Case Teardown: How AI Screens a retained surgical sponge in Under a Minute

By John Mahoney · Founder, MedLegal AI · May 2, 2026

We talk a lot about what our free tools do. This is a teardown that shows them doing it. Below is a synthetic, de-identified med-mal fact pattern run through three of our free tools — the medical chronology, the case-merit score, and Bates-cited record search. Every timeline entry, every number, and every citation below is the genuine output of an actual tool run. Nothing was hand-edited.

The fact pattern

A patient undergoes a laparoscopic appendectomy. In the weeks after surgery they report persistent abdominal pain and low-grade fever. A CT of the abdomen identifies a retained surgical sponge, and the patient is taken back to the operating room for removal.

If you do plaintiff med-mal intake, you already feel the two questions: is this case real, and is it worth funding? Here is how the tools answer them in about the time it takes to read this paragraph.

Step 1 — The chronology builds itself

Drop the records in and the Medical Chronology tool returns a clean, dated timeline:

• 2025-02-11 08:30 · Procedure — Laparoscopic appendectomy
• 2025-02-11 10:15 · Document — Operative note — sponge/instrument count documented as correct — Operating surgeon
• 2025-02-25 14:00 · Encounter — Clinic — Persistent post-operative abdominal pain, low-grade fever
• 2025-02-25 14:10 · Observation/Lab — Temperature: 38.1 °C
• 2025-03-02 11:20 · Procedure — CT abdomen and pelvis
• 2025-03-02 13:40 · Diagnostic Report — CT abdomen — Retained surgical sponge (gossypiboma) with surrounding inflammatory mass.
• 2025-03-02 14:00 · Diagnosis — Retained foreign body after surgery (active)
• 2025-03-04 07:45 · Procedure — Exploratory laparotomy — foreign body removal

Step 2 — The merit score makes the fund-or-pass call

Feed the same facts to the Case Merit & Defensibility Score and it returns a defensibility read — grounded in the published payment-by-merit literature, not a guess:

• Payment probability: 83%
• Defensibility score: 17 / 100 (lower = weaker defense)
• Recommended posture: RESOLVE / MITIGATE — merit + record favor payment; manage exposure early
• Damages exposure: elevated

Why it lands there — the tool's own drivers:

• Breach evidence "strong" → base payment prior 72% (Studdert gradient)
• Documentation "gaps" raises paid-odds (×1.15) — record quality is a verified payout driver
• CANDOR / early-resolution candidate — see recommendation below (clinician/risk-led disclosure, not a litigation step).

Studdert NEJM 2006 payment-by-merit gradient + Peters 2008 merit/payout synthesis; decision support, not a dollar valuation.

Step 3 — Every fact is traced to its Bates page

When you build the demand, prep the deposition, or write the report, you need the page, not just the fact. The Bates-Cited Record Search stamps each record and ties every answer to its source:

• OKAFOR_000203 — OPERATIVE NOTE 02/11/2025 Laparoscopic appendectomy. Sponge and instru
• OKAFOR_000260 — RADIOLOGY 03/02/2025 CT abdomen/pelvis: retained surgical sponge (goss
• OKAFOR_000274 — OPERATIVE NOTE 03/04/2025 Exploratory laparotomy with removal of retai

Ask it "what did the operative note say about the sponge count?" and it returns the answer with the exact Bates page — no hunting through a 2,000-page production.

The point

Three tools, one synthetic case, under a minute — a screened case with a defensibility number, a timeline with the alteration flagged, and every fact cited to its page. No signup, no card, nothing retained past 30 days, and nothing you upload is used to train AI models.

Run your next intake through them free free: https://medicalai.law/free-tools

This teardown uses a synthetic, de-identified fact pattern; all tool output is genuine but the case is fictional. The tools are decision support for attorneys and experts, not legal advice or a dollar valuation.

← All posts · About John Mahoney · Free Daubert tool · Start a 14-day free trial