Adoption of chargemaster data in health economics and outcomes research (HEOR) is typically less common than more prevalent real-world datasets such as claims or EHR data, primarily due to chargemaster being a hospital specific dataset. However, dismissing it as a data source could result in researchers overlooking invaluable insights into treatment outcomes, procedure side effects and healthcare cost dynamics that are missing from more traditional data sources.
We spoke to Ginny Gleason, FinThrive’s Principal Clinical Expert of Data Insights, about hospital chargemaster data, its value for health economics and outcomes research (HEOR) and how FinThrive has addressed some of the challenges of using this unique data source.
Responses have been edited for clarity and length.
Gleason: Chargemaster data essentially refers to patient encounter data captured from the hospital’s charge description master database. This is the system that houses a comprehensive list of all hospital charges for services, procedures, supplies, devices and medications provided to patients. It is ultimately where the bill starts and includes a price for each item or service based upon the hospital’s associated costs related to overhead, expenses, equipment maintenance, staffing and more.
While the chargemaster data is the starting point for the claim and it’s a critical part of the revenue cycle, billing and coding rules limit the information included on the claims sent to the payer. Consequently, chargemaster data gives users greater insights into what happened to a patient during a facility visit compared to standalone claims data.
Gleason: Chargemaster data is key to showing value. With claims, most oral drugs are classified into HCPCS code categories such as J3490 (Unclassified drugs) and J8499 (Prescription drug, oral, non-chemotherapeutic). As such, the value of most oral drugs can’t easily be determined. However, in the chargemaster, drugs and supplies are each listed with a charge. From this data, the monetary value of a drug or supply can be determined, in addition to any clinical reactions or medications given to treat a reaction. This makes chargemaster data ideal for cost effectiveness or comparative effectiveness analyses.
The assignment of HCPCS codes for emerging technology is also a limiting factor of claims data. Upon FDA approval, new drugs do not have a permanent HCPCS code and it often takes between nine to 18 months to be assigned. Until then, the claim shows only that “unclassified drugs or biologicals” were administered whereas the chargemaster data shows the specific drug from the time the pharmacy begins dispensing the drug. This allows for real-world evidence to be generated in a timelier manner.
Gleason: Chargemaster data is only applicable to hospitals, so is most suited to analysis of drugs or medical devices that are used in an institutional setting. While it includes services provided by outpatient hospital departments and outpatient services, the chargemaster does not include information regarding treatment from providers other than hospitals and hospital-based physicians. Chargemaster data is least suitable for conditions or diseases exclusively treated in an ambulatory setting or for freestanding ambulatory treatments such as outpatient physical therapy; and it will not be useful for research on drugs primarily dispensed from retail pharmacies.
Also, HEOR projects that investigate emerging drug therapies and medical device usage will benefit from chargemaster data. This includes identifying the frequency of use for a specific drug or medical device in the hospital setting as well as the use of alternatives. Chargemaster data is particularly valuable for identifying whether adverse reactions to treatments occurred and the charges/costs associated with combating any such reaction.
Gleason: The story of a patient’s hospitalization can be better determined from chargemaster data due to the specific identification of every drug and supply utilized during the facility visit — all of which is not possible with standalone claims data. For example, if a patient has a longer hospitalization than another patient who had the same procedure and/or diagnosis, the chargemaster data can identify what treatments, supplies, medical devices and drugs were given on each day and from that describe the clear picture of “why” the length of stay was longer. The key value is its ability to provide insight into unexpected outcomes and pain control. These are areas that the claim would not specify which drugs the patient was receiving, but rather only that the patient was receiving a higher number of medications. However, these medications could be the patient’s home medications and not related to the hospitalization.
Detailed information regarding the patient’s hospitalization is, of course, housed in the EHR. However, abstracting information from the EHR can be cumbersome and require knowledge of the specific system. If information is contained in free text notes, it cannot be obtained from the EHR in a report format. Each record would have to be manually reviewed to abstract the desired information. And EHR data also lacks the charge information for each of the drugs or devices used, which limits the ability to measure the cost effectiveness of different treatments.
Gleason: During the COVID-19 pandemic, evolving treatment regimens and patients’ responses to new treatments were key to combating the virus. From chargemaster data, all stakeholders—from government agencies, providers and pharmaceutical manufacturers—could identify the drugs and treatments provided along with patient response and outcome effectiveness. This was key in a rapidly evolving healthcare environment where the patients hospitalized were the sickest and most at-risk. Chargemaster data provided more timely access to information without having to extract data from the medical record. This allowed for fast identification of effective treatments as well as those treatments that contained a greater risk of harm or limited benefit. These treatments and responses could also be compared to patient groups. For example, these patients could have comorbidities, such as asthma, heart failure or obesity.
Gleason: The biggest challenges that FinThrive has overcome with the data are (1) data quality for drug and medical device capture and (2) the ability to identify and track patients over time.
For capturing drugs and medical devices, the big challenge is that this information is manually entered in the charge description field by individuals who may or may not have any clinical knowledge. Unlike billing and coding requirements, there is no industry-required standardization of this field. The data entered includes brand name, manufacturer name, generic name, NDC code, HCPCS code and dosage. Given that this is a manually-entered field, misspellings and abbreviations that may be unique to each facility and/or department within the facility are often found in the field. FinThrive has developed an elaborate drug mapping process that takes the free text form data and maps to the National Library of Medicine (NLM). Through this mapping, FinThrive incorporates drug nomenclatures from the NLM that include the RXCUI and ATC codes, creating unparalleled consistency across the data. The medical device mapping has proved more challenging, but we have stepped up our efforts in this area over the last six months and dramatically improved the quality of our medical device data.
The other challenge is related to tracking patients once they leave the hospital. For tracking patients over time, FinThrive has developed a FinThrive Patient Key that standardizes a combination of PHI and internal patient keys for a unique identifier for each patient in their data. FinThrive also works with every major tokenization vendor to ensure that the chargemaster data can be easily linked to other sources such as EHR data, lab data, longitudinal claims data, genetic data and SDOH data sets.
Contact us to learn more about how FinThrive’s chargemaster data can help with your upcoming HEOR projects.