Anomaly Detection in Health Insurance Claims Using Bayesian Quantile Regression

Table 2

Claims data with other covariate information.

Claim amount	Location	Diagnosis	Gender	Type of care	Pharmacy	Type of health facility
7257	1	1	2	1	1	1
5835	2	1	1	1	2	1
7023	2	1	1	1	3	2
9284	2	1	1	1	1	2
8148	2	1	1	1	3	2
9024	3	1	1	1	4	3
⋮	⋮	⋮	⋮	⋮	⋮	⋮
8256	3	1	1	1	3	1
8941	2	1	1	1	3	2
8385	3	1	1	1	4	2
11731.2	3	2	1	2	5	4

Location: 1, California; 2, Georgia; 3, Alabama; 4, Virginia. Diagnosis: 1, acute myocardial infarction; 2, heart failure. Gender: 1, female; 2, male. Type of care: 1, outpatient; 2, inpatient. Pharmacy: 1, heparin; 2, nitroglycerin; 3, eptifibatide; 4, clopidogrel; 5, furosemide; 6, metoprolol; 7, bisoprolol; 8, hydrochlorothiazide. Type of health facility: 1, ambulatory service; 2, outpatient clinic; 3, acute care; 4, hospital (source: https://data.cms.gov).