Medical Malpractice Statistics: Payments, Claims, and Data Limits

Stethoscope beside a sealed records folder and x-ray film on a desk in a medical exam room

The National Practitioner Data Bank (NPDB) held about 529,000 medical malpractice payment reports as of December 2025. That federal count covers payments made for health care practitioners since the data bank began. A separate American Medical Association survey found that 31.2% of physicians had been sued at some point in their careers by 2022.

Both numbers are real, but they measure different things. Medical malpractice statistics often blur payments, claims, lawsuits, and proven negligence into one idea. They are not the same, and the gap between them is large.

This page keeps each unit separate. We checked every figure against the NPDB and AMA source pages on November 10, 2026.

Key Takeaways

Key NPDB Figures

The NPDB publishes cumulative totals on its What Is the NPDB? page. The figures below are as of December 2025.

MeasurePeriodFigure
Medical malpractice payment reportsCumulative to Dec. 2025About 529,000
Adverse action reportsCumulative to Dec. 2025About 1,354,000
Judgment or conviction reportsCumulative to Dec. 2025About 45,000
Query responses2025More than 15.9 million
New reports received2025More than 71,800

Adding the three report types gives roughly 1.93 million reports. That sum is our calculation. It describes how much the data bank holds, not how much misconduct occurred.

The NPDB is a workforce tool. Hospitals, licensing boards, and other authorized bodies query it when they credential, license, or hire practitioners. The 15.9 million query responses in 2025 show how heavily it is used for that purpose.

What a Payment Report Is

The NPDB defines a medical malpractice payment report as a report of payments made for the benefit of a health care practitioner relating to a written claim or judgment for medical malpractice. Entities are required by law to submit these reports.

That definition sets clear limits on what the count can show:

So "529,000 malpractice cases" would be a misleading way to restate the number. The accurate phrase is "about 529,000 medical malpractice payment reports."

A payment also does not equal a finding of negligence. Many payments settle disputed claims without any court ruling on fault. That is our reading of how settlements work, not a statement printed on the NPDB page, and it is why careful medical malpractice statistics never call a payment proof of an error.

Why the NPDB Misses Most Claims

The AMA's 2023 study, Medical Liability Claim Frequency Among U.S. Physicians, explains the blind spot directly. The NPDB excludes claims where no indemnity payment was made. Citing closed-claim data from the Medical Professional Liability Association, the study says 72% of claims do not involve an indemnity payment.

The same study reports what happened to claims that closed between 2016 and 2018:

Outcome of Closed Claims, 2016 to 2018Share
Dropped, dismissed, or withdrawn65%
Decided by a trial verdict6%
Trial verdicts won by the defendant89%

The AMA author draws a careful conclusion from these figures: getting sued is not necessarily a sign of medical error. Most claims never reach a payment, and most that reach a verdict end in the physician's favor.

This is the biggest gap in many pages about medical malpractice statistics. They cite a payment database, then make claims about all malpractice. The data cannot support that leap.

How Often Physicians Are Sued

The AMA study uses its Physician Practice Benchmark Surveys from 2016 through 2022. Each year, 3,500 physicians completed the survey. Unlike the NPDB, the survey counts claims whether or not they were paid.

Measure20162022
Physicians sued in the previous yearAbout 2.4%1.8%
Physicians ever sued in their careers34.0%31.2%
Claims ever filed per 100 physicians6861

The short-term risk is low. Fewer than 2 in 100 physicians were sued in the 12 months before the 2022 survey. The long-term risk is much higher, because exposure builds over a career. The author notes that changes in the one-year figure over time were not statistically significant.

The survey has limits of its own. It covers physicians who give at least 20 hours of patient care a week, have finished residency, and are not federal employees. Response rates ranged from 31% to 38%, and the results are weighted to represent that national population. Claims are self-reported, so the survey cannot confirm how each one ended.

Age matters for the same reason. The study found that 9.5% of physicians under 40 had ever been sued, compared with almost half of physicians over 54.

Medical Malpractice Statistics by Specialty

Specialty creates the widest differences. Surgical fields carry the highest risk, and internal medicine subspecialties the lowest.

SpecialtySued in Prior YearEver SuedClaims per 100 Physicians
General surgery7.9%59.3%193
Obstetrics and gynecology3.2%62.4%152
Allergy and immunology0% in sample7.1%Not cited here

The AMA also reports that 8.0% of hematologists and oncologists and 8.6% of physicians in endocrinology and diabetes had ever been sued. Those are among the lowest rates in the study.

