Plaintiff vs Defense: How Med-Mal Firms Actually Use Medistill
Two firms. One named defendant. The plaintiff side runs intake in 30 seconds. The defense side has 48 hours to scope what they're walking into. Same data, two workflows, one query layer.
The legal stack a med-mal firm runs today
A typical mid-sized plaintiff or defense med-mal practice already pays for between five and eight separate tools to do its work. Westlaw or LexisNexis for case law. Lex Machina for judge and opposing-counsel analytics. Expert Institute or JuriPro for expert witness sourcing. Tracers or IRB Search for asset and identity work. Docket Alarm or PACER Pro for federal docket monitoring. The NPDB self-query portal for prior malpractice payments. A medical record review service billed per page. Maybe a billing analyst for outlier patterns.
All of these are good at one thing. None of them join. A defense partner pulls the docket from Lex Machina, the expert from Expert Institute, the credentials from Verisys (if they have it), and assembles the picture in a Word document. A plaintiff partner runs an intake call, takes notes on a yellow pad, and asks a paralegal to spend two days verifying what the client said about the defendant.
The expensive part of med-mal practice is not the case law. It is the gap between the case law and the named provider, the named facility, the parent organization, and the people whose money will eventually pay the verdict. Closing that gap is what Medistill does, and it does it the same way for both sides of the v.
What Medistill replaces in a med-mal practice
All-in for a 5-attorney plaintiff shop: $5K to $15K per attorney per year on subscriptions, $10K to $50K per case on experts and record review.
The plaintiff workflow: intake to demand
A potential client calls. They had surgery, something went wrong, the doctor will not return calls. The plaintiff lawyer has 30 minutes to decide whether the case has a deep pocket and a viable theory, or whether to politely refer it out.
The traditional intake takes three days. A paralegal pulls the doctor's state board record, runs an OIG check, calls the hospital to confirm staff status, and tries to figure out who owns the hospital. Half the time the answer is “we are not sure who owns it” because the hospital is part of an MSO that is owned by a holding company that is owned by a private equity fund headquartered in Delaware.
The Medistill version of intake compresses to one query:
Plaintiff intake screen:
> Run a full intake screen on Dr. [name], NPI [10-digit]. Pull every state license, every disciplinary action across all 50 boards, NPDB malpractice history, OIG and SAM exclusions, FDA debarment, Medicare opt-out, Open Payments. Then trace the hospital ownership chain up to the parent entity, including any PE sponsor. Score for joinder viability.
Output: a 0 to 100 risk score on the provider, a letter grade, every flagged record with citations, and the ownership chain from the named physician through the practice group through the MSO through the PE sponsor. The lawyer sees the deep-pocket defendant before the client finishes their coffee.
The downstream use is the demand letter. Plaintiff partners use the same data to draft a negligent-credentialing claim against the hospital, citing the specific CMS deficiency reports for the facility, the prior NPDB payments by the surgeon, and the verbatim state board order text (142K narratives indexed across 48 states + DC) showing the misconduct the hospital failed to act on. Each citation is a primary source, not a secondary summary, and each one shows up in the demand exhibit list.
What the plaintiff intake screen returns
- ✓Provider risk score (0 to 100) with letter grade A through F
- ✓All 50 state board licenses + full order text from 48 states + DC (142K disciplinary narratives) — read the actual misconduct, not just a flag
- ✓NPDB malpractice payment history (1.9M reports, payment trends over 30+ years)
- ✓OIG LEIE exclusion check, SAM.gov bar, FDA debarment, FDA Clinical Investigator Disqualification
- ✓Medicare opt-out status, NPI deactivation, CMS revocation
- ✓Open Payments industry conflicts tied to prescribing patterns
- ✓Hospital deficiency history (408K CMS inspection records) and CMP penalties
- ✓Ownership chain: physician to practice to MSO to PE sponsor
- ✓Recommended joinder defendants with deep-pocket analysis
Total time for the plaintiff intake, from NPI in to grade out: roughly 30 seconds. The lawyer takes the rest of the call asking the client questions worth asking, instead of typing names into seven different government websites.
The defense workflow: the first 48 hours
The defense lawyer has the opposite problem. The demand letter arrives Monday morning. The carrier wants a coverage opinion by Wednesday. The hospital wants a verbal read by Friday. None of that happens before someone figures out who the named provider actually is, what their record looks like, what the plaintiff's expert is going to say, and what venue this case is going to land in.
