By Bob Beliveau.
Many labs begin operations by capturing data in spreadsheets. Excel is ubiquitous and versatile, and the ability to add new columns to capture unforeseen data is a plus. Furthermore, instrument vendors provide spreadsheet export capabilities that make Excel data exchanges natural.
As data volume grows and the lab processes mature, it becomes desirable to capture data in an IT system or database. The benefits of centralized storage and reporting as well as automated data capture and analysis drive labs to this decision. This is the point where most labs begin surveying the Laboratory Information Management Systems (LIMS) market. First they capture their requirements. They develop a short list of vendors. Then they stack the products up against their requirements; carefully balancing fit, function, and cost.
I know this process well; having led similar projects as an IT department-head and as an IT consultant. A few years ago, my company helped select a LIMS for a clinical immunology project. Once the vendor was selected, we played no further part in the project and I’ve wondered since how that project unfolded. Recently I got my answer. The company’s new head of IT was very displeased with the system and was looking forward to the day it could be replaced.
Why are so many companies disappointed with their LIMS systems?
Talk to some of your industry colleagues or search around the web and you’re bound to find stories of folks who feel shackled by their LIMS. Notoriously inflexible, they require substantial customization to closely match your workflow. Alternatively, you can accept the “out-of-the-box” solution and be left with large functionality gaps.
Couple these shortcomings with the dynamism of modern laboratories and many LIMS systems have trouble keeping up. And resource-poor IT departments have to stretch to provide maintenance and support; let alone enhancements. No wonder people feel trapped.
The Lack of Context
The aspect of LIMS systems that most disappoints me is the lack of context. Let me explain. At their core, LIMS systems are sample-oriented. But samples are always related to something bigger; a manufacturing lot, a clinical trial, an experiment, a facility/room, etc. When you log-in a sample, you annotate it so that it can be related back to its source. But these are simply annotations and don’t help you track your end-to-end business process.
For instance, clinical LIMS are great at tracking incoming samples. But they don’t typically know about the trial design and how many overall samples to expect over the course of the study. Simply put, they can tell you how many samples you’ve received, but they can’t tell you how many samples are yet to be received. They also can’t help you locate straggler samples when your team is scrambling to close-out a clinical study. Since the LIMS lacks context, it can’t answer these very important questions.
Focus on the Questions
It was Albert Einstein who said “If I had an hour to solve a problem and my life depended on the solution, I would spend the first 55 minutes determining the proper question to ask, for once I know the proper question, I could solve the problem in less than five minutes.”
This approach works when considering a LIMS too. Think about the questions you want to answer when contemplating your LIMS system. For instance, here are 3 questions that your LIMS system should be able to answer:
- What are all the measurements and characteristics about my material? (i.e. Concentration, purity, identity, etc.)
- What ingredients went into making my material and where was it used? (backward/forwards traceability)
- How full are my freezers and do we need to purchase a new refrigerator because of overflow?
LIMS systems are only designed to answer the first question. If you want to answer the second question, you need to implement an Enterprise Resource Planning (ERP) or a Materials Resource Planning (MRP) system. If your focus is the third question, you need a Biorepository.
This is madness. One system should be able to answer all 3 classes of questions. That is why I felt compelled to write this article. Collectively, we’ve been trained to think of vertical “best-in-breed” IT systems. By accepting these artificial boundaries, we haven’t been able to marry cross-domain information without costly (and complex) systems integration.
So when your lab grows to the point where you’re considering automation, think about the questions you want answered. Your Lab doesn’t need a LIMS, it needs a system that can answer your most important questions