Gender showed a gap too. In the survey, 36.8% of male physicians had ever been sued, compared with 23.8% of women. The author notes that these differences shrink after adjusting for factors such as specialty.

Three Ways to Count Malpractice

Most medical malpractice statistics come from one of three places. Each sees a different slice of the same system.

Source TypeWhat It CountsWhat It Misses
NPDB payment reportsPayments made for practitioners on written claims or judgmentsClaims that close with no payment
Physician surveys, such as the AMA'sClaims a physician reports, paid or unpaidPhysicians who do not respond; exact claim outcomes
Insurer closed-claim filesClaims an insurer handled to closureClaims at other insurers or self-insured systems

Notice that the AMA survey reports a far higher share of physicians facing claims than a payment count would suggest. That is expected. A survey catches every claim, while the NPDB only sees the minority that end in a payment.

Neither source tracks the patient's side directly. A preventable injury that never becomes a claim leaves no trace in either count. Researchers who study that gap use medical record reviews, which are a separate field of work with their own methods.

When you read medical malpractice statistics, the first question is always which of these three sources produced the number. The answer tells you what the figure can and cannot support.

What the Public Data File Covers

Researchers who want record-level data can use the NPDB's Public Use Data File. The latest release, updated August 2026, covers reports from September 1, 1990, through June 30, 2026.

The file includes selected variables from malpractice payment reports and adverse action reports. It does not include any information that identifies individual practitioners or reporting entities. Users must accept a data use agreement that bans attempts to re-identify anyone.

The NPDB updates the file four times a year, with each release generally available within two months of its cutoff date. Its online Data Analysis Tool covers the same span, from 1990 through June 30, 2026, with filters, maps, and graphs.

The NPDB flags two timing issues. Because entities have a 30-day reporting window, actions taken at the end of a quarter may not appear until the next quarter. The analysis tool's yearly results may also not match the quarterly public file. Note your extract date before comparing numbers.

Why Payment Dollars Need Context

A payment amount is not a clean measure of how serious an injury was. Settlements reflect future care, lost income, legal risk, insurance limits, and the state where a case was filed. Two similar injuries can settle for very different sums.

Long-term comparisons also need inflation adjustment. A dollar paid in 1995 bought far more than a dollar paid in 2026. Any trend in payment totals should say whether it uses nominal or inflation-adjusted dollars.

Medians usually tell you more than averages. A small number of very large payments can pull an average far above what a typical claim receives. Any medical malpractice statistics that report only an average payment deserve a second look.

What These Data Cannot Show

Neither the NPDB nor the AMA survey can tell you:

Each of those questions needs a different source, such as medical record reviews, court data, insurer files, or patient-safety studies.

How to Use These Numbers Responsibly

Name the unit, period, and source every time. "The NPDB held about 529,000 medical malpractice payment reports as of December 2025" can be checked. "There have been 529,000 malpractice cases" cannot.

When you compare states, use a denominator such as population or physician count. Raw counts mostly reflect state size, so a large state will almost always lead a raw ranking.

The same discipline applies across legal data. Like divorce statistics, malpractice payment counts are administrative records with built-in gaps. And like patent litigation statistics, they track filings and outcomes, not the merits of every dispute. For the business side of the profession, see the legal industry statistics by firm size.

How We Checked the Data

We used two primary publishers. The NPDB supplied cumulative report counts, definitions, and public file details. The AMA's April 2023 study supplied survey-based claim frequency and closed-claim outcomes. We opened each page or file on November 10, 2026, and matched every figure to the text.

Our only calculation is the combined report total of about 1.93 million. Rounded NPDB counts are labeled as approximate. The AMA's closed-claim shares come from industry data cited in that study, and we attribute them that way.

Frequently Asked Questions

How Many Malpractice Payment Reports Are in the NPDB?

The NPDB held about 529,000 medical malpractice payment reports as of December 2025.

What Share of Doctors Get Sued?

In the AMA's 2022 survey, 31.2% of physicians said they had been sued at some point in their careers.

Does a Payment Report Prove Negligence?

No. Many payments settle disputed claims without any court finding of fault.

Which Specialties Are Sued Most Often?

OB/GYNs and general surgeons. About 62.4% and 59.3% of them, respectively, have ever been sued.

How Often Do Malpractice Claims Pay Out?

Industry data cited by the AMA show that 72% of claims involve no indemnity payment.

Can the Public Identify Doctors in NPDB Data?

No. The public use file contains no information that identifies practitioners or reporting entities.

Resources

These sources hold the counts, definitions, and survey results used above.

Verdict Point provides general statistical information, not legal or medical advice.