User:Shawndouglas/sandbox/sublevel11

Sandbox begins below

Computers in the laboratory are not a recent phenomenon. The mid-1960s saw laboratory computerization become increasingly popular, though that enthusiasm was often based on the potential of the computers themselves rather than their actual capabilities. Laboratorians imagined potentials such as automatic sample label generation, daily log and report management, instrument interfacing and data processing, results comparisons, and time management tools. However, it would take time for some of those potentials to be realized.

In 1970, Temple University's Marion Ball, M.A., an assistant professor in the Department of Medical Physics, conducted a survey of pathology directors in clinical laboratories that were using computers. Asking their opinions about the advantages and disadvantages of computerized systems in the lab, she received responses from directors in 15 U.S. states, as well as from three other countries. Responses included:

The ability to rapidly prepare cumulative records and then to inspect them for possible errors through analysis trends has been proven to be of tremendous advantage in a number of laboratories. We can prevent errors in our analytical systems, but we are not prepared to prevent errors in the collecting of the sample, the mislabeling of the sample, or the accidental use of an incorrect sample. Thus, the ability to inspect data trends presents the only real tool that we currently have to pick out these kinds of errors. - Max E. Chilcote, Ph.D, Meyer Memorial Hospital Division

There is little argument about whether an operating computer system can be an advantage in a laboratory, but the most critical time is the installation and transition from a "manual" to a "computer" oriented laboratory. - Robert L. Habig, Duke University Medical Center

Reading about these potentials and opinions today, some 50 years later, we see both clear similarities and definite advances. For example, Habig's statement about transitioning from manual to more automated processes still rings true today: it can be nerve wracking and critical to get the transition right. Conversely, while the systems of decades past weren't able to "prevent errors in the collecting of the sample, the mislabeling of the sample, or the accidental use of an incorrect sample," modern laboratory informatics systems provide many assurances to sample management in the lab. In many cases, activities such as label generation, reporting, results analysis, workflow control, test ordering, and broad interoperability are commonplace in modern systems. And those systems continue to advance, with machine learning now finding its way into a few laboratory data management and analysis workflows.

We've come a long way since the 1960s, to a point where the question is no longer "can a computerized system help my manufacturing-based lab?" but rather "how do I choose and implement an informatics system to help my manufacturing-based lab?" As the previous chapter highlighted, a wide variety of standards and regulations further drive the narrative of "not if but when" for informatics systems acquisition. What follows is information to help you with this necessary acquisition, while considering the technology, features, security, cost, implementation, and vendor guarantees that come with such a system.

3.1 Evaluation and selection

What exactly is a laboratory information management system (LIMS) anyway? Do I need one? What options are available and how do I compare them? What about a request for information (RFI), request for proposal (RFP), or request for quotation (RFQ)? These are questions laboratory professionals typically ponder upon finding themselves charged with the mission of finding software for their food and beverage lab. For many the task can be a daunting proposition.

You may know the workflow-related needs of your laboratory, but perhaps you don't know much about data management solutions like LIMS, leaving you intimidated by all the options. You'll first need to gauge your lab's informatics needs in order to determine which products are worth investigating further. Of course your lab's analysis requirements, reporting and data sharing constraints, instrument interfacing needs, barcoding and tracking requirements, quality assurance processes, etc. are very important factors. But these systems vary in numerous ways, and other important factors exist. Price should certainly be considered, although value is ultimately more important than a low price. Other important questions that get asked include:

• Should we purchase software licenses or "rent" the software via a subscription-based model?
• Does the software need to be on-site, or is a SaaS hosted option more practical?
• Is a modular or complete system better for us?
• What is the best licensing/rental scheme for us? Should we consider site, named user, concurrent user, or workstation licenses?
• Is the company qualified and trustworthy?
• What functionality is available to help our lab not only accomplish workflow tasks but also remain regulatory compliant?

These and other questions are addressed in this chapter.

3.1.1 Technology considerations

Your laboratory's workflow, instruments, data management requirements, budget, technological expertise, business goals, and risk tolerances will all play a role in deciding what technology to invest in. The textile testing lab, for example, may depend less on instrument integration than the stability, cycle, and challenge testing lab, with its microbiological workflows. As such, look at your laboratory's short- and long-term goals, budget, workflow, and regulatory requirements to gain a better understanding of what technology will be involved.

