"';;•••·•
• . . • , -
•
This is the second in a series of articles addressing the issues of information technology (IT) within the analytical laboratory. In the first article, we considered some of the operational, logistic, and strategic issues that must be addressed to enable an analytical laboratory to meet current and future demands (/). Here R. D. McDowall of the University of Surrey concentrates on laboratory information management systems (LIMS) and considers the intent and implementation of LIMS in order to stimulate discussion and debate. He also focuses on the impact that a LIMS should—but rarely does—make on both a laboratory and an organization. To overcome this problem, McDowall proposes a matrix for the development of a strategic LIMS. It can be used to aid system implementation or, alternatively, to evaluate the effectiveness of existing applications and chart their further development. The overall purpose is to give managers and analytical chemists a tool they can use to ensure that any LIMS reaches its full potential. Raymond E. Dessy Series Coordinator
R. D. McDowall 1 Department of Chemistry University of Surrey Guildford, Surrey GU2 5HX United Kingdom
The major function of most analytical l a b o r a t o r i e s is to c r e a t e a n d present information in a timely manner t h a t will allow clients to make decisions. A LIMS is one of the prim a r y a u t o m a t i o n tools at t h e disposal of analytical chemists to help them achieve this goal (2). Although a LIMS does not perform analyses, it can be pivotal in integrating both the laboratory operations and the laboratory itself within an efficient organization. It can provide the means to automate the processes of information creation and presentation and, because it can also be the platform for information dissemination to clients and senior management, a LIMS should serve as a s t r a t e g i c s y s t e m for a n y l a b o r a t o r y to add value to the information that is generated. However, t h i s is not always t h e case. A large number of systems fail to meet i n i t i a l e x p e c t a t i o n s , most probably because LIMS are not fully understood (3). To address this prob-
896 A • ANALYTICAL CHEMISTRY, VOL. 65, NO. 20, OCTOBER 15, 1993
lem, in 1990 we proposed a LIMS model (3, 4) t h a t made it possible to conceptually visualize t h e requirements of a system (4, 5). It has been adopted, modified, and used as the basis of the LIMS concept model in the ASTM LIMS guide t h a t has r e cently been finalized (6). However, t h e original model focuses on t h e user functions within the laboratory environment and not on the strategic placement of a system. In this article, we will review the need for strategic IT planning, discuss the limitations of current LIMS and their implementations, and describe how LIMS fit into t h e p r o cesses of l a b o r a t o r y m a n a g e m e n t and decision making. In addition, we propose a m a t r i x for t h e developm e n t of a s t r a t e g i c LIMS t h a t is formed by plotting three types of information systems versus the scope of l a b o r a t o r y a n d o r g a n i z a t i o n a l t a s k s t h a t can be u n d e r t a k e n . A strategic LIMS will facilitate efficient organizations rather t h a n j u s t t h e d e v e l o p m e n t of o p e r a t i o n a l LIMS. Strategic IT planning Strategic IT planning is essential to support business objectives, and yet a dichotomy exists (7). Many current IT investments have failed to deliver t h e a n t i c i p a t e d b u s i n e s s benefits, 1 Mailing address: 73 M u r r a y Ave., Bromley, Kent BRI 3DJ, United Kingdom
0003 - 2700/93/0365 -896A/$04.00/0 © 1993 American Chemical Society
A Matrix for the Development of a Strategic Laboratory Information Management System and some organizations have become concerned about this problem. At the same time, the application of IT by some organizations h a s provided a competitive advantage. The lack of a coherent IT strategy can produce problems such as missing business opportunities, duplicating efforts, adopting system priorities t h a t are not based on business n e e d s , s e l e c t i n g i n c o m p a t i b l e options, and spending considerable resources in an a t t e m p t to i n t e g r a t e systems retrospectively. These difficulties illustrate the need to plan IT strategically (2, 7, 8). Most computer applications tend to be planned from the bottom up to solve local needs r a t h e r t h a n to deliver the information needed to run the organization. There is an overwhelming tendency to develop solutions in an ad hoc manner with little coordination among systems and little or no sharing of information. To demonstrate the problems t h a t such an approach creates, imagine t h e impact of an a u t o m a t i o n or IT project as a pebble t h r o w n into a pond (9). The pebble (project) creates ripples that emanate throughout the pond (organization). The pebble's impact is proportional to its size and the m a n n e r in which it was i n t r o duced to the pond. When more t h a n one pebble is thrown, t h e i r ripples interact; some reinforce each other a n d o t h e r s c a n c e l e a c h o t h e r or change the shape of the pond itself. Ad hoc IT solutions rarely interact; they become islands of automation or information. In describing technical center automation, Dessy has stated t h a t
