Printer Friendly

Selection of a LIMS systems in a small laboratory.

Selection of a LIMS Systems in a Small Laboratory

Since the early 1970's, microprocessor-controlled instruments, programmable integrators, and other laboratory computers have brought great benefits to analytical laboratories by increasing both the sophistication and quantity of testing which can be performed routinely. Hand-in-hand with this has been a nearly exponential increase in the amount of data to be handled. In attempting to manage this larger flow of information it is inevitable that sooner or later even small laboratories begin to consider the purchase of a Laboratory Information Management System. An integrated, computerized LIMS presents vast potential for improving communication, quality control, and productivity in any analytical laboratory. The problem usually faced by the operator of a small or medium-sized laboratory is that the need may be patently obvious but the cost appears prohibitive.

The process of selecting a LIMS system touches on virtually every aspect of the laboratory operation. The decision process must consider the complete information-handling task from sample log-in through data collection to reporting the result, sending an invoice, and disposing of the sample. An additional practical problem is most laboratories already have in place microprocessor-controlled hardware which represents a substantial investment. These must be able to communicate with the LIMS system without creating a new set of data management problems.

At the start We entered this decision-making process as a commercial environmental testing laboratory with 30 employees and an annual workload of approximately 15,000 samples. In the lab we had GC, GC-MS, HPLC, atomic absorption, ICP, and autoanalyzer instrumentation from five major vendors. Each had its own level of capability to do "on-board" data reduction; in addition there were the usual suite of wet chemistry procedures calculated and reported manually. The starting point was a general overview of our data management needs without considering the capabilities of any commercially available LIMS products. This paralleled the initial systems analysis of any data processing problem, and is an important exercise in understanding the information flow in the lab and the inherent strengths and weaknesses of existing procedures. We also considered carefully how large our workload might grow over an expected five-year system lifetime. This led to a detailed functional description of the tasks we expected the LIMS system to perform. Since this analysis was being carried out in the midst of a rapidly increasing workload, it led to a number of interim improvements in our existing manual record keeping system. Our analysis of the major components of a LIMS system identified four major functions.

Sample tracking and control This is the overall area of communicating to the laboratory the work to be done. In many respects it is very similar to the order processing system of a custom manufacturing operation. It begins with sample log-in which records samples received, who the client is, what tests are required, when and where the report is to be sent, and disposal instructions for the completed samples. The manual equivalent is recording information in a logbook where individual analysts look up the required information.

In this area the LIMS system has potential to become an important supervisory tool since it provides rapid, clear communication of sample information to individual analysts in the lab. It also improves scheduling and productivity as the analyst can readily identify all samples on hand needing a particular test and when the results are required. Rather than the "first-in, first-out" approach to scheduling that usually goes with manual record-keeping, there can be better commitments to promised delivery dates. Productivity and quality control improvements result from longer analytical runs for each test.

It was as this stage of LIMS selection that the temptation to develop a system in-house was difficult to resist: after all, we knew better than anyone else what we wanted. In addition we had a good collection of PC-hardware readily available in the lab, and we had an enthusiastic staff with reasonable programming proficiency. Following some initial consultation with our systems group on data base design we were able to develop a worklist management system which gave us many of the sample tracking features we wanted. This included design of sample entry screens which incorporated many error-trapping features commonly found in commercially available LIMS software and reporting features which greatly improved the communication of sample information to analysts in the lab. However, once we began to consider the problems of collating and reporting analytical results and introducing a multi-user capability to the system, the complexity of the problems quickly grew beyond our capability.

A cursory survey of other laboratories has revealed our experience at in-house development was not unusual and in fact serves to illustrate some of the technical problems inherent in LIMS design. In the jargon of professional programmers, a LIMS system is a "data-intensive", "user-intensive" data processing application. Standard data processing system design tends to be overwhelmed by the sheer volume of information handled through a LIMS system and by the problems created by simultaneous, multi-user access to a single data base. Although some fourth-generation database languages (eg. Oracle) go a long way toward addressing some of these problems, inherent limitations of MS-DOS in handling multi-user, multi-tasking applications places practical limits on what a PC based LIMS running under MS-DOS can do.

Management and reporting of analytical results This aspect of a LIMS system separates it from a standard data processing application. Carrying the analogy to a manufacturing operation a step further, a fully functional LIMS must deliver the product (ie. data) to the client. In analyzing the flow of information in our laboratory we found it to be the most complex part of the LIMS selection process since it had to address a variety of existing procedures for getting data from the laboratory bench to a final, reportable format. For some manual tests (eg. pH value, BOD) the system would be a simple extension of manual reporting which would eliminate much of the manual data transcription. In other cases (eg. GC-MS priority pollutant scans) existing procedures already involved electronic data transfer and a significant amount of computerized data reduction using spreadsheet on PCs. An essential feature we required was an ability to retain these data reduction protocols without creating any new manual transcription steps.