First, what are the laboratory's goals? Does the third-party laboratory owner envision a small investment, taking in a slow but steady flow of basic analytical requests, or do they envision expansive growth, expanding into multiple manufacturing-based testing domains? If the lab is starting small but is confidently expecting to grow, technological investments early on may want to take into account future technologies that may shape data management and security processes. Second, what kind of work will the lab be doing, and what regulatory responsibilities will guide hardware and software investment at the lab? If your lab will be conducting extractable and leachable testing, you'll be considering chromatography and spectroscopy instruments, as well as requirements for retaining analytical results for regulators. The accredited pharmaceutical manufacturing lab will likely have many more instruments to cover all its testing needs, and its data management system will likely need to be able to interface to U.S. Food and Drug Administration (FDA) systems, or at a minimum report in their specific format. Third, your laboratory's budget is always important. Does the budget allow for on-site hardware and software systems, with the personnel to maintain them? Is it easier to pay up-front or find a vendor willing to work with you on leasing or rental terms? (We talk about other cost considerations a bit later.)

Finally, will the lab have someone on-site or on-call to resolve technology issues, including set-up and maintenance of software systems? If your lab will have little in the way of available tech help locally, you'll want to consider the distribution model you want to use for any installed software, i.e., you may want to consider software as a service (SaaS). An increasing number of software services are hosted using cloud computing, which when done well is an increasingly reliable option. Having someone else host the software for you typically means the hosting provider will carry a non-trivial portion of responsibility for technology maintenance and security. Speaking of security, you'll also want to consider the cybersecurity (addressed later) of not only your software solution but also your overall laboratory operations. Does your organization have a cybersecurity plan already in place, or has the decision to make one been postponed? What extra investment is required to ensure your sensitive data is secure? Remember that how you rank your cybersecurity preparedness and implement a cybersecurity plan will also guide your technology investment decisions.

3.1.1.1 Laboratory informatics options

Keeping the above in mind, what are the common software solutions used within the manufacturing-based laboratory? One of the more commonly discussed options is the LIMS, a laboratory informatics solution designed to assist laboratories with managing testing workflows, data, and other aspects of their operations. The use of LIMS in production facilities and labs is not a new concept. However, little information can be found as to the percentage of today's manufacturing-based laboratories using a LIMS in their workflow. In 2020, Astrix Technology conducted a LIMS market survey, with results being compiled from "137 professionals from a wide variety of industries," with some of those industries being related to manufacturing efforts. Astrix indicated some 77 percent of the surveyed companies reported having a LIMS deployment in their lab, but Astrix didn't break out those deployments by industry, making it difficult to determine what percentage of manufacturing-related companies had a LIMS deployment in 2020.

That said, professional discussions about manufacturers and their LIMS use would seem to indicate the importance of an information management system being high in order to remain competitive in tighter economies. When also taking into consideration the importance of integrating instrumentation and produced data in a manufacturing and testing environment, a LIMS or other informatics solution appears to be increasingly critical to eliminating manual processes, improving sample management, increasing productivity, and improving regulatory conformance. This, of course, lends to the manufacturing lab's focus on safety, quality, and compliance.

A LIMS can improve laboratory workflows and workloads while enhancing safety, quality, and compliance in a number of ways. A fragmented mix of paper-based and electronic information sources can be a detriment to the traceability of or rapid accessibility to manufacturing ingredients, additives, quality control samples, standard operating procedures (SOPs), environmental monitoring data, chain of custody data, and other vital aspects of the production of goods. A well-implemented LIMS can reduce the silos of information and data, while at the same time make that information and data more secure and readily accessible. Producers of pharmaceuticals and other goods are increasingly adding automation, and as such, a LIMS is able to better integrate and support digital data and information management while improving the adaptability and flexibility of workflows. Given the regulatory demands for providing rapid proof of traceable product movement and relevant quality control data, the LIMS can also act as the central integrator and audit trail for that information. Because the LIMS improves traceability—including through its automated interfaces with instruments and other data systems—real-time monitoring of supply chain issues, quality control data, instrument use, and more is further enabled, particularly when paired with configurable dashboards and alert mechanisms. By extension, manufacturers can more rapidly act on insights gained from those real-time dashboards. This also means that the lab conducting analyses for the manufacturer can react more rapidly to issues that compromise compliance when certified to, e.g., the ISO/IEC 17025 standard. Finally, many modern LIMS tailored to a manufacturing-based industry come pre-configured out of the box with analytical and quality control workflow support tools that can be further optimized to a lab's unique workflow.

However, the LIMS is not the sole information management solution for manufacturers and the laboratories they use. Software-based information management solutions are being marketed to manufacturing-based labs in other ways. Some vendors have taken to marketing the somewhat related laboratory execution system (LES), which tends to focus more on laboratory test method execution at the process level while integrating other R&D functionalities found in, for example, electronic laboratory notebooks (ELNs). Some software vendors incorporate artificial intelligence (AI) and machine learning (ML) mechanisms into their LIMS-like manufacturing labortatory software and package it differently, e.g., as a "materials informatics platform" or an "applied sciences platform." This platform approach to software is arguably growing as a trend, so these laboratory informatics platforms may have utility in specific manufacturing-based labs. In comparison, some LIMS may or may not address these and other requirements; this functionality will be discussed further in the next subsection.

