Guide to Health Care

Administrative Claims Data

UTHealth Houston School of Public Health
Center for Health Care Data
Table of Contents
• Purpose
• Overview of Health Care Administrative Claims Data
• Elements in Claims Data for Analysis
• What Claims Data Does NOT Include
• Insurance Types
• Texas All-Payor Claims Database (TX-APCD)
• Planning Your Project Utilizing Claims Data
• Contact Us
Purpose
This document provides a high-level overview of administrative health care
claims data and serves as a general reference for qualified researchers
requesting access to claims datasets licensed and administered by the
UTHealth Houston School of Public Health Center for Health Care Data.

While an introduction to claims data is included, researchers who are new to

using administrative claims for research purposes are strongly encouraged to
complete a dedicated tutorial or formal course to gain the foundational
knowledge necessary for effective and responsible use of claims data.

3
Overview of Health Care
Administrative Claims Data
 Member
Enrollment Counts

Healthcare Episode of Care

Administrative
Claims Data Healthcare Costs

A collection of data that is used

for insurance billing purposes Place of Service
and captures information from all
providers and payors.
Prescription Data
Claims data can be used to
study healthcare delivery,
benefits, harms, and costs. Patient Demographic
& Clinical Information
 Tracking the Patient Journey with Claims Data

Plan enrollment
start date

Physician Visit
Lab Work

Hospital
Admission/ Rx Fill
Discharge

Rx Fill

Yellow areas indicate where electronic health Physician Visit

record (EHR) data can supplement claims data.
How Claims Data Adds Value

Learn more about

Study a large
patient
portion of the
procedures and
population
diagnoses

Use aggregated
Follow insured data on episodes
patients over time of care
(inpatient, ED visits,
office visits, etc.)
Using Claims Data to Explore Diverse Health Topics

• Adolescent Health • Mental Disorders • Substance use disorder

• Chronic Disease • Prescription Use • Treatment and
• Comorbidities • Preventive Care Therapies

• Dental Health • Maternal Health • Women’s Health

• Healthcare Service • Men’s Health • And more!
Utilization
• Morbidity and Mortality
• Immunizations
Claims Data as a Complementary Data Source

Primary Data
Collection

Electronic Health Non-Medical

Records (EHR) Drivers of
Health
Datasets
Patient Registry
 Disease Surveillance
Data can be used to monitor disease prevalence, incidence, and
health outcomes across regions in real time.

Common Uses of Predictive Analytics

Claims Data Predictive analytics, along with statistical modeling, can be used to
help identify potential risks, patterns of care, and more.

Program/Intervention Evaluation
Using data to examine the effectiveness of program delivery

Health Economics & Outcomes Research

An evaluation of services/interventions provided by healthcare
companies to analyze costs, utilization, and health outcomes.
Elements in Claims Data
for Analysis
 Claims Data Tables
Claims data tables contain the following information:
• Member Enrollment
• Patient and plan information
• Medical Claims
• Includes inpatient, outpatient, home health, skilled nursing, professional, and others
• Pharmacy Claims
• Filled prescriptions
• Dental Claims

Add a footer 12
 Medical Claims
• Medical claims are submitted by providers for an
encounter with a patient. This includes licensed
healthcare professional (physicians, Physician
Assistant, etc.) and facilities (hospitals, clinics,
etc.)
• Information includes:
• Patient Demographics, Diagnoses, and Procedures
recorded at the time of encounter
• Provider – Both physician and facility level. Billing and
rendering providers.
• Payment – Charged amount, paid amount, copay, etc.

Add a footer 13
 What is in a medical claim?
• Patient ID • Type of Bill
• Claim ID • Facility Type Code (link) and Service
Classification Type Code (link)
• Provider (billing, rendering, etc.)
• Diagnosis Related Group (DRG):
• Diagnosis codes • All Patient Refined Diagnosis Related
• ICD10-CM Group (APR DRG) (link)
• Medicare Severity DRG (MS-DRG) by
• Procedure codes Center for Medicare and Medicaid
• CPT/HCPCS or ICD10-PCS Services (CMS) (link)

• Modifiers and Units • Revenue codes

• Place of Service • Payment amounts
• Allowed, paid, charged, etc.
*It is up to the data requestor to identify which codes are relevant for
their research. We highly recommend performing literature reviews to
develop comprehensive lists of diagnosis, procedure, and other
Addrelevant
a footer codes. 14
 Pharmacy Claims
• Pharmacy claims are submitted by a pharmacy when a
member fills the prescription.
• Drugs administered by a physician are coded in
medical claims
• Drugs given by Provider as IV Injections are coded in
medical claims
• HCPCS J Codes: Charges only cover the cost of the
drug/product. Charges associated with the administration
are not included
• Not all medical members have pharmacy (Rx) benefits
• Medicaid – all enrollees get Rx benefits
• Medicare – Requires enrollment in a Part D pharmacy
plan

