Introduction

Agile software development is a group of software development methods based on iterative and incremental development, where requirements and solutions evolve through collaboration between self-organizing, cross-functional teams. A time-boxed iterative approach, Agile software development methodology promotes adaptive planning, and encourages rapid and flexible response to change. It also provides a vehicle for evolutionary development and delivery of the product. Agile methodology is a conceptual framework that promotes foreseen tight interactions between functional groups throughout the development cycle.

The Agile approach to software development, configuration and project planning has recently become very popular for implementing and developing complex systems, and for good reasons:

  1. Builds confidence in the product
  2. Something in the hands of the users sooner, rather than later
  3. Constant contact with stakeholders allows for speedy design changes and requirement refinement

Purpose

To determine and analyze some of the obstacles and issues that are likely to be encountered doing a LIMS implementation using Agile methods.

Scope

The scope of this article is limited to the Scrum Agile method used in LIMS implementation of a CLIA (or otherwise regulated) laboratory. Not intended for GMP or IVD software solutions, but some information can be leveraged for these.

Analysis

The below are a list of issues to be considered and mitigated when using an agile method on a LIMS implementation.

Data Management and Continuity

Clinical laboratories collect an array of different data sets to conduct their business. Data sets include client and patient demographics, laboratory workflow data, inventory management transactions, QC data, intermediate and final patient results, and billing information. Data from the different data sets are often sourced from separate systems and interfaces, and must be integrated to ensure positive patient identification and result traceability.

Using an Agile method means that not all of these data sets will be implemented at the same time. There needs to be stakeholder buy-in for the planned data flow gaps in early releases. It may require some other methods for transferring data from the current system to the new one.

A traditional waterfall project management approach can require a data migration task as a part of the cutover activities, the scope of which had been previously determined and approved. After the migration, all of the data (agreed upon in the plan) will reside in the new system.

How do you address data continuity in an Agile method?

Regulatory requirements affecting user stories and priorities

There are regulatory requirements and quality system guidelines that the laboratory follows in conducting their business. Certain requirements and guidelines are specific to electronic/computer-based systems. These items must be identified as such in the user stories to ensure that the system requirements derived for regulatory compliance are appropriately represented and complete in early sprints.

Resource availability

One of the most effective features described in agile methods has to do with the rhythmic, continuous involvement with the business stakeholders. With increasing costs and shirking reimbursement rates, clinical laboratories are running increasingly leaner when it comes to staffing. Finding availability for the subject matter experts operating in production can be daunting.
Timing of sprint cycles and meetings should be approved by the business prior to finalization of the Project Plan. If possible, the business should provide proof that they have adequate resources to accommodate the proposed schedule.

Validation

Although CLIA does not require a validation of computer systems in the way that the FDA does, CAP and other voluntary accrediting agencies require documentation of system testing before initial implementation and during change events. Therefore, many laboratories may choose to validate its CLIA-regulated systems.

A waterfall project management approach can require validation and/or user acceptance testing of the computer-based system, either in whole or in part, prior to release to production. Upon approval of a given sprint implementation, a validation plan must be created, approved and executed prior to release to production. The project timings should account for validation efforts.

One suggested approach to validation a system developed using an Agile method is to write an “Agile” validation plan at the beginning of the project that calls for risk assessment at the planning phase of each sprint cycle. The results of the risk assessment will be used to determine which testing activities (unit testing, integration testing and regression testing) to perform for the upcoming sprint. As a part of the project plan, the validation plan can also describe specific milestones of “doneness” to perform full or partial system validations.

When planning the project, it may be useful to discuss the level of effort in validating the proposed requirements in a sprint to determine the proper scope of the sprint.

Conclusion

Traditionally-approached LIMS implementation projects are prone to pitfalls that can lead to projects being delayed, over budget and ultimately failed. Agile methodology attempts to diminish the likelihood of falling into these hazards by maintaining close engagement of users and project members and using iterative, adaptive project planning. However, Agile methods can introduce new challenges when implementing a system in a regulated, resource-lean environment such as a clinical laboratory.

The challenges described in this article may deter many project managers from attempting to use such an approach, especially the first time. As in life, identifying the challenges, planning for them, and facing them head on can lead to great success.

NO EVENT SHALL LAB INSIGHTS, LLC BE LIABLE, WHETHER IN CONTRACT, TORT, WARRANT, OR UNDER ANY STATUTE OR ON ANY OTHER BASIS FOR SPECIAL, INCIDENTAL, INDIRECT, PUNITIVE, MULTIPLE OR CONSEQUENTIAL DAMAGES IN CONNECTION OR ARISING FROM LAB INSIGHTS, LLC SERVICES OR USE OF THIS DOCUMENT.

Introduction

The completion of Human Genome Project marks a defining moment in the history of medical science. New technologies in molecular diagnostics and computational biology have furthered understanding of the transcriptome and proteome at a rate previously unimaginable. Since then, a flood of gene expression profile tests and other biomarker panels have been introduced to the clinical diagnostics market. These tests were all developed in a similar manner—by measuring and comparing the abundance of a group of specific gene transcripts, proteins, or other biomarkers in populations of patients or tissues with a known pathology or prognosis state. This comparative analysis results in the development of a predictive algorithm that can determine the probability of disease or prognosis in unknown samples.