3.1.2 Features and functions

Given the above, it's clear LIMS adoption and use is important to the continued success of manufacturing-based labs. However, in most cases, a generic LIMS won't do; it's imperative the lab find a solution that meets all or most of its workflow requirements. This more often than not requires a configurable solution that enables trained users to quickly make the changes they need, if those changes make sense within the overall data structure of the LIMS. It also requires a solution that has been thoughtfully developed and continues to be carefully maintained to address the ever-shifting standard- and regulation-based requirements of the manufacturing-based laboratory. The following examines both the base features and specialty requirements of a manufacturing LIMS.

3.1.2.1 Base features

What follows is a list of system functionality important to most any manufacturing-based laboratory, with a majority of that functionality found in many vendor software solutions.

Test, sample and result management

• Sample log-in and management, with support for unique IDs
• Sample batching
• Barcode and RFID support
• End-to-end sample and inventory tracking
• Pre-defined and configurable industry-specific test and method management, including for bacteria (i.e., microbiology), heavy metals (i.e., chemistry), drug residues (i.e., pharmaceutical chemistry), and other substances
• Pre-defined and configurable industry-specific workflows
• Configurable screens and data fields
• Specification management
• Test, sampling, instrument, etc. scheduling and assignment
• Test requesting
• Data import and export
• Robust query tools
• Analytical tools, including data visualization, statistical analysis, and data mining tools
• Document and image management
• Version control
• Project management
• Method and protocol management
• Investigation management
• Facility and sampling site management
• Storage management and monitoring
• Contractor lab and outsourcing support

Quality, security, and compliance

• Quality assurance / quality control mechanisms
• Mechanisms for compliance with ISO/IEC 17025 and HACCP, including support for critical control point (CCP) specifications and limits
• Result, method, protocol, batch, and material validation, review, and release
• Data validation
• Trend and control charting for statistical analysis and measurement of uncertainty
• User qualification, performance, and training management
• Audit trails and chain of custody support
• Configurable and granular role-based security
• Configurable system access and use (i.e., authentication requirements, account usage rules, account locking, etc.)
• Electronic signature support
• Data encryption and secure communication protocols
• Archiving and retention of data and information
• Configurable data backups
• Status updates and alerts
• Environmental monitoring support
• Incident and non-conformance notification, tracking, and management

Operations management and reporting

• Configurable dashboards for monitoring, by product, process, facility, etc.
• Customizable rich-text reporting, with multiple supported output formats
• Custom and industry-specific reporting, including certificates of analysis (CoAs)
• Industry-compliant labeling
• Email integration
• Instrument interfacing and data management
• Third-party software interfacing (e.g., LES, scientific data management system [SDMS], other database)
• Data import, export, and archiving
• Instrument calibration and maintenance tracking
• Inventory and material management
• Supplier/vendor/customer management
• Integrated (or online) system help

3.1.2.2 Specialty features As noted previously, some software vendors are addressing manufacturing-based laboratories' needs beyond basic organizational LIMS functionality with their specialty solutions. A standard LIMS tailored for various manufacturing industries may already contribute to some of these wider organizational functions, as well as more advanced laboratory workflow requirements, but many may not, or may vary in what additional functionality they provide. In that regard, a LIMS vendor may also include specialized functionality that goes beyond the basic and addresses specialized activities of the manufacturer and its laboratories:

• Manage stability studies: From the pharmaceutical industry to the food and beverage industry, stability studies are an important specialty activity. These studies require careful statistical analysis, predictive modelling, sensory analysis, quantitative descriptive testing, discrimination testing, microbiology testing, and more. This translates to a need for a wide variety of analytical and visualization tools, as well as LIMS support for a wide variety of test methods and limits. A robust LIMS should have these abilities, but not all do.
• Manage recipes and formulations, as well as master and batch production records: This functionality is more in the domain of the LES or manufacturing execution system (MES). However, a few LIMS vendors may extend their LIMS to provide these features. Given the Hazard Analysis and Critical Control Points (HACCP) rules of the food and beverage industry, as well as other master and batch record requirements for other regulated industries like the pharmaceutical industry, manufacturing labs testing batches and manufacturing materials may appreciate support in this regard.
• Support molecular biology workflows: Molecular biology is an important tool in improving foods and beverages, ensuring the safety of pharmaceuticals, and meeting cannabis product testing requirements of state governments. However, not all LIMS are ideally equipped to handle related workflow aspects such as nucleic acid extraction, protein and cell isolation, and genotyping. A lab using such techniques may have to do extra due diligence in finding a manufacturing LIMS that also supports these workflow tasks.
• Take advantage of ELN functionality: Given the level of R&D to be found in manufacturing facilities, the ELN is a familiar companion to other informatics systems. A few LIMS vendors may have a built-in ELN with their LIMS or offer an ELN that comes readily integrated with the LIMS. Some elements of ELN functionality may even be found in a few solutions. At a minimum—and nodded to in the base functionality above—the LIMS should support ELN functionality through its ability to effectively connect to a third-party ELN.
• Develop regulatory-driven safety plans: The HACCP quality control method is recommended or required for food and testing labs and is an influence on ISO/IEC 17025, which applies to many other manufacturing labs. Some LIMS vendors have recognized this and integrated support for building HACCP steps into laboratory workflows. In some cases this may be as sophisticated as allowing the user to diagram HACCP in their lab or facility as a visualization tool.
• Generate schedules for environmental monitoring/testing in the facility/lab: While a LIMS can help assign and schedule a variety of laboratory tasks, broader organizational goals of testing the production environment on a scheduled, reportable basis may not be so straightforward, particularly without facility and sampling site management functionality that allows for highlighting specific test points in the facility. Even offsite or randomized testing may not be fully supported by a generic LIMS, requiring a LIMS flexible enough to compensate for the need for broader scheduled and randomized testing and retesting.
• Improve reaction time to non-conformances: Many LIMS will have some basic form of non-conformance and incident management tools, but the robustness and extensibility of that functionality may be lacking. Can it send an SMS or email to the appropriate supplier in real-time when a pre-defined set of circumstances concerning that supplier's ingredients occurs? Can it re-prioritize or pause other related activities that are scheduled due to the identified non-conformance or incident? This is a useful area of functionality for the potential LIMS buyer to confirm with a vendor.
• Improve audit readiness and reporting: A LIMS worth its weight will have a robust audit trail, to be sure. But can your LIMS help you audit your suppliers? Can it capture internal audit data on-demand and directly from the facility floor via mobile-friendly forms? Can HACCP- and audit-related data be flagged as such to make retrieval more efficient for audit purposes? These and other considerations may be important to a manufacturing facility, and not all LIMS for manufacturers can provide that functionality.

3.1.3 Cybersecurity considerations

From law firms to automotive manufacturers, the need to address cybersecurity is increasingly apparent. In 2018, the Center for Strategic & International Studies estimated that cybercrime causes close to $600 billion in damages to the global economy every year, though due to underreporting of crimes, that number may be much higher. That number also likely doesn't take into account lost business, fines, litigation, and intangible losses In the end, businesses of all sizes average about $200,000 in losses due to a cybersecurity incident, and nearly 60 percent of small and midsize businesses go bankrupt within six months because of it.

Manufacturing-based laboratories are no exception, regardless of business size. Even tiny labs whose primary digital footprint is a WordPress website advertising their lab are at risk, as hackers could still spread malware, steal user data, add the website to a bot network, hack the site for the learning experience, or even hack it just for fun. Even more importantly are those labs performing digital data management tasks that handle sensitive proprietary manufacturer data, requiring additional cybersecurity considerations.

A manufacturer and its associated laboratories can integrate cybersecurity thinking into its laboratory informatics product selection in several ways. First, the organization should have a cybersecurity plan in place, or if not, it should be on the radar. This is a good resource to tap into in regards to deciding what cybersecurity considerations should be made for the software. Can the software help your organization meet your cybersecurity goals? What regulatory requirements for your lab are or are not covered by the software? Another tool to consider—which may have been used in any prior cybersecurity planning efforts—is a cybersecurity framework. Many, but not all, cybersecurity frameworks include a catalog of security controls. Each control is "a safeguard or countermeasure prescribed for an information system or an organization designed to protect the confidentiality, integrity, and availability of its information and to meet a set of defined security requirements." These controls give the implementing organization a concrete set of configurable goals to apply to their overall cybersecurity strategy. Other frameworks may be less oriented to security controls and more program-based or risk-based. Choosing the best frameworks will likely depend on multiple factors, including the organization's industry type, the amount of technical expertise within the organization, the budget, the organizational goals, the amount of buy-in from key organizational stakeholders, and those stakeholders' preferred approach.

Finally, having an organizational cybersecurity plan that incorporates one or more cybersecurity frameworks gives the laboratory ample opportunity to apply stated goals and chosen security controls to the evaluation and selection process for its informatics software. In particular, a user requirements specification (URS) that incorporates cybersecurity considerations will certainly help a laboratory with meeting regulatory requirements while also protecting its data systems. A USR that is pre-built with cybersecurity controls in mind—such as LIMSpec, discussed later—makes the evaluation process even easier.

