In one of our previous articles, we touched on the topic of service classification (coding).

The idea of unifying services, pharmaceuticals, and medical supplies is not new, but it remains relevant. In this article, we focus on these three types of health insurance items - services, medications, and medical devices - in the context of monetary value, and how it impacts the key stakeholders in health insurance: insurers, healthcare providers, and government regulators.

Coding is not just a reporting tool — it is a mechanism for:

tariff setting;
fraud prevention & detection;
loss forecasting;
and even legal protection in dispute resolution.

The essence of coding lies not in filling in for the shortcomings of classification systems — those have been around since the mid-20th century and are mature enough — but in the uneven pace of digitization across healthcare systems. For insurance companies, coding goes far beyond administrative reporting: it is the foundation of tariff calculations, risk management, loss mitigation, and fraud detection.

An incorrect, missing, or unreadable code turns a claim into a risk — whether due to a payment denial, legal dispute, or distortion of actuarial models.

In many parts of the world, healthcare digitization outpaces standardization. This is especially evident in regions such as the Middle East (MENA), South Africa, India, Latin America, and Southeast Asia. While digital transformation is underway, standards and infrastructure remain underdeveloped. Even when electronic claim submission exists, up to 30–40% of claims are submitted in unstructured, free-text formats. This complicates automation, and reduces transparency.

Due to the rapid development of pharmaceuticals, healthcare reforms, new service types (e.g., telemedicine), and constant changes to coding systems, strict adherence to standards is often unfeasible. As a result, we encounter problems like these:

errors in actuarial modeling,
claim rejections,
regulatory penalties,
software integration issues.

Coding Systems

Dozens of systems are in use, such as:

For medications

ATC, NDC (US), WHODrug, TUSS (Brazil), Russian drug register based on ATC

2 For medical devices

UDI, GTIN, GMDN, etc.

3 For services / procedures

ICHI (WHO), ACHI (Australia), OPCS (UK), Nomenclador Nacional (Argentina), Russian medical services nomenclature

1

Homegrown Systems Can Be Better

Unlike clinical guidelines, which tend to be international, coding systems may and should be national — and that’s an advantage:

legal frameworks exist to support them;
national catalogs can cover all services, drugs, and devices used locally;
detailed granularity can reflect dosage forms, registration status, and more.

Maintaining such systems locally is more practical and reliable.

For instance, in Saudi Arabia, there is a national medical solution - information exchange platform NPHIES. The platform is developed by CHI and NHIC, who have mandated since 2023 the use of Saudi Billing System (SBS) and GTIN/GMDN codes for drugs and devices. However, despite major progress, challenges in integration with the providers and insurers and a shortage of trained coding staff remain significant obstacles.

Reduced financial loss

Faster claim processing

Expert-driven feedback for system learning

Staff training via real cases

Better actuarial modeling

Semi-automated validation

Let’s add some fire

Where there's a need, solutions emerge and evolve. Since the 1970s, various standards have appeared to support medical data exchange — for lab results, medical imaging, insurance claims, and EHRs. Over time, many of these systems overcame weaknesses like:

limited flexibility,
narrow specialization,
high implementation and support costs.

Today, the leading global standard for medical data exchange is FHIR (Fast Healthcare Interoperability Resources), developed by HL7 (Health Level Seven International). FHIR is:

modular,
developer-friendly,
web-native and scalable.

It defines the format, structure, and transfer method of data. FHIR is supported by major companies (Apple, Google, Microsoft, Amazon, Epic, Cerner) and used in national platforms like US Core, NPHIES (Saudi Arabia), and X-eHealth (EU).

That said, FHIR itself does not contain specific codes for procedures or medications — it provides a technical framework for transmitting structured data, enabling interoperability. For example, Daman Health (UAE) uses standardized APIs and protocols to simplify data exchange between insurers and providers.

Crutches or Steroids?

Most integrations between insurers, providers, and governments still involve manual steps — checklists, human review, bureaucratic control — especially in systems lacking FHIR or equivalents. These are necessary "crutches".

AI is promising, but only when paired with human supervision. Selective expert review of AI results is essential to maintaining quality. Think of it as “steroids” for a weak system.

Solutions — including those developed by Mains Lab — enable a semi-automated workflow:

Input: free-text prescriptions, patient info, diagnoses.
Output: accurate codes from target nomenclatures (e.g., ICHI, ACHI, SBS).

This is achieved through:

machine learning,
NLP for context extraction,
expert review, including local medical coders — for high-quality output.

These systems don’t just match codes — they predict which services logically follow a diagnosis, helping prevent service fragmentation and abuse. That’s more than steroids — it’s a true solution.

They also provide a training base for staff, reinforcing expertise and creating a feedback loop that improves the system over time.

Author: Alexander Petrov

Conclusion

Using the right tools and methodologies helps close the gap between free-text practice and structured data requirements. For insurers, coding is no longer optional — it is the foundation of sustainability, reliability, and forecasting accuracy.

Thanks to digitalization and the availability of specialized vendors, even regional insurers — not just global giants — can now build modern coding workflows if they have the vision, commitment, and the right partners.

Insurer Benefits

BACK TO NEWS