On July 26 2007, the Food and Drug Administration (FDA) published a draft guidance to address this emerging field of In Vitro Diagnostics for Multivariate Index Assays (IVDMA). An IVDMIA is defined by the FDA as a device that:

  1. Combines the values of multiple variables using an interpretation function to yield a single, patient-specific result (e.g., a “classification,” “score,” “index,” etc.), that is intended for use in the diagnosis of disease or other conditions, or in the cure, mitigation, treatment or prevention of disease, and
  2. Provides a result whose derivation is non-transparent and cannot be independently derived or verified by the end user.As a result, several organizations filed their products for de novo 510(k) clearance, while most waited for a final guidance. The topic has since been shelved by the FDA as they now consider broadening the guidance to govern complex Laboratory Developed Tests (LDTs) currently governed under CLIA. Although the final guidance has been postponed, the final version will most likely still contain most of the items identified in the draft.

Purpose

The purpose of this document is to summarize the draft guidance as it pertains to Laboratory Developed Tests and discuss its implications at an independent clinical laboratory.

Scope

The scope of this document is limited to studying the draft guidance as it was written in 2007. It does not contain any commentary regarding any potential changes to the guidance as a response to industry comments or a changing regulatory climate.

Analysis

Approach

The FDA continues its efforts at administrating a “least burdensome approach” for IVDMIA. The guidance states that, although there is a potential for complicated and emerging technologies to be included in IVDMIA submissions, the classification of devices will be based on risks associated with intended use. (The guidance includes an example of a test that indicates the likely prognosis of a cancer will most likely be considered Class II, versus a test that indicates the therapy regimen would be considered Class III.) The FDA acknowledges the fact the nature of the types of disease states analyzed by these new tests will most likely require a classification of II or III.

The guidance states that safety and effectiveness determination should include “review of the performance of the entire system, including the accurate measurement of the input variables, directions for use, and expected analytical or clinical performance, rather than a review of only certain subcomponents of the test” (i.e. just the algorithm). This approach is consistent with classification determination of other devices, like clinical chemistry and clinical toxicology test systems.

Premarket Regulation

In the spirit of using a “least burdensome approach”, the FDA proposes a flexible approach to safety and effectiveness determination. Although a prospective study is preferred, the administration will also consider archived samples and/or retrospective studies, as long as the study design, sample composition, sample selection and sample storage processes reflect the intended use and intended population for use of the device.

In lieu of properly designed retrospective studies, IVDMIA manufacturers may also file for “Investigational Use Only” labeling as defined by 21 CFR 809.10 and apply for an investigational device exemption (IDE). Other labeling requirements are to be consistent with those regarding other devices.

Postmarket Regulation

Consistent with other devices, the FDA states that IVDMIAs are subject to Quality System regulation as described in 21 CFR Part 820, however, the FDA intends to continue to exercise enforcement discretion for CLIA-regulated laboratories until a final guidance is approved for laboratory developed tests.

The Administration states that all IVDIAs must comply with 21 CFR 803 – Medical Device

Reporting standards. Also, laboratories that run IVDMIA test systems are considered “user facilities”, and must submit serious injury/device malfunction reports as such (CFR 21 830.50).

Enforcement

In an effort to reduce the effect on innovation costs in the market, FDA intends to exercise enforcement discretion with respect to all regulatory requirements for currently marketed, laboratory-developed IVDMIAs for 12 months following publication of the final guidance document. In an effort to encourage early adoption to the new standards, FDA intends to exercise enforcement discretion for an additional 6 months for any currently marketed, laboratory-developed IVDMIAs if the manufacturer submits a 510(k) or PMA within the initial 12 month period following publication of the final guidance.

Conclusion

While the debate on whether Laboratory Developed Tests are medical services (and should be regulated by CMS under CLIA) or medical devices (and should be regulated by the FDA as IVDs) seems to be coming to a head, LDT developers and IVD manufactures brace for a dramatic shift in regulatory environment. FDA has used “enforcement discretion” in exerting what it sees as its legal authority to regulate medical devices, but many analysts believe that the time will soon come when enforcement will be applied to some areas currently left up to CLIA. In particular, multivariate index assays have come under particular attention in recent years; enough to spur publication of draft guidance by the FDA.

This emerging development may provide some interesting opportunites for device manufacturers. Since they many systems already in place to govern compliance to FDA device manufacturing requirements for new and existing products, these systems can and should be leveraged in order to bring their own CLIA laboratory into compliance for a potential 510(k) or PMA submission when/if the time comes. Other leaders in IVDMIA laboratory testing have already begun to retrofit their current processes and build their new products with the FDA in mind, but without the institutional regulatory architecture, they will be at a disadvantage. Life Technologies has the opportunity to create IVDMIA products with competitive compliance that surpasses the expectations of regulators and their customers while controlling costs.

References

  1. 2007. Food and Drug administration. Draft Guidance for Industry, Clinical Laboratories, and FDA Staff. In Vitro Diagnostic Multivariate Index Assays. http://www.fda.gov
  2. 2011. Smith, Katie M. Exploring FDA-Approved IVDMIAs. http://www.ivdtechnology.com/article/exploring-fda-approved-ivdmias
  3. 2012. Wiess, Ronald L. The Long and Winding Regulatory Road for Laboratory-Developed Tests. American Journal of Clinical Pathology, 138, 20-26.

NO EVENT SHALL LAB INSIGHTS, LLC BE LIABLE, WHETHER IN CONTRACT, TORT, WARRANT, OR UNDER ANY STATUTE OR ON ANY OTHER BASIS FOR SPECIAL, INCIDENTAL, INDIRECT, PUNITIVE, MULTIPLE OR CONSEQUENTIAL DAMAGES IN CONNECTION OR ARISING FROM LAB INSIGHTS, LLC SERVICES OR USE OF THIS DOCUMENT.