3.1.4 Regulatory compliance considerations

Without a doubt, it's vital that manufacturing-based laboratories operate within the bounds of a regulatory atmosphere, not only to better ensure the best consumer satisfaction outcomes but also to ensure the quality of analytical results, the safety of end users, and the promise of maintaining traceability across the manufacturing and distribution chain. Maintaining regulatory compliance requires deliberate approaches to developing and enforcing processes and procedures, quality training, consistent communication, and knowledgeable personnel. It also requires a top-down appreciation and commitment to a culture of quality. From ISO/TS 22002-1:2009 Prerequisite programmes on food safety — Part 1: Food manufacturing and ISO 10993-1:2018 Biological evaluation of medical devices — Part 1: Evaluation and testing within a risk management process to 21 CFR Part 175 and 176 (concerning the materials used to package food) and current good manufacturing practice (cGMP), laboratories have much to consider in regards to what standards and regulations impact them.

That said, consider approaching the question of regulatory compliance from the standpoint of adopting standards. Consider first that the risks and consequences of performing a task poorly drives regulation and, more preferably, standardization, which in turn moves the "goalposts" of quality and security among organizations. In the case of regulations, those organization that get caught not conforming to the necessary regulations tend to suffer negative consequences, providing some incentive for them to improve organizational processes and procedures.

One of the downsides of regulations is that they can at times be "imprecise" or "disconnected" from what actually occurs within the organization and its information systems. Rather than focusing heavily on regulatory conformance, well-designed standards may, when adopted, provide a clearer path of opportunity for organizations to improve their operational culture and outcomes, particularly since standards are usually developed with a broader consensus of interested individuals with expertise in a given field. In turn, the organizations that adopt well-designed standards likely have a better chance of conforming to the regulations they must, and they'll likely have more interest in maintaining and improving the goalposts of quality and security in the lab.

Additionally, reputable software developers of laboratory informatics software will not only adopt their own industry standards for software development but also understand the standards and regulations that affect manufacturing-based laboratories. In turn, the developed software should meet regulations and standards, help the laboratory comply with its regulations and standards, and be of reliably good quality.

If you're a potential buyer of a laboratory informatics solution, it may be that you know a bit about your laboratory's workflow and a few of the regulations and standards that influence how that workflow is conducted, but you're not entirely informed about all the regulations and standards that affect your lab. Turning to a URS such as LIMSpec—which was developed around laboratory regulations and standards—and reviewing the various statements contained within may be necessary to help further inform you. Additionally, as you investigate various informatics options, you can then use the requirements in the URS as a base for your laboratory's own requirements list. Using the categories and their subdivisions, you can then add those requirements that are unique to your laboratory and industry that are not sufficiently covered by the base URS. As you review the various options available to you and narrow down your search, your own list of requirements can be used as both as a personal checklist and as a requirements list you hand over to the vendor you query. And since your URS is based off the standards and regulations affecting your lab, you can feel more confident in your acquisition and its integration into your laboratory workflow.

3.1.5 System flexibility

Before selecting a solution, your laboratory should also have internal discussions about how diversified its offered services are, as well as what the future may bring to the lab. If, for example, your lab is currently configured for microbiological testing, does your existing laboratory informatics system—or the ones you may be considering—have the flexibility to add other types of testing, protocols, and workflows? Will you be doing the footwork to add them, or will the vendor of your system support you in that effort? If you're a start-up, will your lab be focusing solely on a specific type of testing and expand into other types of analytical work later, or will your test menu need to be much broader right from the start? In most of these cases, you'll desire a LIMS that is flexible enough to allow for not only running the specific tests you need now, but also sufficiently expandable for any future testing services your lab may conduct in the mid- and long-term. Having the ability to create and customize sample registration screens, test protocols, labels, reports, specification limit sets, measurement units, and substrates/matrices while being able to interface with practically most any instrument and software system required will go a long way towards making your expanding test menu and workflows integrates as smoothly as possible.

Such a system will typically be marketed as being highly user-configurable, giving labs a relatively painless means to adapt to rapid changes in test volume and type over time. However, once you've internally addressed current and anticipated future growth, your lab will want to learn what explicitly makes any given vendor's system user-configurable. How easy is it to configure the system to new tests? Add custom reports? What knowledge or skills will be required of your lab in order to make the necessary changes, i.e., will your staff require programming skills, or are the administrator and advanced user functions robust enough to make changes without hard-coding? These and other such questions should be fully addressed by the vendor in order to set your mind at ease towards a system's stated flexibility. Ultimately, you want the system to be flexible enough to change with the laboratory—and industry—itself, while minimizing overall costs and reducing the time required to make any necessary modifications.