Add a footer 15
 What is in a pharmacy claim?
• Patient ID
• Allowed amount
• Drug information:
• Therapeutic Class, Generic vs. Brand Name
Drug, Dose, Strength
• National Drug Classification (NDC) Codes
• Dispensing information:
• Date filled, New or Refill, Days Supply,
Quantity Dispensed
• Pharmacy Information

Add a footer 16
 National Drug Classification (NDC) Codes
• There are three segments to each NDC code
• Labeler code – 4 or 5 digits – not necessarily the manufacturer – assigned by Food
& Drug Administration (FDA)
• Product code – 3 or 4 digits – assigned by labeler
• Packaging code -2 or 1 digit – assigned by labeler

NDC Labeler Product Package

00054003621 00054 Roxane Lab 0036 Clarithromycin 21 60 tablet bottle
250mg tab

0002515152031 00025 Pfizer US 1520 Celebrex 100 mg 31 100 capsule bottle

cap

Add a footer 17
 NDC Codes – Continued
• NDC are displayed on drug packaging in a 10-digit format: 4-4-2 or 5-3-2 or 5-4-1
• CMS: Uses NDC 11 (5-4-2) by adding a leading zero to comply with the format.

Add a footer 18
Dental Claims
Professional Claims

• Some dental procedures are

medical claims – dental surgery,
orthodontal, periodontal
• Dental claims use Current Dental
Terminology (CDT) codes
• Claim fields are different from
medical—for example, claim fields
document which tooth received
treatment

19
Common Payment Fields in Claims Data
• Charges – amount charged by the provider of services
• Excluded/Discount – not allowed or discounted via contractual
arrangements
• Allowed: maximum amount plan will pay for covered health care service
• Also called “eligible expense,” “payment allowance,” or “negotiated rate.”
• If provider charges more than allowed amount, patient may have to pay
difference
• Paid: Total payments paid to a provider for a service. Payment equals
amount eligible for payment under medical plan terms after applying
rules such as discounts, but before applying Coordination of Benefits
(COB), Copayments, and Deductibles
• Out-of-pocket costs: deductibles, coinsurance, copayments for covered
services plus all costs for services that aren't covered
• Coordination of Benefits (COB): a patient is covered under more than
one insurance plan
20
What Claims Data Does Not Include
What is NOT in a medical claim
• Detailed clinical information
• Lab values: Lab values are typically found in electronic health records -
not claims data.
• Patient vitals: Patient vitals such as height, weight, blood pressure, etc.

• Race/ethnicity (for commercial insurance providers)

• Commercial insurance providers typically do not record information related
to race and ethnicity.

• Nonmedical Drivers of Health

• Claims data generally do not include information related to patients’ socio-
economic status, education, transportation, food insecurity, etc. However,
there may be opportunities to leverage NMDOH databases alongside
claims data to understand nonmedical drivers of health for patients.

22
NOT in a medical claim – continued
• Information on family members
• Spouses and/or children of the insured may not be found in claims
data, as they may be covered under separate insurance plans
• Some plans are individual enrollment only
• Subjective information
• Medical claims do not include healthcare providers’ opinions,
feelings, or beliefs about the patient’s condition
• Premiums
• Insurance premiums—the regular payments members make to
keep coverage—are not included in medical claims.

23
What is NOT in a pharmacy claim
• Non-prescription items such as:
• Over the counter medications
• Food or dietary supplements
• Medication that was prescribed, but not filled by the patient
• Medication that was taken in a hospital (this will be in the medical claim)
• Lab results and diagnostic test results
• Subjective information (opinions, feelings, or beliefs about the patient’s
condition)
• Details on patients within the same household
• Prescription adherence
24
Insurance Types
Insurance Types
• Commercial
• These include both HMO and PPO plans. Provide
plans to individuals and groups, often through
employers. Common carriers that offer
commercial plans are Blue Cross Blue Shield,
Aetna, United Health Care, etc. These include
both HMO, POS and PPO plans.
• Medicare Advantage (Part C): a Medicare-
approved plan offered by private insurers and
offers a bundled Medicare plan (Part A, B and
often a prescription drug benefit). It may also
include extra benefits such as dental, vision, etc.
26
Insurance Types - Continued
• Government-sponsored plans
Medicare Fee For Service: This federal health insurance program is administered by the
Centers for Medicare & Medicaid Services and includes Parts A (hospital) and B (medical),
and Part D (when enrolled). This program is for:
• those ages 65 and older
• younger people with disabilities
• People with End Stage Renal Disease or Amyotrophic Lateral Sclerosis (ALS)
Medicaid: This federal health insurance program is administered by the state of Texas and
provides free or low-cost health coverage. In Texas, coverage is for:
• Low-income children and families (e.g. CHIP)
• Pregnant women with qualifying incomes
• People with disabilities
• Seniors ages 65 and older with limited income
• Texas programs include STAR, STAR kids, STAR health, and STAR + PLUS.
27
Texas All-Payor Claims
Database (TX-APCD)
Texas All-Payor Claims Database (TX-APCD)
The Texas All-Payor Claims Database is a large, statewide data system
that identifies cost and quality drivers from billions of medical, pharmacy
and dental claims from a wide range of payors — private insurance
companies, Medicaid, Medicare, and some self-funded employer health
plans.
Its main purposes are to:
• Increase transparency in health care cost and quality
• Support evidence-based policymaking by giving lawmakers and
agencies better insight into health spending, utilization, and trends
• Provide data access to qualified researchers to identify ways to
improve population health, reduce costs, and address variations in care
• Inform consumers through public tools that compare costs and quality
across providers
29
TX-APCD: What is NOT included
• Claims from persons residing outside of Texas
• Claims from persons insured under plans outside of
Texas
• Claims from persons who are Texas residents and:
• Are covered under Federal insurance plans; e.g.,
• Veterans Administration
• Tricare
• Medicare Fee for Service (however, the CHCD provides this data
from another resource to be used in conjunction with the TX-
APCD)
• Insured under an employer ERISA plan that opted to no
participate
• Uninsured individuals or individuals who choose to pay cash at
the time of service 30
Who are TX-APCD Submitters?
The Texas Department of Insurance (TDI) requires all payors
that are regulated by TDI to submit claims to TX-APCD.