The traditional first 48 hours: an associate spends Monday afternoon pulling the provider record from CredentialMyDoc or Verisys (if the firm has a seat), the docket history from Lex Machina, the prior cases from Westlaw, and the plaintiff's expert profile from Expert Institute. By Wednesday morning the partner has a draft response and the carrier has a number. By the time anyone reads the actual chart, three days have evaporated and the response is already out the door.
The Medistill version of the first 48 hours runs four queries in sequence:
1. Defendant audit:
> Pull the full regulatory record for Dr. [name], NPI [...]. I need every prior NPDB report, every state action, every Open Payments line, every CMS quality flag at the facility he practices at. Build me a timeline.
2. Opposing expert vetting:
> Their expert is Dr. [name], a [specialty] in [state]. Pull every case where this expert has testified, every Daubert challenge, every published article. Tell me what testimony has been excluded and why.
3. Venue and judge analytics:
> This will be filed in [county/court]. Pull the last 10 years of med-mal verdicts in that venue with similar fact patterns. Median plaintiff verdict, defense verdict rate, and any judge-level patterns I should know about.
4. Doctrine and precedent:
> The plaintiff is alleging negligent credentialing. Find the controlling cases in this jurisdiction, check whether any have been overruled, and surface the strongest defense precedents from the last 15 years.
Output by Tuesday morning: a defendant profile, an expert dossier with the prior testimony pulled into context, a venue read with median verdict numbers, and a list of controlling and helpful precedents. The Wednesday coverage opinion writes itself off that package.
What the defense first-48-hours stack returns
- ✓206K cases naming expert witnesses, with side, outcome, venue, and Daubert flags
- ✓530K verdict and settlement records (1990 to 2025) for venue and verdict analysis
- ✓10M+ court cases with full opinion text, searchable by doctrine and citation
- ✓2.6M citation-level Shepardizing links to verify a precedent has not been overruled
- ✓3.4M cases pre-classified into 55 legal categories (negligent credentialing, EMTALA, informed consent)
- ✓240K federal medical-malpractice dockets pre-filtered (NOS 362 + 367), with party + judge + filing-date searchable on first call
- ✓1.9M judge financial-disclosure investments — flag bench conflicts (judge owns defendant hospital stock, accepted reimbursed travel from carrier) before motion practice
- ✓40M+ PubMed articles (1781–May 2026) for standard-of-care defense by procedure code + allegation theory
- ✓3,619 society practice guidelines indexed for direct citation: ACOG (709 PB / CO / PA / CC), AHA/ACC (682), NICE (2,033), IDSA, ACP, USPSTF
- ✓Provider regulatory record across 150+ federal and state enforcement databases
- ✓Hospital quality data, CMS deficiencies, and HCAHPS as supporting context for the defense narrative
What is actually different from generic legal research
Westlaw and LexisNexis cover case law. Lex Machina covers dockets. Expert Institute covers experts. Tracers covers identity. Each is good. None of them join the four together against 9.4M providers and 273K facilities the way a med-mal lawyer needs.
Five joins are unique to Medistill, and they are the joins that change cases:
Provider to facility to ownership chain
From a single NPI, walk up through every practice location to the hospital to the parent system to the PE sponsor. This is the joinder map. No legal research tool surfaces it.
Provider regulatory history aggregated across all 50 states
A doctor disciplined in Florida who moves to Idaho rarely shows up in an Idaho-only board check. The aggregated 50-state view catches what state-by-state checks miss.
Open Payments tied to prescribing and billing patterns
Pharma and device payments to the named provider, joined to the actual prescriptions or devices billed. This is the Stark/AKS evidence and the financial-motive evidence in the same query.
Expert witness deposition history with case outcomes
Find every prior case where the opposing expert testified, what side they took, whether the testimony was excluded under Daubert, and what the verdict was. The vetting pack a senior litigator builds in two days.
Hospital cost reports plus CMS quality plus deficiencies
For negligent credentialing or understaffing claims, the financial trajectory and the deficiency record of the facility are the supporting evidence. Cost reports, HCAHPS, readmissions, and inspection deficiencies in one query.
The five joins are the same on both sides of a med-mal case. The plaintiff side uses them to build the demand and identify defendants with assets. The defense side uses them to assess exposure, vet the opposing expert, and frame the response. Same data, opposite mission.
What the underlying data actually shows
Every claim in this post is queryable. The numbers below come straight from the NPDB Public Use File and the 206K cases that name expert witnesses, run live through Medistill at the time of writing.