3.1.6 Cost considerations

First, you'll want to be clear on what will be included in the sales agreement. Whether through an estimate or statement of work (SOW), it is important it includes exactly what is expected, being as specific as possible, since this will be the entire contractual obligation for both you the buyer and them the vendor. Note that line items may differ slightly from system to system, according to what features and functions are included by default with each vendor's solution and which, if any, are additional. Also keep in mind that any hourly amount in the the estimate or SOW is usually a best estimate; however, if sufficient attention to detailed requirements has been given, then it should be quite accurate, and in fact the final cost may even be below the quoted cost if you prioritize your own obligations so that the vendor's hours are used sparingly and efficiently.

The estimate or SOW should optimally include:

• licensing or subscription rates;
• required core items to meet federal, state, and local regulations;
• additional optional items and totals; and
• required services (implementation, maintenance and support, optional add-ons).

There are two primary ways to price a laboratory informatics solution: a one-time license fee or a subscription rate (cloud-hosted SaaS). If you have your own dedicated IT department and staff, you may prefer the former (although many system administrators are just as happy to let it be hosted elsewhere rather than add to their workload). Otherwise, a SaaS subscription may well be the better and more cost-effective way to go (since the primary IT cost is simply internet access). This item will be part of your up-front cost and, in the case of subscription, it will also figure into your first year and ongoing costs; otherwise only associated maintenance, support, and warranty (MSW) will figure in. Typically, your first year's subscription costs will be due at signing. More often, the vendor may require three months or even the first year up front, so be prepared to factor that into up-front costs. However, it still is almost always less expensive at the outset (and over time, if you factor in IT costs and annual MSW) than paying for a license fee.

In addition to the two types of software pricing, there are also sub-types. Generally these are based on the number of users (or, in some cases, "nodes," which are simply any entities that access the informatics system, including other systems, instruments, etc.). How these are counted can vary.

• Named users: This method bases pricing on the actual individual users of the system, even if they only log in sporadically. Users may not use each other's logins (this is a no-no regardless of pricing structure, for good laboratory practice and other regulatory reasons).
• Concurrent users: This bases pricing on the maximum number of users who will be logged in at any given time. You can define an unlimited number of named users in the system, each with their own login credentials. However, only the number of concurrent users specified in the license or subscription may be logged in at any one time. For example, you may have 10 staff, but due to work processes, shifts, etc., only up to six might ever be logged in simultaneously. Whereas this would require a named user license for 10, it would only require a concurrent user license for six.
• Unlimited users: In the case of very large labs (typically 30 to 50 and up), the license or subscription may simply be a flat fee that allows any number of users.

The line items in the estimate or SOW should reflect these nuances, as well as whether the listed costs are monthly or annual (for subscription services), hourly (typically for support and training), or a fixed one-time cost. Additionally, be cautious with fixed costs, as they typically represent one of two possible scenarios:

1. Final fixed cost: In this case, the cost has been figured by the vendor so as to cover their worst-case hourly labor total. If a line item (e.g., an interface) is not "worst case," then you are overpaying.
2. "Expandable" fixed cost: This is as bad as final fixed cost, and maybe even worse because it's almost a case of "bait-and-switch," popping up as a surprise. The initial "fixed cost" number is low, and additional hourly services are needed to actually deliver the item. This will have been provided for somewhere in the small print.

The bottom line is that everything in a laboratory informatics solution is really either licensing or hourly services. Just be careful if they are portrayed as anything else.

It is important to be clear which category each line item falls under when figuring costs: up-front (due upon signing), annual, or ongoing (e.g., SaaS subscription). It is useful to clearly lay out each and compute initial costs, as well as first-year and subsequent years' costings. For example, your initial obligation may be as little as your first year's subscription plus the first 40 hours of services. Different vendors have different policies, however, and you may be required to pay for your first full year's subscription and all services, or some other combination. Normally, though, any instrument interface or other service charges aren't due until the they are implemented, which may be a few weeks or even a month down the road. This may depend on your budget, complexity of the SOW, and urgency. Your first year's expenses will include everything, including initial license fees; all setup and training; any interfaces and additional configurations or customization; and first annual MSW. (If this isn't included in the SaaS subscription, then it usually commences on full system delivery). Afterwards, your subscription and MSW will be the only ongoing expenses (included as one in this example), unless you choose to have additional interfaces or other services performed at any time.