work flows in m o s t o r g a n i z a t i o n s need feedback loops to make better use of the h u m a n , physical, and informational resources (10). The primary way to achieve feedback is to integrate IT completely, which may necessitate work flow reengineering. Formulating an IT/laboratory automation strategy is a creative process (2, 8). A s s e s s m e n t s m u s t be made about the c u r r e n t s t a t u s and future goals of an organization and decisions must be made about which methods are best to reach objectives, given the necessary resources and alternative routes. The history of IT within an organization will influence its potential for the future, for better or worse. Whether the strategy will constrain or facilitate development
i n f o r m a t i o n or t h e l a b o r a t o r y , a s Dessy h a s i n d i c a t e d (1, 10). T h e y support only r u d i m e n t a r y m a n a g e m e n t a n d p l a n n i n g t a s k s , such as m a k i n g simple calculations of how m a n y samples of a p a r t i c u l a r type were assayed in a month or updating information about workload s t a t u s and sample backlogs. These are not the only tasks facing laboratory m a n a g e r s . They are also concerned with problems such as organizing future work schedules, planning how to use staff effectively, and justifying resources. Determining future demand for services can be difficult. The ability to extrapolate d a t a or calculate how future workloads will affect resources is v i r t u ally nonexistent with present LIMS.
A/C INTERFACE must be considered. A key aspect of any strategy is to build on strengths and overcome weaknesses. The business situation, the envir o n m e n t in which the organization exists, the competitive pressures, and the company's long-term strategic plan m u s t be understood to enable the IT to focus on critical areas for the future. The IT strategy should also enhance the business strategy. T h e r e s u l t i n g d e m a n d s become a portfolio of system requirements that will evolve over time (2, 8). The " o l d " and " n e w " LIMS
C u r r e n t LIMS are actually laboratory d a t a m a n a g e m e n t systems because they manage data rather than
How c a n t h i s p r o b l e m "be overcome? The solution depends on the attitudes of the LIMS suppliers and c u s t o m e r s . If t h e s u p p l i e r s do not perceive the need for these functions they will not be added to the repertoire of LIMS. Thus it behooves the customers to demand these capabilities from their suppliers. The difficulty is t h a t t h e c u s t o m e r is concerned about managing data and not the laboratory. A shift in attitudes is essential. To demonstrate a true LIMS, one might ask (11), "What resources do I need to accommodate 3000 samples for the determination of X t h a t will be submitted during the next three m o n t h s ? " The LIMS s h o u l d be
ANALYTICAL CHEMISTRY, VOL. 65, NO. 20, OCTOBER 15, 1993 • 897 A
A/C
INTERFACE
able to calculate an answer for the manager, basing the decision on the known workload and the c u r r e n t throughput of the assay in question. Priorities would also play a p a r t in the process, and several scenarios would be given. Managers would be able to plan effectively by using this information {11). Other desirable capabilities of the new LIMS include project m a n a g e ment and artificial intelligence tools (3) as well as the ability to provide combined operational, logistic, and strategic tools for the analytical lab oratory (1). LIMS and laboratory management Most LIMS are used for speeding up t a s k s w i t h i n t h e l a b o r a t o r y ; often the use of LIMS is justified on the basis of the ever popular "improve ments in laboratory m a n a g e m e n t . " This is a fallacy that is often sold by naive sponsors of LIMS to naive se nior management. As more seasoned individuals can attest, m a n a g i n g a laboratory is a delicate balancing act involving four main factors: h u m a n resources, analytical equipment, an alytical methods, and workload (12). Each area can be divided further into a number of individual items. The four areas are interrelated; a change in one can have drastic im pacts in one or more of the others. For e x a m p l e , a n i n c r e a s e in staff turnover may slow the t u r n a r o u n d time of samples, increase the overall reporting times, and affect the abil
ity to use available equipment fully. In contrast, automation may have ef fects on s a m p l e t h r o u g h p u t a n d backlog, staff m o t i v a t i o n , a n d t h e need to train or recruit personnel to carry out the new work. This is a dy namic s i t u a t i o n t h a t needs careful and continuous management. D e c i s i o n m a k i n g . Making deci sions is a complex process t h a t in volves i n f o r m a t i o n g a t h e r i n g a n d sifting before t h e decision can be made. Because it is unusual to have a complete picture of a situation, an element of uncertainty accompanies a decision. How l a r g e t h i s u n c e r tainty is depends on the implications of the decision as well as the manager's experience and his or her willingness to take risks. F o u r l e v e l s of i n f o r m a t i o n or knowledge that affect decision mak ing can be distinguished (13). Level 1 is a data set with unknown relation ships between the data; Level 2 is a reduction of data into characteristic figures (e.g., m e a n value, s t a n d a r d deviation, or correlation coefficient); Level 3 is a descriptive model that il l u s t r a t e s the relationships between the characteristic figures or blocks of figures; and Level 4 is a predictive model that forecasts the effect of de cisions on the value of the character istic figures. Levels 3 and 4 are t h e most a d vanced tools for decision making be cause they provide users with direct information on the consequences of a l t e r n a t i v e decisions. Most labora