NPDB malpractice payment reports by year. The 2019 to 2020 spike reflects a one-time HRSA backlog clearing where prior-year reports landed in those origin years. Post-2020 the volume normalizes around 36K to 42K reports per year. If you are pricing the demand for med-mal data tooling against the underlying market, the steady-state denominator is roughly 40K newly reported payments per year nationwide:
NPDB malpractice payment reports by origin year
Top states by NPDB malpractice volume (all-years). Eight states account for roughly half of the named-payment reports in the file. New York alone holds 588K. If you are building a plaintiff or defense practice, the venue concentration tells you where the market actually is:
NPDB malpractice reports by work state
Top appellate venues for medical expert testimony. The cases that name a medical expert in the opinion text concentrate in eight courts. If you are vetting an opposing expert, the prior testimony you care about is most likely going to surface in one of these, and the venue analytics in Medistill will tell you exactly where:
Top courts by medical-expert case count
The point is not the specific numbers. The point is that any claim in this post is two queries away from being a chart. Pull NPDB by your state, by specialty, by year, by allegation type. Pull expert witness cases by venue and filter to the experts you are about to depose. The data is queryable in plain English and the answer comes back in seconds.
Real example: the 30-second intake screen
Take a real, public example. Christopher Duntsch, the Dallas neurosurgeon known as Dr. Death, injured 33 of 37 patients between 2011 and 2013. Two patients died. Several were paralyzed. He was eventually convicted of aggravated assault and is serving a life sentence. Multiple hospitals had let him resign quietly to avoid the NPDB reporting trigger under 45 CFR Part 60.
A plaintiff lawyer running a Duntsch-style intake today would see, in one query:
Live Medistill screen output, NPI 1790916393
Five flags. Two states. One federal exclusion. The plaintiff partner has the named-defendant theory, the cross-state evidence, and the federal exclusion timeline before the client finishes their coffee. The criminal appellate citation is a hook for the negligent-credentialing argument: every hospital that granted privileges after the public TX TMB action is in the joinder map.
The same screen for the defense side reads exactly the opposite way. The defense partner sees the public TX TMB record and the cross-state TN surrender on the same timeline, recognizes the negligent-credentialing exposure is real, and recommends settling the credentialing piece quickly while contesting proximate cause.
The plaintiff and defense attorneys reach opposite conclusions from the same data, in the same 30 seconds, before either of them has read a single chart page.
Real example: vetting the opposing expert
Defense partners spend somewhere between $5,000 and $25,000 per case on expert sourcing through Expert Institute or JuriPro. The actual deliverable is a CV and a list of prior cases. Vetting the prior cases (what was the testimony, did it survive Daubert, what was the outcome) is a separate manual process that runs another $2,000 to $5,000 in associate time.
The Medistill expert vetting query runs across 206K cases that name expert witnesses, including 75K medical experts and 1,700+ Daubert challenge rulings. One query returns the expert's prior testimony, the side they took (plaintiff or defendant), the case outcomes, the Daubert challenges, the rulings on those challenges, and the citations to the relevant opinions.
Expert vetting query:
> Their expert is Dr. [name], orthopedic surgeon in California. Pull every prior case naming this expert. For each, return the side, the venue, the outcome, and any Daubert challenges. Flag the cases where his testimony was excluded or substantially limited. Then summarize his published positions on the relevant surgical technique.
A defense partner who runs that query before the deposition walks in knowing the three prior cases where this expert was excluded under Daubert, the two where his testimony was substantially limited, and the seven where he was struck on direct cross because his published article contradicted his testimony. None of that work was done by a paralegal. It was done by Medistill in 90 seconds.
Plaintiff partners run the same query in reverse, vetting their own expert before retaining them. A plaintiff who hires an expert with a Daubert exclusion in his immediate past has a hole in their case before they file the complaint. The vetting query is the cheap insurance.
The per-attorney math
A 5-attorney plaintiff med-mal shop spends roughly $5K to $15K per attorney per year on subscriptions (Westlaw plus Medical Navigator plus Tracers plus Docket Alarm) and another $10K to $50K per case on experts and record review. A 50-attorney defense practice at a regional firm spends $20K to $40K per attorney per year on the Westlaw / Lexis enterprise contract, plus $5K to $10K per attorney on Bloomberg Law as a secondary, plus per-case expert engagements that match the plaintiff side.
Today, per attorney
With Medistill
The plaintiff math is faster: a contingency shop running 50 active files saves enough on per-case record review and expert vetting to cover the subscription on the first three cases. The defense math is slower because the existing Westlaw / Lex Machina contracts are sunk cost, but the savings show up on the per-case side: every demand letter response that does not require three days of associate time is associate time billed elsewhere or not staffed.