3.2 Implementation

If you've ever worked through a system implementation process with a vendor, it was hopefully a smooth process. However, there are plenty of horror stories out there, highlighting the need of the laboratory to discuss in detail how a potential vendor will handle installation, validation, and training for the informatics solution. Does the vendor truly understand the industry and your needs? Does the vendor assign a project manager who will work with you, from planning to go-live and beyond? Can they offer you references of other labs who have gone through implementation so you can compare notes with those labs? How much attention does the potential vendor give to related issues such as data integrity of migrated data? Do they have the means to properly handle your legacy data? And are they able to work with your schedule, even if it means implementing software at off-peak work hours?

As you finally get down to the ultimate decision on which vendor to work with, you may wish to start setting up an implementation checklist as part of your early project planning. Do you receive a help desk account as part of the implementation process, and if so, what information is included? If not, you'll need to keep track of specific details such as business associate agreement (BAA), sales agreement, scope documents, welcome letters, documentation, and approved staff who can utilize the vendor's support. You'll likely need to share other configuration details with the vendor, including time zone requirements, DNS and URL requirements, up-time monitors, and administrative account requirements. Finally, you'll want to ensure you and the vendor are on the same page concerning any additional customization, integration, and system validation requirements, ensuring the roll-out period is pain-free and efficient.

3.2.1 Internal and external integrations

Manufacturing-based labs acquire data management software for many reasons, including improving accuracy, saving time, increasing productivity, and adding capabilities. One way of doing all of those activities is to integrate or interface the manufacturer's systems, databases, and instruments so that human error is greatly reduced or eliminated, workflows are automated and sped up, and each component's capabilities are brought into play in the most efficient and effective ways possible. In the world of manufacturing, choosing what gets integrated makes for a challenging decision, however, given the other systems that may be at work in the organization. Other systems that may used in the manufacturing enterprise include:

• manufacturing execution systems (MES)
• quality management systems (QMS)
• enterprise resource planning systems (ERP)
• supervisory control and data acquisition systems (SCADA)
• statistical process control systems (SPC)
• product life cycle management systems (PLM)
• distributed control systems (DCS)

These systems all have their strengths within the manufacturing enterprise but typically lack functionality necessary for laboratory processes. Not all systems need to be integrated with a laboratory system like a LIMS, but some may benefit, for example with the MES and QMS. As such, you'll want to inquire with the vendor about its solution's hardware and software integration capabilities. Is it designed to interface with every laboratory instrument or software that can output any readable electronic file? Or are integrations limited to certain instruments and systems? How does it connect, i.e., what protocols does the software depend on to connect with other systems? Does the system allow a user to map their own file imports and exports? Can system processes be set to detect new instances of file outputs at regular intervals? Ask these and other questions to make sure the vendor clearly describes what internal and external integrations are supported with their application.

In many cases, a vendor's LIMS solution will have instrument integration capability built into the software, but occasionally such interfaces are separate from the main software. Today's instrument interfaces are generally built on standardized communication protocols such as RS-232, RS-422, IEEE-488 (GPIB), USB, Ethernet, and more. The LIMS that can support such instrument integrations is increasingly vital to the manufacturing-based laboratory. Manufacturing labs may also want their laboratory informatics solution to be able to communicate with other software and databases. This is often done using application programming interfaces (APIs) that depend on web services implementation protocols such as REST and SOAP. These messaging protocols actually allow for the creation of an API that receives communication requests and sends responses between two software systems. A more practical example is wanting your laboratory informatics solution to communicate with an enterprise resource planning (ERP) application. Perhaps the ERP system needs to create sample batches within the informatics solution, and when testing is done, have the results returned to the ERP. APIs and communication protocols make this happen.

3.3 MSW, updates, and other contracted services

The MSW offered with the vendor's solution is almost as important as the solution itself. The laboratory informatics solution you acquire is more than than the software you operate: it's mission-critical and deserves having a reliable and responsive team with the necessary resources to ensure it remains operational. Downtime can negatively affect both immediate customer satisfaction and your reputation. As such, it's imperative you ask the vendor about the details of its MSW, making sure you understand what is and isn't covered, as well as how much it will cost. Cost-wise, industry norms are anywhere from 15% to 25% of either the license fee or total contract, levied annually to provide this coverage. Alternatively, it may simply be included with your subscription. The MSW will include a specified number of support and maintenance hours or guarantees. The actual warranty should be unlimited for as long as the MSW or subscription is kept current.