tory management decisions are based on future trends or workloads; there fore, a predictive model t h a t deter mines how future workloads would affect r e s o u r c e s is n e e d e d . S u c h functions are virtually nonexistent in present LIMS. Laboratory m o d e l i n g . Modeling the laboratory can help predict the result of changes in events. This pro cess should be linked to t h e LIMS database to access the needed infor mation. Some laboratory models that are based on digital filtering, a tech nique of operations research that can represent analytical procedures and i n s t r u m e n t s , h a v e been published (14). However, they are relatively ru d i m e n t a r y and c o n c e n t r a t e on the low operational levels of laboratories. An a l t e r n a t i v e would be to collect and store operations research data in a LIMS and t h e n formulate and as sess different options with an expert system or a suitable model. K l a e s s e n s a n d c o - w o r k e r s (15) have collated historical d a t a stored in a LIMS into a list and, using an algorithm based on fuzzy set theory to incorporate u n c e r t a i n t y into the process, classified the entries into a l i m i t e d n u m b e r of c l u s t e r s called sample types. This classification is completely user-defined and t r a n s forms the historical data into a form that can be used to study the perfor mance of a laboratory. Retrospective analysis of the workload (e.g., client/ source, t e s t s c a r r i e d out, or batch size and throughput) is possible by
Organization
Laboratory
Strategic LIMS
Integrate with client operations
Compare requests with capacity
Resources based on workload
Integrate with document management software
Decision support system integration
EDI and CANDAs Rapid commercialization
Logistic LIMS
Integrate laboratory operations
Automate regulatory compliance
Monitor test use
Electronic reports to clients
Highlight out-of-specifica tion results
Remote on-line inquiry integration
Automate existing operations
Monitor and approve test results
Sample status Billing
Paper reports to clients
Display results vs. specifications
Remote printing of results E-mail between customer and laboratory
Laboratory operations
Monitor and control
Laboratory management
Reporting and communication
Analytical decision making
Organizational integration
I φ •3
α. Ε ο ϋ
Operational LIMS
System scope
Figure 1. LIMS matrix and functions associated with each cell. EDI, electronic data interchange; CANDAs, computer-assisted new drug applications. 898 A • ANALYTICAL CHEMISTRY, VOL. 65, NO. 20, OCTOBER 15, 1993
interrogating the LIMS database. The resulting information is presented in histograms. Although a representation of historical data is possible by using this technique, the selection of the path leading to the optimal r e p r e s e n t a t i o n must be found by trial and error, and hence the need for a user-definable classification is clear. Klaessens extended this work to include a simulation model of an analytical laboratory in which samples, analysts, and instruments are represented as a production line (16). A sample generator represents the flow of samples into the laboratory, and a planner schedules the samples for analysis. SIMULA, a dedicated simulation language, is used to construct an expert system with which to implement rules concerning the flow of data and samples in a laboratory as well as to produce predicative information. Van de Wijdeven and co-workers have used digital simulation to model an analytical laboratory with information derived from laboratory records and the experience of laboratory staff (17). They monitored the flow of samples through the laboratory and the analyses performed there. By changing various factors in the model, the simulation was able to identify bottlenecks and improve scheduling, thus increasing the work capacity of the laboratory. Constructing a simulation model is both time-consuming and expensive, and it must be weighed against a decision to increase the work capacity without the benefit of the model. Developing simulation models will give managers additional information from which better decisions can be made. Quo v a d i s LIMS? As shown above, some of the requirements of the new LIMS are proper laboratory management, prediction capability, and decision support. However, the current generation of systems cannot undertake these functions. To build a LIMS that serves both the laboratory and the organization requires strategic thinking that most analysts lack. Therefore, we propose a LIMS matrix as a tool to help users, impie menters, and vendors install effective systems. The LIMS matrix
The rectangular matrix presented in Figure 1 is simply a plot with information systems on the vertical axis and laboratory and organizational tasks on the horizontal axis. The components of the matrix and details of
