Analyze this! Administrative claims data or EHR data in health services research?

One of the ongoing debates in health services research concerns the relative merits of using administrative claims data versus electronic health record (EHR) data for research. Should one be preferred over the other? Some question the degree to which administrative claims data continue to be valuable for health services research given the growth of EHR systems.

Administrative Claims Dataare generated primarily for the administration of payment for health services delivered by healthcare providers and facilities. Most administrative claims are based on the format of the CMS 1500 form for outpatient and provider services and the Universal Billing (UB-04) form for inpatient services. These forms collect patient information such as patient demographics (name, address, birthdate, gender and marital status), employment and insurance status, occupational limitations, dates of service, diagnoses and procedures, service provider information and charges for services. Due to the nature of the adjudication process for administrative claims, there is generally a 90-day claims “run-out” period during which payer “allowed” amounts are finalized before claims are added to a research database.

Electronic Health Record Datareflect a digital version of a patient’s paper medical chart. EHR’s are real-time, patient-centered records that can make information available instantly and securely to authorized users. One of the most valuable assets of an EHR is the results of laboratory, imaging and other diagnostic tests and records of health behaviors that are not included in claims data. In addition to containing medical and treatment histories of patients, an EHR system can contain more than the standard clinical data collected in a provider’s office. CMS Meaningful Use guidelines for EHR development expand upon clinical measurement and test elements to include patient demographics, medication use, potential medication interactions, clinical quality measures and the protection of electronic health information.

 The Use of Claims and Clinical Data for Health Services ResearchAs discussed above, administrative claims are produced primarily for the purpose of billing and paying for health services, while EHRs are produced primarily for recording and managing patient care. While some studies have found claims data to be less sensitive in identifying patients when compared to EHR records,1 others have found that combining claims and clinical data can improve quality measurement.

A recent study of pediatric patients concluded “[pediatric] care quality measures may not be accurate when assessed using only administrative claims. Adding EHR data to administrative claims data may yield more complete measurement.”2

A study sponsored by the Agency for Healthcare Research and Quality (AHRQ) evaluated the process and analytic yield of linking administrative claims data and electronic medical records using state Medicaid population claims and an academic medical center’s EHRs.3 The study group concluded that, although many challenges exist in combining and analyzing claims and clinical data, the combination of these two sources of healthcare data creates an analytic resource stronger than either source individually, and a process worth continued evaluation and improvement.

Our research group conducted a pilot study of the utility of combining administrative claims and clinical data to investigate differences in health status, clinical test results and charges between diabetics and non-diabetics in a population of commercially insured adults. 4 As part of our study we demonstrated that we can conduct useful analyses that provide insight into patient care and health status that is not available using either claims data or clinical data alone. For example, we were able to examine HbA1c, LDL, BMI, and blood pressure values in persons with and without a diagnosis of diabetes present on claims. This analysis produced an estimate of the proportion of diabetics with “controlled” disease and yielded a small number of patients who had out-of-range HbA1c values with no diagnosis of diabetes in claims.

Claims have the advantage of collecting data from various sites of services that may not be included in a single EHR and, thus, allow better risk classification of a patient and analysis of their overall utilization of health services. EHRs, on the other hand, contain valuable clinical detail not found in claims records and generally include data spanning multiple payers.

Future State — Clearly both administrative claims and electronic health records have their strengths and weaknesses. Future efforts to develop integrated databases will provide the most well-rounded picture of the health status, service utilization and financial profile of both individuals and populations. Linking an individual patient record with population data from similar patients could provide, for example, useful clinical benchmarks, outcome probabilities for potential courses of action and medication adherence statistics for medications being considered by the physician. Adding metrics from additional sources, such as census or survey data, could add a layer of social determinants of health (SDH) which could estimate the probability a patient will successfully comply with a treatment protocol. The future is not “or” but “and” when it comes to data for health services research.

Paul LaBrec is Research Director for Populations and Payment Solutions with 3M Health Information Systems.

1 Tang, PC et al. “Comparison of Methodologies for Calculating Quality Measures Based on Administrative Data versus Clinical Data from an Electronic Health Record System: Implications for Performance Measures.” J Am Med Inform Assoc. 2007;14:10 –15. DOI 10.1197/jamia.M2198. Available at:

2 Angier H et al. “Variation in Outcomes of Quality Measurement by Data Source.” Pediatrics 2014;133:e1676–e1682. Available at:

3 West SL, et al. “Use of electronic medical records and administrative claims data for assessing type 2 diabetes care.” Effective Health Care Research Report No. 18. (Prepared by University of North Carolina at Chapel Hill DEcIDE Center Under Contract No. 290-05-0040-1). Rockville, MD: Agency for Healthcare Research and Quality. January 2010. Available at:

4 LaBrec P. “Linking Healthcare Claims and Electronic Health Records (EHR) for Patient Management – Diabetes Case Study.” Paper presented at the 2014 Pharmaceutical Industry SAS User’s Group (PharmaSUG) conference (paper HA04), San Diego CA, June 2014. Available at:


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s