• Medical plans • Medicaid Managed Care Monthly Submission Claims

Organizations (MCO’s)
• Dental plans • Enrollment Claims

• Medicaid fee for service • Provider Claims

• Behavioral Health plans • Medical Claims
• Non-ERISA self-funded
• Medicare Advantage plans • Dental Claims
plans & Non-Preempted • Pharmacy Claims
• Medicare Supplemental ERISA )100% of Medicare
plans (voluntary) Fee for Service claims are
• County and Municipal available through CHCD
Sponsored Plans
• State Plans

31
Planning Your Project Utilizing
Claims Data
Using claims to answer your question
• Who is in my population and how can I identify them?
• Inclusion and exclusion criteria
• What is considered an event or case?
• How much time should be covered?
• You only need data over many years if you are: (1) studying a very rare condition/event
and are concerned about sample size, or (2) specifically measuring trends over time.
• Restricting to less time can make overall data management easier. Claims data can get
very large, making programming and file storage efforts cumbersome and costly
• Am I interested in costs, and if so, which costs? Cost charged to insurance?
Cost paid by patients? Amounts received by providers?

33
Prepare ahead of time
• If you have not frequently used claims for research, we recommend a
literature review of your topic searching specifically for ways to identify your
target population, event, case, or others within administrative claims.
• It can also be beneficial to collect information from clinicians or billing
departments, if you work in a medical setting, for advice on which codes are
commonly used.
• While the CHCD is well-versed in claims data, we are not medical experts
and cannot tell you which diagnosis codes or procedure codes should be
used for your study. You are the Subject Matter Expert.
What to Prepare
• Diagnosis codes
• ICD-10-CM, APR DRG, MS DRG
• Procedure codes
• CPT/HCPCS, ICD-10-PCS
• Modifiers
• Type of bill, place of service, revenue codes
• Drug codes
• NDC, therapeutic classes, generic names
• Provider codes
• NPI, Specialty, NPI Taxonomy
35
Example 1 - Diabetes Prevalence
• Research question: What is the prevalence of diabetes in Texas?
• Who is in my population?
• Everyone? Adults 18+? Pregnant women? Children and adolescents? 65+?
• Should anyone be excluded? Those with diseases that cause secondary diabetes? Cystic
fibrosis, pancreatitis, corticosteroid use?
• Who counts as diabetic?
• Type1, 2, or gestational?
• Is one instance of a diabetes diagnosis sufficient? Two diagnoses at least 30 days apart for
increased positive predicted value? Should they also have a prescription for insulin?
• When?
• If this is a 1-year cross-sectional, should you only consider those with a full year of insurance
coverage? Is less time of surveillance sufficient?
• If someone is diagnosed in the year before your research year and is still alive the following
year, should we still consider them as diabetic?
Example 2: Mammography with
questions answered
• Research question: What was the screening mammography rate for Harris
County in 2024?
• Who is in my population?
• Women, age 40-75, living in Harris County at the end of the measurement period,
continuously enrolled in insurance for the entire measurement period, exclude those who
were in long-term care, were frail or had an advanced illness, or had a previous bilateral
mastectomy or two unilateral mastectomies (right and left)
• What counts as a mammogram?
• At least 1 mammogram identified using HCPCS/CPT procedure codes during the
measurement period.
• When?
• Current recommendations are for every 2 years, so for the rate in 2024, we must include
mammograms from 2023 & 2024.

37
Example 2 Mammography (cont.):
Prepared Codes
• 538 ICD-10-CM Diagnosis codes identifying
• Absence of left or right breast, advanced illnesses, frailty, history of mastectomy, hospice
encounter
• 121 HCPCS/CPT/ICD-10-PCS Procedure codes
• Mastectomy, frailty device, frailty encounter, hospice encounter, palliative care
• 14 Revenue codes for hospice
• 6 Mammography CPT codes

• This is just to identify the numerator and denominator!!!

38
For any questions regarding health care data,
or to submit a data request, please contact
our team for more information.

Thank you!

39