its construction are explained below. Classification of information s y s t e m s . Computer applications have been divided into three hierarchical types (data processing, control, and planning systems) by Anthony (18). Each type is responsible for different tasks within an organization. This concept was modified by Nolan (19), and the three types of systems are now called operational control,
management control, and strategic planning. This concept can be applied to LIMS as well to produce three types of systems: operational, logistic, and strategic. The functions and aims of these systems are described below in more detail and summarized in the box below. A strategic LIMS integrates information and applications from different functional areas. From this infor-
Impact of the different types of LIMS on laboratories and organizations Strategic LIMS Overall impact is on laboratory competitiveness; it has the greatest impact on the business operation Integrates information and applications from different functional areas and reshapes operations Logistic LIMS Overall impact is on laboratory effectiveness; it has intermediate impact on the business operation Optimizes and enhances operations by providing information to manage and control the functional area Operational LIMS Overall impact is on laboratory efficiency; it has the greatest impact on laboratory operations and little impact on the business operation Improves operational efficiency by automating laboratory processes and is typified by automating the status quo
Scope of LIMS functional areas Laboratory operations Overall scope is to automate and structure work Automate basic laboratory operations such as sample entry, work list generation, and results entry Rationalize the work performed Monitor and control Overall scope is to evaluate performance Monitor and control laboratory operations by processes such as approving results, using quality control schemes, and checking for transcription errors Laboratory management Overall scope is to support intellectual processes Organize and manage laboratory functions and operations Plan projects and work Reporting and communication Overall scope is to augment human communication Provide the means to transmit results or reports and communicate with laboratory clients Analytical decision making Overall scope is to aid and speed decision making Provide high-quality information in a timely manner and in the correct format needed to make decisions Support processes in production, development, or research Organization integration Overall scope is to facilitate transactions within and between organizations Integrate with other functional groups in the corporation and between organizations (e.g., electronic data interchange or computerassisted new drug applications)
ANALYTICAL CHEMISTRY, VOL. 65, NO. 20, OCTOBER 15, 1993 • 899 A
A/C I N T E R F A C E mation it may be possible to reshape operations. A strategic LIMS has the greatest impact on the company by increasing the competitiveness of the laboratory. The m a i n function of a logistic LIMS is to provide users, especially m a n a g e r s , w i t h both m a n a g e m e n t and quality control information and to increase the effectiveness of a lab oratory. An operational LIMS a u t o m a t e s basic processes within the laboratory and can be considered to be an oper ational control system (19). The im pact of a L I M S , as it is n o r m a l l y implemented, is usually local and in creases the efficiency of a laboratory. An o p e r a t i o n a l LIMS would a u t o mate functions such as sample entry, work list g e n e r a t i o n , a n d r e p o r t preparation. S y s t e m s c o p e . Six a r e a s of sys tem scope comprise t h e horizontal axis: laboratory operations, monitor ing and control of operations, labora tory m a n a g e m e n t , r e p o r t i n g a n d communication, analytical decision making, and organizational integra tion. This system scope is derived from the work of Markus (20). Ana lytical decision m a k i n g h a s b e e n added to the five areas she described to make the matrix pertinent to the analytical laboratory. Each area of scope is d e s c r i b e d in t h e box on p. 899 A with an overview of its func tion. Matrix c o n s t r u c t i o n . The matrix together with t h e associated func tions of each cell can be viewed as the 3 x 6 block illustrated in Figure 1. The a r e a s of l a b o r a t o r y o p e r a tions, monitoring and control of op erations, and laboratory manage ment on t h e system scope axis are concerned w i t h functions inside a laboratory. Reporting and communi cations, analytical decision making, and organizational i n t e g r a t i o n are concerned with organizational func tions. A LIMS may be considered a localized tool if it is i m p l e m e n t e d only in t h e l a b o r a t o r y a r e a of t h e matrix. When a LIMS extends be yond t h i s division, t h e l a b o r a t o r y (via the LIMS) can have a greater ef fect on the organization. The third d i m e n s i o n of the ma trix. Although the matrix is shown here in two dimensions, it is possible to add a third dimension to resolve a problem t h a t often troubles organi zations: prioritizing automation and IT projects (21). The third dimension could involve any of m a n y diverse b u s i n e s s p r e s s u r e s such as r a p i d identification of potential products, faster development t i m e , or lower
production costs. These factors can be customized for an industry or an individual company.
