Sign up for the Healthcare NLP Summit 2024, April 2-3. Register now.
was successfully added to your cart.

Diagnosis Related Groups (DRGs) and Big Data

For a clinical decision maker, determining a definite diagnosis is not always an easy task.  A definite diagnosis needs you to have a meticulous review for the patient medical history, laboratory data, and some radiographs to see and interpret.

When it is hard to reach a definite diagnosis, the physician may use descriptive terms rather than a definite known diagnostic term.  Sometimes, they may add the term “idiopathic” to the patient symptoms.

The need for DRGs

There is a great need to obtain a definite diagnosis considering the medicolegal issues, third-party reimbursement, epidemiologic monitoring, and the prospective payment system.

Even, the emergence of Electronic Medical Records (EMRs) and the need for hermetic integration and interoperability between different systems imposed a need for a standardized diagnostic procedure especially if we target the automation of the billing process.

What is the Diagnosis Related Groups (DRGs)?

A Diagnosis Related Groups (DRG) is a system that classifies patients by a standardized prospective payment to hospitals.  It supports cost containment as it should cover all the charges paid starting from the time of the inpatient admission to discharge.

It categorized patients according to the following variables.:

– principal diagnosis

– secondary diagnosis

– surgical procedures performed

– comorbidities and complications

– patient age and sex

– discharge status

There are currently three commonly used DRG versions:

  • The basic DRGs which are used by the Health Care Financing Administration (HCFA). This system is used for the payment of inpatient Medicare beneficiaries.
  • The All Patient DRGs (AP-DRGs) is considered an expansion of the basic DRGs. It targets non-Medicare populations such as pediatric patients.
  • The All Patient Refined DRGs (APR-DRG).

Later subclasses like illness DRGs (S-DRG) and risk of mortality DRGs were added.

Outliers

Patient care cost is compared to a typical patient within the group. If it appears to be higher than the typical patient, the case is referred to as “Outlier”.

Grouper

It is a software developed to assign the DRG classification.

Hawaii Medical Service Association (HMSA’s) grouper can be taken as a natural language processing case study as it uses the DRG case designation categories as Medicare, as defined in the annual “Inpatient Prospective Payment System” final rule.

DRG Grouper is updated at October each year (with the update of ICD-9-CM/ICD-10-CM)

John Snow Labs catalog includes a valuable dataset which could provide insights about hospital-specific charges for more than 3,000 U.S. hospitals that receive Medicare Inpatient Prospective Payment System (IPPS) payments for the top 100 most frequently billed discharges, paid under Medicare based on a rate per discharge using the Medicare Severity Diagnosis Related Group (MS-DRG).

This can allow the reader to have a better understanding of what is meant by “Inpatient Prospective Payment System” and how could it be used to obtain accurate insights.

History of the DRGs development in the US

– In 1983, the DRG system was introduced as a mandatory methodology for hospitals inpatient reimbursement.

– In August 2006, Centers for Medicare and Medicaid Services (CMS) issued a plan to transfer the CMS-DRG with a classification methodology that will reflect the severity of the disease.

– Patient groupings were based on medical diagnosis (like ICD-10).

– DRG system focus more on procedures rather than diseases and targets more the financial use rather than the actual scientific classification of diseases (on the contrary to ICD).

– A successful DRG system must take into consideration:

– The physician’s Clinical concerns.

– Documentation needs for the 3rd-party reimbursement.

– Feasibility of recording on patient chart.

– For discharges beginning on or after January 1, 2017, the SIWs (Service Intensity Weights), cost thresholds and ALOS (Average Length of Stay) effective July 1, 2014, were used for payment purposes with the updated APR-DRG grouper version 33.

You can find the APR-DRG grouper version 33 dataset in John Snow Labs catalog.

DRG Worldwide

For any country to develop a DRG system, it can either develop their own system or import an existing one.

In their initial trials with the DRGs, England, Portugal, France, and Ireland imported DRG systems that were developed in the US.

Ireland, Spain, and Portugal adopted the Health Care Financing Administration (HCFA-DRGs) or All Patient (AP-DRGs).

Ireland, Slovenia, Croatia, and Romania adopted the Australian-Refined (AR-DRG) system.

Germany developed its own standard German DRG (G-DRG) based on the AR-DRG system.

Poland also developed its own system but this time it was based on the English system.

Most countries started using an existing DRG system as a starting point and then develop its own.  Choosing which system to adopt is even a big challenge where there many factors to consider, like the compliance of the adopted system national hospital system considering the applied coding systems for diagnoses and procedures, the presence of training materials and staff and the technical support for the system.

Costs of adopting an available system might be high as the country that needs to apply the system have to pay for the copyrights for this system. Costs will be higher if the system will apply a system that is based on a coding system that is not available or applied within the target country.  This will entail purchasing the copyrights to use this coding system as well.

For example, in 2003, Ireland adopted the Australian coding system when it decided to change its used DRG system from HCFA-DRGs to AR-DRGs.

Either adopting an existing system or developing a new one is a very complex process and entails a lot of challenges to consider.  However, all countries are still exerting more efforts and funds to either improve their DRG systems or take their first steps to apply it.

Medical Imaging and Big Data - Part II: Medical Imaging and Big Data Challenges

The emergence of new artificial intelligence concepts like face recognition, supervised-learning science, and predictive analytics have imposed new challenges to medical imaging....
preloader