By tracing the evolution of QSI's WinLIMS from the early days, Clive Collier shows how a LIMS is now within the reach of even the smallest laboratory.

By tracing the evolution of QSI's WinLIMS from the early days, Clive Collier shows how a LIMS is now within the reach of even the smallest laboratory.

The acronym LIMS entered the laboratory vocabulary in the 1970s, when analytical instruments capable of generating large quantities of data first appeared on the scene. The first rudimentary LIMS were developed in-house by large organisations needing to improve the accuracy and efficiency of their internal data acquisition and reporting. The next step was the emergence of specialist companies that designed custom-built and therefore highly individual systems to suit the needs of specific laboratories. Analytical instrument companies, realising the commercial potential of laboratory data management, began to introduce data systems to accompany their own instruments, and the first commercial LIMS appeared in the 1980's. The same transitions that were taking place in other areas of Information Technology were happening here too, with the move from dedicated hardware, to minicomputer, to PC and the efficiency of client/server architecture. By the 1990's, most laboratory data handling was PC based, and since then LIMS has embraced many of the developments enabled by the remarkable processing power of the modern PC, such as web connectivity and wireless computing.

Remarkably, in spite of these advances some large organisations, particularly in the pharmaceutical industry, are still developing in-house solutions, and spending large sums of money in the process. Many others are devoting considerable time and resource to protracted and expensive product selection and evaluation exercises, perhaps motivated by fears of buying a product that might not perform as expected - may not be so surprising when a typical LIMS for a small contract lab could cost £10-20,000. This fear can be, and often is, enhanced by industry jargon - after all, analytical chemists are not necessarily IT experts - a process known in the computer industry as FUD. FUD stands for Fear, Uncertainty and Doubt (or Fear Under Duress!), and was first coined to describe a technique used in computer marketing to ignite these emotions about competitive products. A similar technique can be used to make the product seem immensely complicated, possibly one way of justifying a high price.

There are many data handling tools on the market that have evolved in a similar way to LIMS and over a similar time frame. As an example, consider accounting packages. Before the advent of PC's, the only companies that could afford computerised accounting were large international corporations, often running on mainframes. Then came PC's, and accounting software could be had for £7-8K for a stand-alone package, still a significant investment for a small business. Then networking arrived, enabling multi-user working and further reducing the costs. Now, companies like Sage offer highly competent accounting packages for as little as £100. At this price, any minor difficulties that might arise from not having a tailor-made solution are far outweighed by the benefits. Most users of any popular PC software package - word processor, spreadsheet, etc - recognise that it contains far more functionality than they will probably ever need, and will be regularly updated and properly supported, so are quite comfortable with buying it off the shelf. Where is LIMS in this evolutionary process? At QSI we believe that we are now at virtually the same point, and that LIMS has evolved so that it is not only affordable by almost any laboratory, but can handle virtually any functional demand placed upon it. This is an easy statement to make, rather more difficult to justify, but we think the arguments are convincing.

According to one estimate, there are 14-15,000 laboratories of all kinds in the UK, of which only 20% have a LIMS installed, so there are many laboratories out there that could enjoy the benefits of a LIMS. However the relatively small number of laboratories with LIMS installed has kept the price of these systems artificially high, and also limited the funds available to LIMS companies for R&D, particularly as many are parts of large corporations that demand a bottom line result. When QSI was formed to develop the core functionality for WinLIMS, released in 1992, it was in an environment in which a high proportion of competitive LIMS installations were failing - as high as 60%.

Our first priority was clearly to establish why the failure rate was so high, and investigation showed that there were three main reasons: 1) Poor technical communication between client and vendor. Customer specifications, sometimes generated by people with limited data processing expertise, didn't always convey to the supplier exactly what was wanted, or demanded that the system should work in a non-ideal way (in software engineering terms) - rather like trying to specify a motor car without understanding the internal combustion engine. 2) A lack of resource at client sites. 3) A failure to include all end users in the specification process These conspired to create solutions that sometimes didn't deliver the client's expectations, required lengthy commissioning and revision, or encountered resistance from the people charged with using the system day to day.

The QSI strategy that was devised to address this situation can be summarised as: 1) No sale without a comprehensive site survey, however detailed the potential client's written specification might be. If this was denied, QSI would reluctantly decline the business. 2) Concentrate development investment on industry-specific code rather than client-specific code. Over the years, this has led to the accumulation of a huge knowledge base on the needs of a whole range of different industry sectors, and the creation of software applications matching those requirements. 3) The parallel development of wholly-owned international subsidiaries, not only to promote and support the product but by pooling world-wide knowledge, to ensure its global relevance. 4) A corporate decision to re-invest all profits in R&D. By using this strategy installations have become robust, and by not having to "re-invent the wheel" with specially developed custom code for every client, our programmers have been able to concentrate on developments that add real value for the client - namely new functionality. This focus has led to a rapid expansion of the functional scope of the basic WinLIMS package. The QSI support operation achieved the highest overall score in the latest "Forrester Report", an international survey of LIMS customers by Scientific Computing magazine. This is all very well, but we still have to justify the claim that LIMS is now a tool for every laboratory. In order for that to be true, we would suggest that a number of conditions must be satisfied: 1) The functionality provided must be capable of handling virtually any laboratory task situation without modification, and be out-of-the-box. 2) Installation must be straightforward, with minimal disruption to normal workflow and laboratory revenue. 3) User training and familiarisation must be easy to implement. 4)  Support must be immediately and economically available. 5) Initial and running costs must be acceptable. QSI believe that we can certainly satisfy the first four of these criteria, but for many laboratories, the historically high cost of LIMS has been the principal barrier to its acquisition. We believe that our strategy has led to a product that is sufficiently mature, comprehensive and robust such that we can now amortise the development costs and dramatically reduce the cost to the user. To that end we have introduced WinLIMS Rental, a concept that provides full-functionality LIMS software for a few hundred pounds a month - typically £265 per month for a three-user system. Let's now compare the rental charge with the very real cost savings that can be achieved. The following information is from a company that recently implemented WinLIMS Rental: • The company has 250 product specifications • On average it ships 20 batches of product per day • Customers require COA and SQC reports • Management requires monthly productivity and summary reports • The lab uses 20 instruments that are scheduled for calibration and maintenance with continuous control charts • The compensation package for a trained lab technician (total of salary and benefits) is £23,000 per year • Their WinLIMS monthly rental fee is £265 A time and motion study was conducted to look at the impact of the new LIMS. Looking at just one WinLIMS function, automated reporting, and comparing it with the previous error-prone and time-consuming spreadsheet system, significant savings were immediately identified:

With an average 8-hour working day, this is a saving of over 70 man-days, or more than £6400 a year. Put another way, the monthly cost saving achieved by this piece of functionality alone is more than twice the monthly rental fee. Other functions, like automatic invoicing, would yield parallel savings, not to mention those accruing indirectly from the elimination of data transcription errors and rework. Direct costs, however, are not the only issue, and the concept also attempts to address some of the other perceived risks that have, in our experience, discouraged potential LIMS users in the past:

Clive Collier has 21 years experience within the LIMS industry sector and has been the Managing Director for Quality Systems International (UK) Ltd for 10 years