developed to their full potential, ei ther as a result of a lack of personnel or limited management vision. The axes of the m a t r i x represent the effect of the system on the organ ization (computer system axis) ver sus the effect of the laboratory on the organization (system scope axis). The further toward the end of each axis t h a t one aims the development, the greater the strategic focus of the sys tem. However, as a direct corollary, the further one develops the system
Impact of the matrix on systems development The matrix is a tool for visualizing a LIMS and its consequences. Not all systems need to be developed to the full potential of t h i s m a t r i x . Some systems need to interface with other applications t h a t perform required matrix functions. Most LIMS are not
(a) Document management
Strategic LIMS
Logistic LIMS
Operational LIMS
LIMS—Phase I LO
l
MC
ι
LM
ι
RC
ι
AD
l
01
(b) Document management
Strategic LIMS
Logistic LIMS
Phase II
Phase II
Operational LIMS LO
I
MC
I
LM
I
RC
I
AD
'
01
(o) Document management
Strategic LIMS
Logistic LIMS
Phase I
Operational LIMS
LIMS
TO
I
MC
Phase 111
!
LM
RC
!
AD
ι
01
Figure 2. Overall scope of the design for a LIMS in a pharmaceutical company. (a) First phase of development, (b) Second phase of development, (c) Third phase of development. LO, laboratory operations; MC, monitor and control; LM, laboratory management; RC, reporting and communication; AD, analytical decision making; 01, organizational integration.
900 A • ANALYTICAL CHEMISTRY, VOL. 65, NO. 20, OCTOBER 15, 1993
along each axis, the greater the risk attached to the overall development. S t r a t e g i c LIMS i m p l e m e n t a t i o n . P r u d e n t implementation of a strategic LIMS should be performed in a s t a g e d s e q u e n c e a n d be well planned and funded (8). It is important to realize that a LIMS cannot be implemented w i t h o u t the involvement of the laboratory t h a t it is intended to a u t o m a t e . Therefore, the foundation of t h e system m u s t always be based on implementing the LIMS in t h e l a b o r a t o r y first, a n d t h e n e x p a n d i n g t h e s y s t e m scope into the organization. The 10-yearold s t a t e m e n t t h a t t h e essence of LIMS d e v e l o p m e n t is "to provide simple services first with more sophisticated ones following later " (22) still holds true and always will. The overall scope for a LIMS in a pharmaceutical company is shown in Figure 2. The LIMS must interface with a company-wide document m a n a g e m e n t application t h a t will prepare computer-assisted new drug applications (CANDAs). The LIMS will deliver analytical reports to this software. The first phase of development is to establish the LIMS in the analytical laboratory (Figure 2a). Reports in the first instance will be on paper. The second phase is to build the interface to the document management application and integrate the analytical i n s t r u m e n t s w i t h i n the LIMS environment (Figure 2b). The latter t a s k will enable a u t o m a t i c downloading of sample files and methods coupled with instrument control via feedback loops and uploading of analytical r e s u l t s into t h e LIMS. The third phase is to communicate electronically w i t h c l i e n t s , which i n cludes having the ability to display results and print reports in remote locations (Figure 2c). T h r e e LIMS g e n e r a t i o n s . The matrix can also be used to evaluate vendor products. For example, vendor publications have claimed the release of t h i r d - , f o u r t h - , or s i x t h generation products, but how does one define the degree of change that constitutes a new generation? The matrix can be used to define a new LIMS generation by considering the three types of LIMS on the vertical axis (operational, logistic, and strategic) as j u s t t h r e e g e n e r a t i o n s of LIMS. For a LIMS to qualify as a higher order, it must possess at least three functions of the higher system category. The majority of functions incorpor a t e d in a c o m m e r c i a l L I M S a r e driven by existing and potential us-
e r s . Because u s e r s c o n c e n t r a t e on laboratory functions r a t h e r t h a n on o r g a n i z a t i o n a l f u n c t i o n s , can t h e vendors be blamed for responding to market forces? This leads to the LIMS catch-22 situation. Laboratory users and their clients should realize that the organization requires delivery of information to run the business. This necessitates systems that operate conjointly with the objectives of the organization. These systems should provide a means to capture, store, and manipulate a wide range of data to produce information t h a t can be delivered in the places it is needed, in the format t h a t it is n e e d e d , a n d w h e n it is needed. However, to achieve this objective, it is essential that systems be developed with a solid foundation. This foundation is the analytical laboratory. Impact of the LIMS matrix on the LIMS model It is n a t u r a l to compare the roles of the development matrix with the LIMS model (3, 4). When conceived, the LIMS model was intended to describe and classify the main user and s y s t e m f u n c t i o n s . By i t s n a t u r e , however, t h e model is u s e r - b a s e d and concentrates mainly on the laboratory r a t h e r t h a n on the organization. The model is concerned with reporting methods but barely considers electronic mail or desktop publishing. In contrast, the LIMS matrix is a tool that has been conceived strategically. It is more b u s i n e s s - o r i e n t e d and encompassing t h a n the model. Therefore it is envisioned t h a t the matrix should be used first to chart the overall scope of a LIMS within an organization, defining the interaction with other IT systems such as process control systems (23). Once the breadth of the system has been defined, the LIMS model can be used to determine the scope of the functional components within the laboratory environment. I would like to t h a n k Ray Dessy, Rich Mahaffey, and Gerst Gibbon for helpful advice, encouragement, and comments during the preparation of this article.
