN: ___
DOES NOT ESTABLISH: ___
Replicated / contradicted by: ___
A worked shape (fictional abstract)
Question: does this abstract support a change in practice? File: a conference abstract reporting a signal in forty adults. No methods appendix, and no full paper on the file. The file shows: one abstract, a stated sample, an endpoint named by the authors. Nothing else. May be drawn: the authors describe a signal. That's their claim. It isn't yet a replicated result. Does not establish: full methods, comparator detail, replication. The desk won't promote this to standard of care. Record: a change in practice isn't established by this file. Human judgment remains the authority.
Sometimes the important answer is: the file doesn't establish it.
How MIA works
- Sign in at mia.medicalintel.org. A conversation is for testing a question. A file is for the paper, scan or exhibit you need to stand behind.
- Put it down. Type, dictate, or attach the PDF. Google Drive and Dropbox can be connected from the desk. A paper stays on that desk and isn't copied to another app.
- Choose the pass. Ask is one pass. Review runs again when the file has changed. Deep Review is two independent workings, then a settlement: agree, contest, or still a gap.
Clinical Conservative by default
The default mode favours high-quality, directly relevant evidence, and keeps standard of care apart from investigational or speculative claims. A hypothesis-generating mode exists for exploratory work. It always opens with a banner saying it isn't standard of care.
Limits
MIA gives no dosing, schedules or implementation parameters, and no help obtaining, travelling for, financing or accessing care or trials. It is educational, not a clinician. The medico-legal opinion remains the expert's.
With LIA Pro
Medico-legal files are legal files too. On an FIA Plan seat, you can send an FIA Brief from MIA to LIA Pro, the legal desk: the question, the conclusion you're willing to show, the caution. The legal working can then use the finding. Only that packet crosses.
Questions
Can AI do a medical literature review for a legal case?
It can do the first working: what each paper shows, what may be drawn, and what it doesn't establish. An expert still gives the opinion.
Is MIA medical advice?
No. MIA is an educational literature and exhibit desk. It isn't a clinician, isn't a medical device, and isn't a substitute for licensed professional judgment.
How does MIA treat a conference abstract?
As an abstract. Without full methods, a comparator or replication, the desk won't promote a signal to standard of care.
Can I connect Google Drive or Dropbox?
Yes. Both can be connected from the desk. A paper stays on that desk.
Does MIA give dosing advice?
No. It gives no dosing, schedules or implementation parameters.