Medical Insurance Fraud Detection
This dataset contains detailed synthetic records of medical insurance claims, including patient demographics, provider information, claim amounts, service dates, and labeled indicators of fraudulent activity. Designed for machine learning and analytics, it enables robust research and development of fraud detection models in healthcare and insurance. The dataset supports granular analysis of claim patterns, provider behaviors, and patient demographics to identify and prevent fraudulent claims.
Sample rows
preview · 8 of 200 rows · all 20 columns| claim_idstring | claim_statusstring | claim_amountfloat | is_fraudboolean | hospital_idstring | patient_idstring | provider_idstring | claim_datedate | admission_datedate | discharge_datedate | diagnosis_codestring | procedure_codestring | approved_amountfloat | service_start_datedate | service_end_datedate | patient_ageinteger | patient_genderstring | provider_specialtystring | billing_codestring | num_servicesinteger |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CLM00001 | approved | 185 | false | blank | PT001 | PVD01 | 2023-02-15 | blank | blank | E11.9 | 99213 | 166.5 | 2023-02-15 | 2023-02-15 | 35 | female | Family Practice | B99213 | 1 |
| CLM00002 | approved | 130 | false | blank | PT002 | PVD02 | 2023-02-16 | blank | blank | F32.0 | 90834 | 117 | 2023-02-15 | 2023-02-15 | 27 | male | Psychiatry | B90834 | 1 |
| CLM00003 | approved | 2700 | false | HOSP001 | PT003 | PVD03 | 2023-02-17 | 2023-02-15 | 2023-02-17 | I10 | 99223 | 2250 | 2023-02-15 | 2023-02-17 | 76 | female | Cardiology | B99223 | 2 |
| CLM00004 | approved | 75 | false | blank | PT004 | PVD04 | 2023-02-18 | blank | blank | M54.5 | 97110 | 75 | 2023-02-18 | 2023-02-18 | 44 | male | Physiotherapy | B97110 | 1 |
| CLM00005 | denied | 110 | false | blank | PT005 | PVD05 | 2023-02-19 | blank | blank | E66.9 | 99401 | 0 | 2023-02-19 | 2023-02-19 | 53 | female | Internal Medicine | blank | 1 |
| CLM00006 | approved | 210 | false | blank | PT006 | PVD06 | 2023-02-20 | blank | blank | J06.9 | 99214 | 189 | 2023-02-20 | 2023-02-20 | 60 | male | Family Practice | B99214 | 1 |
| CLM00007 | approved | 18500 | false | HOSP002 | PT007 | PVD07 | 2023-02-21 | 2023-02-19 | 2023-02-21 | C34.1 | 77427 | 12000 | 2023-02-19 | 2023-02-21 | 84 | female | Oncology | B77427 | 2 |
| CLM00008 | pending | 120 | false | blank | PT008 | PVD08 | 2023-02-22 | blank | blank | E78.5 | 80053 | 100 | 2023-02-21 | 2023-02-21 | 41 | male | Endocrinology | B80053 | 1 |
| CLM00009 | approved | 135 | false | blank | PT009 | PVD09 | 2023-02-23 | blank | blank | F41.1 | 90837 | 121.5 | 2023-02-23 | 2023-02-23 | 29 | female | Psychiatry | B90837 | 1 |
| CLM00010 | pending | 17500 | false | HOSP003 | PT010 | PVD10 | 2023-02-24 | 2023-02-22 | 2023-02-24 | I21.4 | 92928 | 10000 | 2023-02-22 | 2023-02-24 | 79 | female | Cardiology | B92928 | 3 |
| CLM00011 | approved | 240 | false | blank | PT011 | PVD11 | 2023-02-25 | blank | blank | M25.561 | 20610 | 216 | 2023-02-25 | 2023-02-25 | 33 | female | Orthopedic Surgery | B20610 | 1 |
| CLM00012 | denied | 90 | false | blank | PT012 | PVD12 | 2023-02-26 | blank | blank | E03.9 | 84443 | 0 | 2023-02-26 | 2023-02-26 | 25 | male | Endocrinology | blank | 1 |
| CLM00013 | pending | 160 | false | blank | PT013 | PVD13 | 2023-02-27 | blank | blank | J45.909 | 94640 | 128 | 2023-02-27 | 2023-02-27 | 17 | female | Pediatrics | B94640 | 1 |
| CLM00014 | approved | 9200 | false | HOSP004 | PT014 | PVD14 | 2023-02-28 | 2023-02-26 | 2023-02-28 | S06.0X0A | 99285 | 5520 | 2023-02-26 | 2023-02-28 | 9 | male | Emergency Medicine | B99285 | 2 |
| CLM00015 | approved | 220 | false | blank | PT015 | PVD01 | 2023-03-01 | blank | blank | Z00.00 | 99396 | 209 | 2023-03-01 | 2023-03-01 | 52 | male | Internal Medicine | B99396 | 1 |
| CLM00016 | approved | 77 | false | blank | PT016 | PVD02 | 2023-03-02 | blank | blank | E11.65 | 83036 | 77 | 2023-03-02 | 2023-03-02 | 63 | female | Endocrinology | B83036 | 1 |
| CLM00017 | approved | 55 | false | blank | PT017 | PVD03 | 2023-03-03 | blank | blank | Z23 | 90471 | 49.5 | 2023-03-03 | 2023-03-03 | 31 | other | Family Practice | B90471 | 1 |
| CLM00018 | pending | 18500 | false | HOSP005 | PT018 | PVD04 | 2023-03-04 | 2023-03-01 | 2023-03-04 | S72.001A | 27130 | 13000 | 2023-03-01 | 2023-03-04 | 87 | male | Orthopedic Surgery | B27130 | 3 |
| CLM00019 | denied | 85 | false | blank | PT019 | PVD05 | 2023-03-05 | blank | blank | F41.9 | 99406 | 0 | 2023-03-05 | 2023-03-05 | 23 | female | Psychiatry | blank | 1 |
| CLM00020 | approved | 65 | false | blank | PT020 | PVD06 | 2023-03-06 | blank | blank | M79.7 | 97140 | 58.5 | 2023-03-06 | 2023-03-06 | 47 | male | Physiotherapy | B97140 | 1 |
What the 200 rows show
from the 200-row sampleDenied (claim status) stands out: 12 of its 23 rows have is_
- 9%is_
fraud = true - 128.5median claim_
amount - 4patient genders
- 18provider specialties
- 25providers
- 51billing codes
Denied (claim status)'s mean claim_
Median 128.5, from 48.0 to 128,000.
- string 10
- integer 2
- float 2
- date 5
- boolean 1
Columns
20 columns in four groups| column | type | description | example |
|---|---|---|---|
| Text 10 columns | |||
claim_id | string | Unique identifier for each medical insurance claimunique | CLM00001 |
patient_id | string | Unique identifier for the patient associated with the claim | PT001 |
provider_id | string | Unique identifier for the healthcare provider submitting the claim | PVD01 |
diagnosis_code | string | Primary diagnosis code (e.g., ICD-10) associated with the claim | E11.9 |
procedure_code | string | Primary procedure code (e.g., CPT/HCPCS) billed in the claim | 99213 |
claim_status | string | Status of the claim (e.g., submitted, approved, denied, pending)submitted · approved · denied · pending | approved |
patient_gender | string | Gender of the patientmale · female · other · unknown | female |
provider_specialty | string | Medical specialty of the provider submitting the claimoptional | Family Practice |
hospital_id | string | Unique identifier for the hospital or facility where services were renderedoptional | HOSP001 |
billing_code | string | Billing code used for the claim (may be different from procedure code)optional | B99213 |
| Numbers 4 columns | |||
claim_amount | float | Total monetary amount claimed for reimbursement0 or more | 185 |
approved_amount | float | Amount approved by the insurance company for payment0 or more · optional | 166.5 |
patient_age | integer | Age of the patient at the time of claim submission0 to 120 | 35 |
num_services | integer | Number of services or procedures billed in the claim1 or more | 1 |
| Dates and times 5 columns | |||
claim_date | date | Date when the claim was submitted | 2023-02-15 |
admission_date | date | Date when the patient was admitted for treatment (if applicable)optional | 2023-02-15 |
discharge_date | date | Date when the patient was discharged from treatment (if applicable)optional | 2023-02-17 |
service_start_date | date | Date when the medical service began | 2023-02-15 |
service_end_date | date | Date when the medical service ended | 2023-02-15 |
| True or false 1 column | |||
is_fraud | boolean | Label indicating whether the claim is fraudulent (true) or legitimate (false) | false |
15 SQL practice questions for the Medical Insurance Fraud Detection
Written for this data and checked against it: 12 SQL and 3 Power BI exercises from beginner to advanced. Try the first three free; the practice kit has all 15, a starter .sql file and an answer key with every expected result.
Denied claims over 5000
Return claim_id, provider_id and claim_amount for every claim whose claim_status is 'denied' and whose claim_amount is greater than 5000. Sort by claim_amount from highest to lowest, and use claim_id ascending to break ties.
Tables: medical_insurance_fraud_detection · your result: 3 columns, 13 rowsDistinct provider specialties
Return each distinct provider_specialty found in the claims as a single column named provider_specialty, sorted alphabetically.
Tables: medical_insurance_fraud_detection · your result: 1 columns, 18 rowsClaims per status
Count the claims for each claim_status. Return claim_status and claim_count, sorted by claim_count from highest to lowest, then by claim_status alphabetically.
Tables: medical_insurance_fraud_detection · your result: 2 columns, 3 rows
- 04Average claim and approvals by genderBeginner
- 05Fraud-flagged claims by specialtyPower BI
- 06Providers with at least 10 claimsIntermediate
- 07Claim size buckets and fraudIntermediate
- 08Claims above the overall averageIntermediate
- 09Monthly claims and approval rateIntermediate
- 10Specialty average vs overall averagePower BI
- 11Top providers by total claimedAdvanced
- 12Days since previous claimAdvanced
- 13Running total of approved amountsAdvanced
- 14Claims far above specialty averageAdvanced
- 15Share of total claimed by statusPower BI
Get the practice kit: the full dataset, all 15 exercises, a starter .sql and the answer key. One-time $3.99.
Get the practice kit · $3.99Use it for
An insurance dashboard
The is_
fraud rate, claim_ amount by claim_ status and a breakdown of hospital_ id. Excel, Power BI or Tableau. Why do 17 of 200 rows have is_
fraud = true? A root-cause class exercise
Hand out the rows and one question. The answer is in the data, not in the brief.
- Claims200CLM00001185approvedCLM0002324500deniedCLM0003324000approved
A software demo
Believable claims with patient_
id, provider_ id and claim_ date to fill a screen in front of a buyer.
blueprint · medical-insurance-fraud-detection
Behind this dataset
Same schema. As many rows as you need.
These 200 rows came out of a blueprint — 20 columns with generation rules behind each one. Open it in Data Factory to retune a column, add your own, wire in foreign keys, and run it at the size you actually need.
- Claims include varying procedures, costs, and patient demographics
- Fraudulent claims follow patterns like duplicate billing or upcoding
- Legitimate and fraudulent claims distributed 90/10
- Temporal features mimic real-world claim submission timelines
1 credit per row. Preview 20 rows free. Buy rows for the full run.
- Exports
- CSV, JSON, JSONL, Parquet, SQL, Excel, TSV, XML
- Licence
- yours to use, including commercially
- API slug
- medical-insurance-fraud-detection