References (1) Dessy, R. E. Anal. Chem. 1993, 65, 802 A. (2) McDowall, R. D. Chemom. Intell. Lab. Syst. Lab. Info. Mgt. 1992, 17, 265. (3) McDowall, R. D.; Mattes, D. C. Anal. Chem. 1990, 62, 1069 A. (4) Mattes, D. C; McDowall, R. D. In Scientific Computing and Automation (Europe); Karjalainen, E. J., Ed.; Elsevier: Amsterdam, 1990; p. 301.
(5) McDowall, R. D. Chemom. Intell. Lab. Syst. Lab. Info. Mgt. 1992, 17, 181. (6) Standard Guide for Laboratory Information Management Systems (LIMS) E5178; ASTM: Philadelphia, PA, 1993. (7) Ward, J.; Griffiths, P.; Whitmore, P. Strategic Planning for Information Systems; John Wiley and Sons: Chichester, England, 1990. (8) Mole, D.; Mason, R. J.; McDowall, R. O.J. Pharm. Biomed. Anal. 1993, 11, 183. (9) Brooks, M. Chemom. Intell. Lab. Syst. Lab. Info. Mgt. 1991, 13, 51. (10) Dessy, R. E. Anal. Chem. 1992, 64, 733 A. (11) Laboratory Information Management Systems: Concepts, Integration and Implementation; McDowall, R. D., Ed.; Sigma Press: Wilmslow, England, 1988; pp. 47-48. (12) McDowall, R. D.; Leavens, W. J.; Massart, D. L. Chemom. Intell. Lab. Syst. Lab. Info. Mgt. 1992, 13, 221. (13) Massart, D. L.; Vandeginste, B.G.M.; Deming, S. N.; Michotte, Y.; Kaufman, L. Chemometrics: A Textbook; Elsevier: Amsterdam, 1988. (14) Cook, C. F. Anal. Chem. 1976, 48, 724 A. (15) Klaessens, J.; van Schalkwijk, P.; Cox, R.; Bezemer, B.; Vandeginste, B.; Kateman, G. Chemometrics 1988, 3, 81. (16) Klaessens, J. Ph.D Thesis, University of Nijmegen, The Netherlands, 1990. (17) Van de Wijdeven, B.; Lakeman, J.; Klaessens, J.; Vandeginste, B.; Kateman, G. Anal. Chim. Acta 1986, 184, 151. (18) Anthony, R. N. Planning and Control Systems: A Framework for Analysis; Harvard University Press: Boston, 1965. (19) Nolan, R. L. Harvard Business Review 1979, March-April, 115. (20) Markus, M. L. Systems in Organizations; Pitman: London, 1984. (21) Dessy, R. E., Virginia Polytechnic Institute and State University, personal communication. (22) Dessy, R. E. Anal. Chem. 1983, 55, 70 A. (23) Stein, R. R. Process Control and Quality 1990, ;, 3.
R. D. McDowall is a bioanalytical chemist with more than 20years experience, 15 of which have been spent working in the pharmaceutical industry. He is a visiting senior lecturer in the Department of Chemistry at the University of Surrey where he is working on a computer-aided chemistry course. He is editor of "Laboratory Information Management," a section of Chemometrics and Intelligent Laboratory Systems, and consulting editor of LC/GC International. He is also principal ofMcDowall Consulting.
ANALYTICAL CHEMISTRY, VOL. 65, NO. 20, OCTOBER 15, 1993 • 901 A
