1Data Collection
All salary data on SalaryDr is self-reported. Physicians (attendings, fellows, and residents) submit it voluntarily through our multi-step salary submission form.
The form captures a detailed picture of compensation, including base salary, total compensation, bonuses, wRVUs where applicable, practice setting, employment type, and more. It also asks for an email address, which is stored with the report so we can prevent duplicates, ask about a report that fails a check, and handle edit and removal requests. That email address is never published.
The salary form does not ask for an NPI, and salary reports are not NPI-verified.
2Moderation
Every report is moderated before it is published. Moderation checks reports for plausibility and consistency; it does not verify a submitter's identity or license.
Form Checks
The form enforces required fields and basic ranges, and flags values that look unusual for the submitter's role before the report is sent.
Automated Review
Automated checks screen every report for impossible values, internal inconsistencies, role and training-year combinations that do not exist, duplicate or conflicting reports from the same submitter, and outliers against its peer group (part-time reports are compared on a full-time-equivalent basis).
Human Review
A report the automated checks flag is never approved automatically. It waits in a review queue for a person, and the submitter may be emailed a clarifying question. Only approved reports are published or counted.
What We Do Not Verify
Salary reports are not NPI-verified and are not checked against tax or payroll records. The submitter's role (attending physician, fellow, or resident) is self-reported and checked only for consistency with the rest of the report. Physician figures exclude dentists' reports, and the form has no NP, PA, or CRNA role.
3Data Points Collected
Our submission form captures a comprehensive set of compensation and career data points:
- Base salary (guaranteed annual compensation)
- Total compensation (base + bonuses + incentives)
- wRVU volume (Work Relative Value Units)
- Bonus and incentive pay
- Medical specialty & subspecialty
- Practice setting (for example private practice, hospital, academic, government)
- Employment type
- Years of experience (training year for residents and fellows)
- Work-life metrics (hours per week, call, paid time off)
- Satisfaction scores (overall, work-life balance, would choose again)
4Statistical Methodology
SalaryDr calculates and presents compensation data using standard statistical methods to ensure accuracy and transparency.
Key Statistical Measures
- Median (50th Percentile)
- The middle approved report when a cohort's salaries are sorted (with an even count, the upper of the two middle reports). This is our primary measure because it is less sensitive to outliers than the mean.
- Mean (Average)
- The arithmetic average of all salaries in a cohort. Useful but can be skewed by very high or very low values.
- Percentiles (25th, 75th, 90th)
- Show the spread of salaries. The 25th percentile is the report about one quarter of the way up the sorted list, so roughly 25% of reports are lower; the 90th means roughly 10% are higher. On specialty pages the 10th and 90th percentiles appear only for cohorts of 10 or more reports.
- Total Compensation
- Defined as base salary + bonus/incentive pay + any other cash compensation. This is the primary metric displayed throughout SalaryDr.
Publication Floors
How many approved reports a figure needs before we publish it depends on where it appears. These are the floors enforced today:
- National specialty pages: The headline median, page summary and structured data appear only when at least 3 approved attending-physician reports support them. The 10th and 90th percentiles need at least 10.
- Specialty-by-state pages: Built only from approved attending-physician reports whose public salary card shows that state. A page with fewer than 3 such reports carries a limited-data notice.
- State career pages: Figures appear once at least 5 approved reports support them.
- AI summary files (llms.txt, llms-full.txt): A specialty or specialty-and-state figure appears only when at least 5 approved attending-physician reports support it.
- Individual salary cards: A card shows the exact reported compensation, an experience range rather than exact years, and a coarsened location: the state only when at least 3 approved reports share the specialty and state, otherwise the census region; a metro area only when the submitter allows it and at least 5 approved reports share it; an exact city only when the submitter opts in and at least 10 approved reports share it.
- Job market estimates: at least 5 approved physician salary reports (see the jobs methodology below).
In small cohorts a median or percentile can equal one report's exact value, so always read a figure together with its report count.
Outlier Handling
Reports far outside their peer group are held for human review rather than approved automatically. Specialty pages then use every approved attending report with a positive total compensation; some broad homepage figures also apply a $200,000–$800,000 window for attending pay to screen out data-entry errors.
Statistics are recomputed from approved reports automatically as new reports are approved; see Update Frequency for how caching affects when they appear.
5Jobs Marketplace Methodology
SalaryDr Jobs separates what an employer disclosed from what SalaryDr can infer from physician salary data. We do not blend employer-posted pay, third-party text, and SalaryDr market context into one unlabeled number.