Data base query This incorporates several overall management aspects of implementing a LIMS system. One feature we require is an ability to report quickly to a client inquiry on the status of work in progress. In a manual reporting system this can hinder productivity because it invariably means an interruption in work for at least two people. Incorporating an on-line inquiry feature allows faster, more accurate reporting without interrupting the laboratory work flow. Additional database query features would provide reporting of historical data for analysis of workload trends and related general management activity.

Our experience with in-house design of a sample tracking system showed us an ability to custom-tailor the LIMS system to specific sample handling practices without altering the system source code was a critical feature. We also looked for flexibility in the system's ability to grow with us and, if possible, to be transportable to larger hardware as needs grew.

Invoicing We identified this as the fourth major function as part of an integrated LIMS system for our laboratory. Although it is an accounting function quite separate from the technical aspects it is nonetheless essential to the operation of a commercial laboratory. By integrating a computerized billing system in the overall data management system we hoped to streamline invoice preparation and reduce costs.

We don't need to look beyond this year's Pittsburg Conference to see the development of commercial LIMS systems has come of age: no less than 30 vendors had software or integrated hardware/software products. Systems offered by most vendors are based on mainframe or multi-tasking minicomputers. The micro-VAX series from Digital Equipment Corporation is the most popular. In addition, several vendors have software products designed to run on PC hardware as stand-alone single workstations, on local area networks, or multi-user microcomputers.

The decision Based on our existing and projected workload requirements we ruled out the VAX-based systems on costs ($160,000-$200,000 for a working system). This led to a fairly detailed examination of two systems designed to run on PC-hardware, which appear to offer a working system for less than half the cost of VAX-based systems.

SAM from Radian Corporation has been around for several years. We found it to be technically credible as a full-function LIMS product, offering all of the technical features we required. As with any of the PC-based systems we examined, it does not offer direct data acquisition from instruments, but this is not among our technical selection criteria. SAM allows relatively easy transfer of information from Lotus 1-2-3 and other software through their Electronic Benchsheet utility as well as design of custom sample worksheets for data entry and computation on multi-parameter tests. It is designed to run as a single workstation or on a local area network using any of several popular network configurations.

NWA LIMS from Northwest Analytical is a new software product officially introduced at this year's Pittsburg Conference. It is the first commercial system to use PC-hardware operating under the Unix or Xenix operating systems. This creates a multi-user environment analogous to the VAX-based systems in which users access the LIMS database via terminals connected to a single host computer. It is a full-function LIMS system which, like SAM, is designed for small-to-medium sized laboratories with less than 30 workstations. It has good capabilities for custom-tailoring to specific data handling situations, reporting formats, and data base query needs. Since all of the communication with the host is through RS-232 serial lines, there is potential for direct transfer of information to the system from intelligent instruments or other laboratory computers.

An essential difference in approach between SAM and NWA LIMS is the hardware environment used to create multi-user access to the LIMS database. SAM provides this within the context of the "electronic office" capabilities of a local area network. This implies the availability of electronic mail and other features for communications between system users typical of most networks. NWA LIMS, by virtue of the multi-user environment created by the Unix operating system, sacrifices these communication aspects, but provides more efficient multi-user access to the LIMS database and reduces the system hardware cost to less than half the cost of installing a local area network.

We found selecting a LIMS system is a complex decision process which, when finally implemented, will cause some fundamental changes in the way our laboratory operates. Through an evaluation process some 18 months long, we concluded every laboratory is unique and there is probably no single LIMS product that is best for everybody. NWA LIMS appears to be best suited to our operating requirements. It is also the lowest priced full-function LIMS system we found. The experience has given us a far better understanding of our own information handling needs, which has led to improvements even before the implementation of the full LIMS system.
COPYRIGHT 1989 Chemical Institute of Canada
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Laboratory Information Management System
Author:Neaves, William M.
Publication:Canadian Chemical News
Date:May 1, 1989
Previous Article:T-Cubed scientific word processing system.
Next Article:Ia FT-Raman really worth it?

Related Articles
EasyLIMS - laboratory data management.
Cimarron Software Signs License Agreement With Affymetrix
AACC Conference Attracts Analytical Instrument Vendors.
NuGenesis(R) Technologies and LabWare, Inc. Announce Effort To Join Scientific Data Management System (SDMS) and LIMS For Enhanced Laboratory...
Strong Growth Prospects for the Rapidly Evolving U.S. Laboratory Information Management Systems Market.
Laboratories eye new LIMS opportunities.
Integration of LIMS.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters