Insurance
20 columns
200 rows
8 formats

Medical Claims Fraud Scenarios

This dataset contains simulated health insurance claims with detailed patient, provider, and service information, including flags and scores for a variety of fraud scenarios. It is designed to support the development and evaluation of fraud detection algorithms, audit workflows, and compliance monitoring in healthcare insurance. The dataset enables analysis of both common and rare fraudulent patterns for improved anomaly detection.

opened 5 times
last updated 12 Jul 2025
by GoMask

The brief that made it

Training and benchmarking medical claims fraud detection models

Sample rows

preview · 8 of 200 rows · all 20 columns

claim_idstring	service_locationstring	claim_amountfloat	fraud_flagboolean	fraud_scenariostring	patient_idstring	provider_idstring	claim_datedate	service_datedate	diagnosis_codestring	procedure_codestring	fraud_scorefloat	audit_statusstring	reviewer_idstring	review_datedate	patient_ageinteger	patient_genderstring	provider_specialtystring	billing_codestring	notesstring
CLM0001	office	112.4	false	blank	PAT001	PRV1001	2024-02-15	2024-02-13	J01.90	99213	blank	cleared	REV101	2024-02-21	8	female	pediatrics	BILL0001	Routine pediatric checkup, no complications.
CLM0002	office	238.6	false	blank	PAT002	PRV1002	2024-03-10	2024-03-09	E11.9	83036	blank	pending	blank	blank	47	male	endocrinology	BILL0002	Diabetes management follow-up, routine labs ordered.
CLM0003	office	78.55	false	blank	PAT003	PRV1003	2024-01-18	2024-01-16	J00	99214	blank	cleared	REV102	2024-01-22	6	female	otolaryngology	BILL0003	Treated for acute sinusitis, no antibiotics prescribed.
CLM0004	inpatient	25100	true	upcoding	PAT004	PRV1004	2024-03-12	2024-03-10	S72.001A	27130	0.92	under_review	REV103	2024-03-15	72	female	orthopedics	BILL0004	Claim flagged for review due to high claim amount and complex procedure.
CLM0005	office	92.15	false	blank	PAT005	PRV1005	2024-02-20	2024-02-19	R05	99213	blank	cleared	REV104	2024-02-24	27	other	family_medicine	BILL0005	Cough and cold symptoms, patient advised rest and fluids.
CLM0006	office	136	false	blank	PAT006	PRV1001	2024-03-05	2024-03-04	L40.0	11100	blank	cleared	REV105	2024-03-12	38	female	dermatology	BILL0006	Skin biopsy performed for suspected psoriasis.
CLM0007	inpatient	18950	true	phantom billing	PAT007	PRV1006	2024-01-25	2024-01-24	C50.911	19303	0.96	confirmed_fraud	REV106	2024-01-31	77	female	oncology	BILL0007	Unusual billing activity for mastectomy; claim investigated for phantom billing.
CLM0008	telemedicine	64.5	false	blank	PAT008	PRV1007	2024-02-08	2024-02-08	G47.00	95810	blank	cleared	REV101	2024-02-15	35	male	sleep_medicine	BILL0008	Telehealth follow-up for sleep disorder, no medication change.
CLM0009	inpatient	8900	true	duplicate claim	PAT009	PRV1008	2024-03-16	2024-03-15	I21.4	92928	0.89	under_review	REV107	2024-03-18	82	male	cardiology	BILL0009	Possible duplicate billing detected for cardiac stent placement.
CLM0010	office	58.25	false	blank	PAT010	PRV1009	2024-01-07	2024-01-07	H65.3	69433	blank	pending	blank	blank	3	female	pediatrics	BILL0010	Well-child visit with ear infection diagnosis.
CLM0011	office	315	false	blank	PAT011	PRV1010	2024-03-03	2024-03-03	E11.9	99215	blank	cleared	REV102	2024-03-10	52	male	endocrinology	BILL0011	Complex diabetes follow-up, new complications discussed.
CLM0012	outpatient	129.8	false	blank	PAT012	PRV1011	2024-02-28	2024-02-25	M54.5	97110	blank	pending	blank	blank	45	female	physical_therapy	BILL0012	Physical therapy for lower back pain, initial session.
CLM0013	inpatient	18550.75	false	blank	PAT013	PRV1012	2024-02-12	2024-02-10	M16.11	27130	blank	under_review	REV108	2024-02-15	69	male	orthopedics	BILL0013	Hip replacement surgery, notes reviewed for completeness.
CLM0014	office	98.4	false	blank	PAT014	PRV1003	2024-01-30	2024-01-29	J06.9	99213	blank	cleared	REV104	2024-02-03	11	male	general_practice	BILL0014	Treated for upper respiratory infection, routine exam.
CLM0015	telemedicine	51	false	blank	PAT015	PRV1013	2024-03-14	2024-03-13	F41.9	90834	blank	cleared	REV109	2024-03-19	29	female	psychiatry	BILL0015	Psychotherapy session, anxiety disorder.
CLM0016	inpatient	21500	false	blank	PAT016	PRV1014	2024-01-20	2024-01-19	I25.10	92980	blank	under_review	REV110	2024-01-23	74	male	cardiology	BILL0016	Angioplasty performed, high value claim.
CLM0017	office	61.7	false	blank	PAT017	PRV1002	2024-02-24	2024-02-22	R51	99212	blank	pending	blank	blank	44	unknown	urgent_care	BILL0017	Patient presented with headache, no alarming symptoms.
CLM0018	outpatient	74.5	false	blank	PAT018	PRV1015	2024-03-01	2024-02-28	F32.9	90837	blank	cleared	REV107	2024-03-07	36	male	psychiatry	BILL0018	Routine outpatient visit for depression.
CLM0019	inpatient	10056.85	true	unbundling	PAT019	PRV1004	2024-02-03	2024-01-31	C61	52601	0.94	confirmed_fraud	REV111	2024-02-10	80	male	urology	BILL0019	Claim confirmed as fraudulent due to unbundling of services.
CLM0020	office	58	false	blank	PAT020	PRV1016	2024-01-11	2024-01-10	J02.9	99213	blank	cleared	REV103	2024-01-17	9	female	pediatrics	BILL0020	Routine sore throat visit; home care advised.

What the 200 rows show

from the 200-row sample

Inpatient (service location) stands out: mean claim_amount is 9,761, against 128.2 for the rest.

11%fraud_flag = true
126.1median claim_amount
4audit statuses
4patient genders
28provider specialties
0.94median fraud_score

Mean claim_amount by service_location200 rows

claim_amount200 rows, in bands of 5,000

Median 126.1, from 19.5 to 25,100.

fraud_scenario22 rows with a value · 178 left blank

upcoding8
phantom billing5
phantom_billing4
duplicate claim2
duplicate_claim2
unbundling1

20 columns by typefrom the column list below

string 13
integer 1
float 2
date 3
boolean 1

Columns

20 columns in four groups

blueprint · 20 columns
column	type	description	example
Text 13 columns
`claim_id`	string	Unique identifier for each insurance claimunique	CLM0001
`patient_id`	string	Unique identifier for the patient associated with the claim	PAT001
`provider_id`	string	Unique identifier for the healthcare provider submitting the claim	PRV1001
`diagnosis_code`	string	ICD code representing the primary diagnosis for the claim	J01.90
`procedure_code`	string	CPT or similar code representing the procedure performed	99213
`fraud_scenario`	string	Short description or code of the detected fraud scenario (e.g., 'upcoding', 'phantom billing', 'duplicate claim')6 scenarios · optional	upcoding
`audit_status`	string	Current audit status of the claim (e.g., 'pending', 'under_review', 'cleared', 'confirmed_fraud')pending · under_review · cleared · confirmed_fraud · optional	cleared
`reviewer_id`	string	Identifier for the auditor or reviewer assigned to the claimoptional	REV101
`patient_gender`	string	Gender of the patientmale · female · other · unknown · optional	female
`provider_specialty`	string	Medical specialty of the provider (e.g., 'cardiology', 'general_practice')optional	pediatrics
`service_location`	string	Location where the service was provided (e.g., 'inpatient', 'outpatient', 'office', 'telemedicine')4 locations · optional	office
`billing_code`	string	Internal or external billing code used for the claimoptional	BILL0001
`notes`	string	Additional notes or comments related to the claim or audit processoptional	Routine pediatric checkup…
Numbers 3 columns
`claim_amount`	float	Total amount claimed for reimbursement0 or more	112.4
`fraud_score`	float	Numerical score indicating the likelihood of fraud (0-1, where 1 is highest risk)0 to 1 · optional	0.92
`patient_age`	integer	Age of the patient at the time of service0 to 120 · optional	8
Dates and times 3 columns
`claim_date`	date	Date the claim was submitted	2024-02-15
`service_date`	date	Date the medical service was provided	2024-02-13
`review_date`	date	Date the claim was last reviewed or auditedoptional	2024-02-21
True or false 1 column
`fraud_flag`	boolean	Indicates whether the claim is flagged as potentially fraudulent	false

Use it for

An insurance dashboard
The fraud_flag rate, claim_amount by service_location and a breakdown of fraud_scenario. Excel, Power BI or Tableau.
Why do the 51 inpatient rows have a mean claim_amount of 9,761?
A root-cause class exercise
Hand out the rows and one question. The answer is in the data, not in the brief.
Claims200
CLM0001112.4office
CLM000425100inpatient
CLM000718950inpatient
A software demo
Believable claims with patient_id, provider_id and claim_date to fill a screen in front of a buyer.

This dataset200 rows20 columns

Yours1,000 rows20 columnsservice_location: UK only

blueprint · medical-claims-fraud-scenarios

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.

Rules it was built with

Include claim type, provider ID, patient age group
Flag claims as valid or fraudulent
Include reason for flagged claims
No real PHI

Rows1,000 rows10,000 rows100,000 rows1,000,000 rows

Open the blueprint in Data Factory

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-claims-fraud-scenarios

What should your data show?

No signup. No card.

Sample rows

What the 200 rows show

Columns

Use it for

An insurance dashboard

A root-cause class exercise

A software demo

Make it yours.

Same schema. As many rows as you need.

More in Insurance

What should your data show?