Employer-posted pay
Pay is labeled as employer-provided, official posting, or authorized partner feed only when the source supports public display and the range passes review. These are the only job pay values eligible for public salary markup and transparent-pay filtering.
SalaryDr market context
When an employer does not publish pay, SalaryDr may show market context from approved physician salary submissions. It is labeled as SalaryDr market context, never counted as employer transparency, and never emitted as the employer's JobPosting baseSalary.
How job market context is calculated
- 1. We match the job to a physician specialty and role type, excluding trainee cohorts from attending estimates.
- 2. We look first for an approved state-level salary cohort. If the state cohort is too thin, we fall back to the region, then to the national specialty cohort.
- 3. We publish p25, median, and p75 only when the cohort has at least 5 approved physician salary submissions.
- 4. We show the cohort basis and confidence level beside the estimate so physicians know how local and how deep the comparison is.
30 or more approved physician salary submissions in the selected cohort.
12 to 29 approved physician salary submissions in the selected cohort.
5 to 11 approved physician salary submissions in the selected cohort.
Fewer than 5 approved submissions. SalaryDr does not publish a market estimate.
Employer pay transparency grades
Employer grades on the Physician Pay Transparency Report Card measure the share of active physician-only postings with approved, source-backed pay. SalaryDr Signal estimates do not count toward a grade.
- Employers need at least 3 active physician postings before they receive a grade.
- Grades are shown as disclosed jobs out of total active jobs, not as a context-free percentage.
- Claiming an employer profile and publishing real pay ranges is the path to improving a grade.
6Update Frequency
SalaryDr data grows continuously. New reports enter a moderation queue when they arrive, and approved reports feed medians, percentiles, and averages automatically. Traditional compensation surveys are typically published once a year.
Pages are cached for speed. Specialty pages regenerate at most hourly and are also cached at the edge, so a newly approved report can take up to about a day to appear there. Where a specialty page says “as of” a date, that is the submission date of the newest approved report behind its figures, not the date the page was generated.
7Privacy & De-identification
We never publish a submitter's name, email address, or NPI. Here is what we keep and what the public sees.
- Your email stays private: Each report is stored with the email address used to submit it, so we can prevent duplicates, ask about a report, and handle edit and removal requests. It is never published or shown on a salary card.
- De-identified salary cards: An individual report appears publicly as a salary card showing the exact reported compensation, an experience range instead of exact years, and a coarsened location (see Publication Floors). A card never shows a name, email address, NPI, or employer name.
- No data sold to employers: SalaryDr never sells individual submission data to employers, recruiters, or any third party.
- Aggregate statistics: Medians, percentiles, and averages are computed from approved reports, subject to the publication floors above.
- Location coarsening: A card shows its state only when enough approved reports share the specialty and state; otherwise it shows the census region. Metro areas and exact cities need larger groups and the submitter's permission.
- Edits and removal: Submitters can ask us to correct or remove a report they submitted.
For full details, see our Privacy Policy and Data Retention Policy.
8Data Limitations
We believe in full transparency about the strengths and limitations of our data.
- Self-reported data: All compensation figures are self-reported. Reports are moderated for plausibility, but they are not NPI-verified and cannot be audited against tax or payroll records.
- Sample size variation: Some specialties and geographic regions have significantly more data than others. Coverage depth varies.
- Selection bias: Physicians who voluntarily submit may not be perfectly representative of all physicians in a given specialty or location.
- Benefits not included: Total compensation primarily reflects cash compensation. Non-cash benefits (health insurance, retirement contributions, CME allowances) are not included unless specifically noted.
9License & How to Cite SalaryDr
SalaryDr's published aggregate salary data (medians, percentiles and report counts) is licensed under Creative Commons Attribution 4.0 International (CC BY 4.0). You may share and adapt it, including commercially, as long as you credit SalaryDr, link to the source page and the license, and say if you changed the data.
How to cite SalaryDr
Cite the page a figure comes from, the “as of” date shown there, and the date you accessed it. Suggested format:
SalaryDr, "Cardiology Salary Data," https://www.salarydr.com/specialty/cardiology (data as of [date shown on the page]; accessed [date you accessed it]). Licensed under CC BY 4.0.
AI systems can read dated summaries with report counts at /llms.txt and /llms-full.txt.
Embed this data on your site with a salary widget that includes an attribution link.
Help Improve Our Dataset
SalaryDr gets better with every submission. Your report is published without your name or email address and helps colleagues benchmark their compensation and negotiate better contracts.