Abstract
Biorepositories in Africa need significant infrastructural support to meet International Society for Biological and Environmental Repositories (ISBER) Best Practices to support population-based genomics research. ISBER recommends a biorepository information management system which can manage workflows from biospecimen receipt to distribution. The H3Africa Initiative set out to develop regional African biorepositories where Uganda, Nigeria, and South Africa were successfully awarded grants to develop the state-of-the-art biorepositories. The biorepositories carried out an elaborate process to evaluate and choose a laboratory information management system (LIMS) with the aim of integrating the three geographically distinct sites. In this article, we review the processes, African experience, lessons learned, and make recommendations for choosing a biorepository LIMS in the African context.
Introduction
A
Materials and Methods
In 2012 and 2013, the National Institutes of Health-funded H3Africa Program established three regional biorepositories in Africa in Uganda, South Africa, and Nigeria. The biorepositories were developed to archive and facilitate international scientific access to well-curated biospecimens collected by research projects in the H3Africa Consortium Program. To facilitate international access to biospecimens, it was desirable that the biorepositories set up an interoperable LIMS for data integration and exchange. At the time of funding, each of the three biorepositories had a LIMS for their current capacity. The biorepositories adopted a User LIMS Requirement Checklist provided by Autoscribe Informatics (Berkshire, UK) for the LIMS evaluation. The checklist consists of 15 key user requirements, including, but not limited to, general system information, system configuration and customization, user access and security, functionality, static data tables, reporting system, and system support. Six commercial LIMS companies were invited to respond to the questions and requirements in the evaluation tool. Two open-source LIMS were self-evaluated by the H3Africa biorepositories in Uganda and South Africa. The returned results from three commercial LIMS vendors were combined for analysis and discussed during conference calls and online correspondence. Two biorepository consultants provided advice during the evaluation process. The technical description of how the commercial and open-source LIMS compared with each other is beyond the scope of this article. However, we describe the experience and lessons drawn from the LIMS evaluation process from an African/LIMC perspective.
Results and Discussion
LIMS selection at the H3Africa biorepositories
Following the evaluation process, the H3Africa biorepositories made decisions on their biorepository LIMS. The Institute of Human Virology Nigerian–H3Africa Biorepository (I-HAB) in Nigeria decided to upgrade the current LIMS with the latest software updates from the vendor. The Integrated Biorepository of H3Africa Uganda (IBRH3AU) in Uganda decided to acquire a new LIMS and migrate the data from the old platform to the new LIMS. The Clinical Laboratory Services–H3Africa Biorepository (CLS) in South Africa decided to acquire a new LIMS platform in addition to the existing LIMS. Both CLS and IBRH3AU chose the same new LIMS platforms. I-HAB was convinced that an upgrade was necessary and sufficient to achieve the interoperable requirements of the three program biorepositories.
Checklist for choosing a LIMS
The H3Africa project specifications required that the three biorepositories workflows would be harmonized through interoperable LIMS to enable data integration and exchange. Following the evaluation process, the following key elements were considered while choosing a LIMS for the project: (1) customizability and usability; (2) interoperability with other LIMS; (3) access to revisions, updates, patches, and maintenance releases; (4) cost and access to technical support services; (5) maintenance and associated costs; (6) multiuser/site support; (7) robustness to handle large volumes of sample information; (8) security systems (audit trail, user roles, and privileges, etc.); and (9) type of (open source or commercial) LIMS. Table 1 represents a summarized checklist of the elements we considered to be key while choosing a biorepository LIMS. A detailed checklist can be accessed via biorepository.h3africa.org website.
A detailed checklist can be accessed via biorepository.h3africa.org.
CFR, Code for Federal Regulations; LIMS, laboratory information management system.
Commercial versus open-source LIMS
Commercial LIMS are systems whose source codes are developed for sale and need authorization from vendors for licensed use. Open-source LIMS are systems whose source codes are available for distribution at no cost. 7 During the H3Africa biorepository LIMS evaluation phase, it was decided to implement a commercial LIMS in support of the H3Africa biospecimen collections. Commercial LIMS are much more expensive upfront than open-source LIMS, are less flexible for end-user adaptations, but do not need local expertise for support. Despite some features of modified open-source LIMS that might seem more applicable to the African setting, there were significant concerns about the stability of such systems and the lack of standardization. In addition, it was clear that the adaptation and maintenance of such a system would need highly specialized staff at three biorepositories and that this approach could create differences among the biorepositories, which could cause potential delays in interbiorepository transfer of data and material.
Financing of LIMS in the LMIC biorepositories
The functioning of biorepositories requires stability and continuity of the LIMS. In addition to the buying of commercial licenses, some unforeseen expenses were encountered in the implementation of the commercial LIMS in H3Africa included training costs for staff and more complex hardware requirements, unlike for open-source LIMS. In H3Africa, it was possible to budget for these contingencies, but these costs may become prohibitive for smaller biorepositories outside of a funded mechanism. In such cases, potential creative solutions include forming consortia with similar facilities to buy a multiuser license. This approach has the added advantage of enabling a shared forum for dealing with other problems. However, since each biorepository has separate views, specific role-based security would be implemented for each user's collections. It may also only be necessary to acquire specific modules within a commercial package to keep the costs lower.
