Medhealth Review

Performance evaluation design for in vitro diagnostic software

Technological advances in the healthcare industry have led to a paradigm shift in the sector. The rapid pace at which software, including artificial intelligence (AI), has affected the field has also led to a response from regulatory bodies. With the addition of the term “software” to the definition of in vitro diagnostic medical devices (Regulator (EU) 2017/746 of the European Parliament and of the Council of April 5, 2017; IVDR), software – both which is part of a device and which is itself a device – began to be included in the scope of the regulation of in vitro diagnostic medical devices and thus received due attention.

In vitro diagnostic (IVD) software is a specialized computer software used in medical or diagnostic laboratories to analyze patient samples primarily for:
Patients health assessment,
Diagnosing diseases,
Obtaining information on predisposition to a specific disease,
Predicting response or reaction to treatment,
Monitoring therapeutic actions.

Increasingly, IVD software is based on artificial intelligence, which is undoubtedly revolutionizing the development of medical devices and the entire healthcare system. The use of AI enables IVD medical devices in the form of software to dynamically adapt to new conditions and optimize performance in real time. In practice, this translates into an increase in the quality of medical services, while reducing costs. Benefits for the patient of AI-based IVD software are undeniable – improved diagnosis of diseases and health conditions, and even the ability to create early warning systems for life-threatening patients, to name just a few.

What are the types of IVD software?

When considering the term “in vitro diagnostic software”, consider the following forms of software:
Software as a stand-alone in vitro diagnostic medical device, according to IVDR. This type of software is used to provide diagnostic or therapeutic information and is not intended to support other IVD medical devices.
Software as an accessory to an in vitro diagnostic device that is supplied separately from the device, but is intended to operate or affect the device.
Embedded software that is part of an IVD medical device. It controls this device or influences its use. This type of software is not classified independently, but as a part of a medical device.

IVD software regulations and standards

In vitro diagnostic medical devices are subject to a number of standards and guidelines. The selection of the appropriate ones depends primarily on the chosen market and the classification of the device in question applicable to that market.

For software that is an IVD medical device, the following standards should be considered:
ISO 13485:2016 Medical devices – Quality management system – Requirements for regulatory purposes
ISO 14971:2019 Medical devices – Application of risk management to medical devices
IEC 62304:2006 Medical device software – software life cycle processes
IEC 62366-1:2015 Medical devices – Application of usability engineering to medical devices.

Risk classification according to IVDR

In vitro diagnostic software, like other IVD devices, is classified according to the risk to patients – from Class A (low-risk devices) to Class D (high-risk devices). However, the IVD software classification does not affect the safety and performance requirements that all IVD medical devices must meet. Any software, regardless of class, is developed and produced in accordance with the state of the art and taking into account the principles:
Software life cycle (IEC 62304:2006),
Risk management (ISO 14971:2019), including information security,
Usability (IEC 62366-1:2015),
Verification and validation.
However, the classification identifies possible conformity assessment procedures. For Class A IVD medical devices, no documentation review by a notified body is required. Manufacturers also do not need a certified quality management system.

How are the classification rules applied in practice?

The IVD regulation emphasizes that the decisive factor for the classification of IVD software, is the intended purpose of using the device. It is important whether the software allows the provision of clinical information. In other words, what matters is whether the software is used to determine the presence (diagnostic sensitivity) or absence (diagnostic specificity) of a marker associated with a particular disease or health condition.

How should a performance evaluation be conducted for in vitro diagnostic software?

Designing a performance evaluation for in vitro diagnostic software includes several key aspects, such as:
Create performance evaluation requirements: The requirements for evaluating the performance of IVD software should be defined. This includes, among other things, defining the objectives, scope, standards and guidelines that must be met. According to the IVDR, software, which is itself a device, should be designed to ensure reproducible results, reliability and performance consistent with their intended use.
Functionality testing: Tests are conducted to verify that the software meets the specified functions and requirements. Testing occurs for such elements as data entry, results processing, report generation and, where applicable, interactions with other systems.
Performance testing: In vitro diagnostic software is typically provided for big data analysis and real-time operation. Its performance, scalability and load response are being tested.
Security and confidentiality: The software must meet requirements for medical data security and compliance with regulations and standards related to patient privacy.
Validation and certification: The final step is the validation process, which involves proving that the software works as expected and meets certain standards. A positive validation result is required for CE IVD certification.
Manufacturers of medical devices for in vitro diagnostics, face the challenge of developing high-quality, productive devices that meet relevant regulatory requirements. The IVDR regulation requires manufacturers to ensure the safety of their products, and expectations for software documentation are also high.
Pure Clinical, as a full-service provider, will guide you through all stages of your software’s life cycle related to performance evaluation – competently and reliably. The experience and knowledge of our experts allows us to ensure high quality services, accuracy and reliability of the process. We will support you every step of the way, providing comprehensive solutions that meet relevant norms, standards and guidelines.


Regulation (EU) 2017/746 of the European Parliament and of the Council of 5 April 2017 on in vitro diagnostic medical devices and repealing Directive 98/79/EC and Commission Decision 2010/227/EU.
Medical Device Coordination Group (2019). MDCG 2019-11 – Guidance on Qualification and Classification of Software in Regulation (EU) 2017/745 – MDR and Regulation (EU) 2017/746 – IVDR Available at: .
ISO 13485:2016 Medical devices — Quality management systems — Requirements for regulatory purposes.
ISO 14971:2019 Medical devices — Application of risk management to medical devices.

IEC 62304:2006 Medical device software — Software life cycle processes.
IEC 62366-1:2015 Medical devices – Application of usability engineering to medical devices.

By Małgorzata Pawlikowska PhD, Clinical Research Scientist, Pure Clinical Sp. z o.o

A PhD graduate of Biological Sciences and MA studies in Biotechnology at Nicolaus
Copernicus University in Torun as well as post-graduate studies: Conducting and Monitoring
of Clinical Research at Kozminski University in Warsaw with over 14 years of experience in
scientific research industry.
She has a working experience in designing and scientifically coordinating performance studies conducted in accordance with the ICH-GCP rules and the requirements of IVDR 2017/746, FDA and other applicable regulatory bodies. Her responsibilities include providing oversight and analysis of guidelines related to clinical trials, developing study protocols and other study materials for research projects within the IVD Department, as well as analyzing data obtained during clinical and analytical performance studies.

Must Read
Related Articles