Maintenance includes any and all work necessary to keep your system working as designed. It should include updates, patches, or fixes, and most if not all upgrades. (Note, however, a major upgrade to a totally new edition may not be covered, but it may come at a negotiable, significantly lower cost.) The support aspect of MSW generally consists of a specified number of hours dedicated more to helping you with the operation of the system rather than "fixing" anything. Support includes guidance on training, password or login support, and more. Finally, with any professional application you also expect to have a warranty. The warranty should cover anything that doesn't work that otherwise should for the designated period of time. That includes any standard features and functions, as well as any additional ones that were delivered and signed off on, and any other work performed by the vendor or its representatives. However, a typical warranty does not cover anything that was working fine, but upon being manipulated in a way beyond normal operation the functionality ceased. In these cases, you'll probably have to pay to get it fixed.

Beyond the MSW, additional updates and services related to the system may also be required. No matter how well it is pre-configured, any professional laboratory informatics solution will require some amount of standard setup to reflect your particular lab. This includes adding lab branding and demographics for reports and certificates; entering users, their roles, and access permissions; adding and/or modifying tests and workflows; renaming fields; adding or hiding fields; setting up a web portal; and implementing interfaces. Equally indispensable is proper training for both users and administrators. And of course you may later find that you would like additional features or functions. These and other services may prove particularly useful to the laboratory with little in the way of IT and systems expertise. As such, the vendor may provide one or more of the following as a billable service for the laboratory:

• initial implementation meeting (e.g., initial planning, identify delta, set schedule)
• project management
• requirements gathering and documentation
• initial setup
• user and administrator training
• configuration and customization
• interface development and implementation
• custom screen and field development
• custom functionality development
• custom reports and labels
• custom triggers and alerts
• validation or acceptance testing (to a third-party standard or certification, or to agreed manufacturer specs)

3.4 How a user requirements specification fits into the entire process (LIMSpec)

Merriam-Webster defines a "specification" as "a detailed precise presentation of something or of a plan or proposal for something." In other words, an existing or theoretical product, concept, or idea is presented in detail for a particular audience. In a broad sense, detailing the specifics about a project, concept, or idea to others is just common sense. This applies just as well to the world of software development, where a software requirements specification is essential for preventing the second most commonly cited reason for project failure: poor requirements management.

In fact, the ISO/IEC/IEEE 29148:2018 standard (a conglomeration of what was formerly IEEE 830 and other standards) is in place to help specify "the required processes implemented in the engineering activities that result in requirements for systems and software products" and provide guidelines for how to apply those requirements. The standard describes the characteristics that make up quality software requirement development, including aspects such as:

• correctly describing system behavior;
• effectively removing ambiguity from the language used;
• completely covering the system behavior and features;
• accurately prioritizing and ranking the requirements; and
• unequivocally ensuring the requirements are testable, modifiable, and traceable.

A requirement typically comes in the form of a statement that begins with "the system/user/vendor shall/should ..." and focuses on a provided service, reaction to input, or expected behavior in a given situation. The statement may be abstract (high-level) or specific and detailed to a precise function. The statement may also be of a functional nature, describing functionality or services in detail, or of a non-functional nature, describing the constraints of a given functionality or service and how it's rendered. An example of a functional software requirement could be "the user shall be able to query either all of the initial set of databases or select a subset from it." This statement describes specific functionality the system should have. On the other hand, a non-functional requirement, for example, may state "the system's query tool shall conform to the ABC 123-2014 standard." The statement describes a constraint placed upon the system's query functionality.

This is where a requirements specification shines, not only for the software developer but also for those acquiring the software. A set of development requirements, compiled in the form of a software requirements specification, can serve to strengthen the software development process. For those acquiring the software, a set of user requirements, compiled in the form of a URS, can be used for the selection and acquisition of software or a service. In the case of the URS, the acquiring business can approach this several ways. The simple way would be to essentially take the vendor at the word in regards to what they say their system can and can't do, agreeing formally to their description and taking responsibility that it will cover all the applicable regulations required by your business. However, this method isn't comprehensive and leaves the business open to not being able to fully meet its goals.

The other method has the URS be specific to your business' needs. The process is more work but leaves less to chance. Developing your own URS isn't always straightforward. Often times, the developed document turns into a mix of "wishlist" requirements from potential and active clients, as well as regulation-mandated requirements. The wishlist items aren't necessarily ignored by developers, but the URS should in fact clearly prioritize requirements as "nice to have" or "essential to system operation," or something in between. Whatever the URS looks like in the end, it's ultimately up to the vendor to be able to demonstrate how the software does and does not meet its requirements.

In the latter half of this guide, you'll be given an opportunity to see an example of a URS for the manufacturing industry in the form of LIMSpec, an evolving set of software requirements specifications for laboratory informatics systems. Built from requirements found in ASTM E1578-18 Standard Guide for Laboratory Informatics, as well as dozens of other standards and regulations, we will use LIMSpec to demonstrate how a URS can be put to use, while also showing you how an informatics system can help you laboratory better meet regulatory requirements.