Retaining and training staff to use LIMS have presented some challenges for H3Africa. Some strategies were used, including actively searching for personnel with experience, ensuring that the LIMS manager feels integrated into the biorepository management structure, and training junior staff to make sure that there is an adequate succession plan. In addition, including training clauses within purchasing agreements have mitigated some of the risk of purchasing complex licenses, but ongoing training represents an essential need and should be part of a biorepository's quality management system. Discussion of formulating in-house training material for LIMS support is ongoing across all three H3Africa Biorepositories for them to stay aligned with training and SOPs.
IT infrastructure
In addition to high acquisition costs, commercial LIMS requires an IT infrastructure to fully support their function. IT infrastructure is a combined set of the hardware (e.g., servers and computers), software (e.g., operating software), and network systems required to deploy and support the LIMS. As biospecimen numbers grow, there is a corresponding decrease in LIMS functionality such as very slow loading and processing speeds, which is linked to the supporting IT infrastructure. 8 Therefore, a LIMS infrastructure should have the ability to be scaled to meet the needs of the community it serves. Institutions that are unable to support a biorepository LIMS on their own should team with other institutions to develop and deploy a shared biorepository LIMS infrastructure. By using a shared infrastructure, each participating institution could support components of the infrastructure independently, while also collectively managing the entire biorepository architecture. This will not only give an economical benefit but also offers an environment for harmonizing complex, but still critical, components of a LIMS, such as structured data files and data models, as well as standards for data transmission.
User support services
Commercial LIMS often needs user support services such as customization, implementation assistance, annual licenses, maintenance, and updates, although the need for these services may vary over time especially as users become experienced with the system. User support can be provided in several forms, including telephonic support and on-site support. There is an additional cost to get access to such services, which may be minimized by conducting thorough initial training and license negotiation. Many commercial LIMS vendors have no support networks or offices in Africa. This increases the costs of user support services because of airfare, accommodation, and other attendant costs to access the services. In such circumstances, African biorepositories should try to use other remote-access technology to get access to support, such as public IP addresses that enable external access and manipulation of the LIMS. Unlike commercial LIMS, open-source LIMS do not have user support services and the user must troubleshoot locally, which can be challenging or impossible based on technical capacity.
LIMS harmonization in the H3Africa biorepositories
Before the H3Africa program, each biorepository had its own LIMS that met their current needs. Since the commencement of the program, the biorepositories have either acquired new LIMS or upgraded their existing systems following a thorough LIMS assessment program. The biorepositories conducted a harmonization exercise to ensure interoperability of the LIMS across the three sites. Pilot studies conducted between H3Africa biorepositories integrated data sharing and importation protocols through a pilot biospecimen and data exchange. Data exchange harmonization is essential if biorepositories are to work efficiently in networks such as H3Africa to support population genomics studies. Biorepositories need to define an agreed sharable set of data and data format for harmonization and interoperability to ease exchange. Pilot projects with virtual data transfer protocols were undertaken successfully, suggesting that this harmonization has been effective.
LIMS sustainability
The biorepositories should develop strategies for long-term LIMS sustainability. Reliable and adequate sources of funding are keys to sustainability of LIMS. Commercial LIMS vendors require the payment of annual support fees unlike open-source LIMS. The risk of liquidation needs to be considered as it impacts LIMS support and maintenance and the long-term viability of the biorepository. Biorepositories should implement cost recovery measures for users to ensure sustainability.
Conclusion
Developing a state-of-the-art biorepository requires much capacity and staff development, including acquisition of formal training, equipment, and software. Key among the biorepository infrastructure needs is a LIMS. Choosing a LIMS in low- and middle-income countries requires careful consideration of the various factors that could affect its successful and sustainable deployment and use. H3Africa biorepositories operating in a consortium have highlighted key factors that affect recommendations for successful LIMS choice and implementation.
Footnotes
Acknowledgments
The H3Africa Consortium Biorepositories are funded by the U.S. National Institutes of Health grants UH3HG007008, UH3HG007051, UH3HG007438, and UH2HG007092 of the H3Africa Common Fund Initiative. The authors thank the Autoscribe Informatics (Berkshire, UK) that provided the Laboratory Information Management System (LIMS) User Checklist used in the LIMS evaluation process. The following Investigators were members of the H3Africa Biorepository PI Working Group: Jennifer Troyer and Louise Wideroff, National Institutes of Health, Washington, District of Columbia. Sue Penno, Biorepository Consultant, National Institutes of Health, Washington, District of Columbia. Christine M. Beiswanger, Coriell Institute for Medical Research, Institute of Human Virology-Nigeria, Abuja, Nigeria.
Author Disclosure Statement
No conflicting financial interests exist.