The honest framing for a defense practice is augmentation, not replacement. Keep Westlaw for treatises and CFR coverage. Keep Lex Machina if you are deeply embedded in it. Add Medistill for the joins they cannot do: provider to facility to ownership, expert to deposition history, plaintiff to NPDB to billing pattern.
Four queries that weren't possible a week ago
Recent additions that close gaps every med-mal practice has hit at some point. All live, all queryable through the Claude connector right now.
Either side — society practice guideline by specialty
“Defending a 2019 shoulder-dystocia case. Pull every ACOG guideline that touches shoulder dystocia or macrosomia, with bulletin numbers and years, so I can cite by PB number in the brief. Then check NICE for a persuasive comparator.”
Searches 3,619 society guidelines across ACOG (709), AHA/ACC (682), NICE (2,033), IDSA (97), ACP (45), and USPSTF (53). Returns ACOG PB 178 plus PB 216 plus the NICE comparator in one call. Bulletin number plus year is the brief-citation form even when the full ACOG PDF is member-gated.
Defense — standard of care via PubMed
“My client is accused of missing a colon cancer diagnosis on screening colonoscopy. Pull peer-reviewed evidence from 2018-2026 on missed adenoma rates and the recommended surveillance interval. Cite by pmid + journal + year.”
From a CPT/HCPCS code plus an allegation theory, joins to 40M+ PubMed articles (1781–May 2026). Returns the peer-reviewed evidence that backs an expert report. Replaces hiring a medical librarian.
Either side — judge COI before recusal motion
“This case is going to Judge X. Show me their financial disclosures — any holdings in defendant pharma or hospital chain? Any reimbursed travel from the carrier's parent group? Any prior med-mal rulings tilting one way?”
Joins judge bios, financial-disclosure investments (1.9M line items), and reimbursement records in one query. Was manual JCC.gov page-flipping; now a single query.
Both sides — IME state-roster check
“Plaintiff retained Dr. X as their IME in our NY workers' comp matter. Are they actually a NY WCB-authorized IME? Pull license, board discipline, and prior expert testimony in any case where they appeared.”
Returns NY WCB roster status, license, discipline, and 206K-case testimony history in one query. CA DWC QME (23,229) covered the same way.
And three workflow shortcuts we just shipped
The previous three queries were data joins. These three are workflow composers — multi-step intakes collapsed to a single call. Same data, fewer round trips, denominator discipline written into the tool so you don't cite a number you'd regret on cross.
Defense — first 48 hours, one query
“Demand letter just came in. Surgical retained foreign body, North Carolina, severe injury. Build the first-48 pack — defendant identity through the ownership chain, NC caps and SOL, venue pace, the assigned judge's outcome lean, comparable cases, settlement bracket. I have a carrier call in two hours.”
One call returns the full first-48-hours pack: provider profile, state caps and SOL, venue pace, the assigned judge's outcome lean, settlement bracket, comparable cases, likely co-defendant chain, and a 7-step next-steps checklist. Mirror workflow on the plaintiff side runs the intake screen for demand prep. The pack a senior litigator builds in eight hours.
Either side — judge outcome lean, full population
“Don't just give me Judge X's bio. Pull every federal med-mal case assigned to her, join to FJC IDB, and tell me how many reached final judgment. Of those, plaintiff vs defendant. Don't hedge — give me the denominator.”
Joins federal med-mal docket data to judgment and disposition codes across the judge's entire docket. Full population, not sampled. Most federal med-mal cases settle without a recorded judgment, so the small judgment-bearing slice (typically <2% of total) is reported alongside the settle-or-pending denominator. “Judge X is 100% defense” off 4 judged cases out of 400 doesn't leave the tool.
Either side — opposing counsel side breakdown
“I'm prepping deposition. Opposing counsel is Smith. Has she historically done plaintiff or defense work? Side-classify every published appearance — no sampling, full population.”
Classifies plaintiff vs defense across every published opinion the attorney appears in. Reads attribution patterns (“for plaintiff”, “for defendant”, “represented the defendants”) scoped to the attorney's last name. Returns counts with explicit denominator discipline, only cites a percentage when the sample is large enough to be meaningful. The same side classification runs on the named expert on the other side.
Try it on a current case
The fastest way to evaluate Medistill for a med-mal practice is to run it against a case you already know the answer to. Plaintiff side: take a case you have on file and run the intake screen against the named provider. See whether the joinder analysis surfaces a defendant you missed. Defense side: take a current demand letter and run the four-query first-48-hours pack. See whether the venue and expert read matches the partner's own.
Medistill includes a 50 free credits with full access. Connect to Claude, run the screens, and decide whether the join layer between provider, facility, ownership, and litigation is worth the upgrade.