1-5.Standardization of laboratory automation systems:information and communication

The structural components of a total clinical laboratory automation system (TLA) are a laboratory information system (LIS), a sample transportation system (STS) and analyzers; information is exchanged among these components. In smaller laboratories, the LIS and the analyzers are generally linked, while in medium-sized to large laboratories, a laboratory system can be efficiently operated through the use of an STS. An STS is a system that carries specimens to analyzers and automatically activates analyzers. In recent years, STSs have been programmed to handle specimens with more "intelligence". A laboratory automation system (LAS) is a system in which analyzers and a STS are linked.
Through the use of STSs or robots, LASs can deliver specimens according to a plan and optimally allocate specimens depending on factors such as the status of each analyzer or the need for reruns. To achieve this type of integrated control, it is necessary to manage information such as requests and test results controlled by an LIS, the state of tests controlled by an LAS, and reagents and quality control of analyzers, so that appropriate action can be taken. Therefore, information must be exchanged among the LIS, LAS and analyzers. At present, since the specifications for communication among these components differ among manufacturers, it takes a great deal of time and money to link them as part of an effective system.
This article will discuss how these components are used to construct a system and how information is exchanged among these components, and report the necessary points in achieving the standardization of communication.

Regarding the standardization of information and communication on a global scale, the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC) began their investigation by forming a committee on SC16 (standardization of open system interconnection) within TC97 (standardization of electronic calculators and information processing) in 1977. Later, a joint committee was held between TC83 (information instruments) and SC47B (microprocessors), and Open System Interconnection (OSI) was designed.
As shown in figure 1, OSI consists of seven layers: the physical, data link, network, transport, session, presentation, and application layers.

As the standardization of information and communication in Japanese clinical laboratories, the Japan Society of Clinical Laboratory Automation established "Bit Serial Interface Standard Guideline" (BSI) ^1)-4) in 1981. BSI regulates the physical and apart of data link layers of OSI. Meanwhile, the American Society of Testing and Materials (ASTM) established E-1381-915) and E-1394-916). E-1381-91 is a low level protocol, and E-1394-91 is a high level protocol that can exchange field information such as patient attributes, requests, and test results. When compared with BSI, ASTM is superior in terms of reliability and flexibility. However, due to the fear of problems with real-time transfer, often associated with ASTM in large analyzers, ASTM has not been widely utilized in Japan.
Then in 1996, the Japanese Association of Healthcare Information Systems Industry (JAHIS), the Japan Analyzers Industry Association, and the working group of the Committee on Analytical Systems of the Japan Society of Clinical Chemistry (JSCC) began establishing communication specifications.
As to the standardization of the communication between LAS and analyzers, the Laboratory at Kochi Medical School and A&T cooperatively began a study on open specification transportation systems and proposed parallel and serial transfer specifications in 19957,8). Furthermore, in 1997, JSCC-CAS proposed a serial interface specification plan9-12).
A working group was formed, also in 1997, within the Japan Council for Clinical Laboratory Standardization (JCCLS) to investigate the specification plan of various organizations, gather opinions and establish Japanese specifications.

In smaller laboratories, the LIS and analyzer are generally linked using an RS232C cable in order to select the necessary tests and collect the test results. In this type of system, medical technologists set specimens in the analyzer and perform batch tests (Figure 2).

As LISs and analyzers have been used together for a long time, there are many established procedures for linking them. Therefore, there have been many studies on methods that can link both units in a short period of time by separating and compartmentalizing processing layers inside LISs that correspond to the OSI layer protocol model13-16).

The components of another type of laboratory system are an LIS, analyzers, and an STS. As shown in figure 3, this system forms a triangle composition among the three components. In this system, once specimens are delivered to the inlet of the STS, the system will read the ID of each specimen and deliver each specimen to a pre-designated analyzer, so that the analyzer can automatically perform the necessary tests and send the test results. The communication procedures between the LIS and analyzers are the same as those shown in figure 2, and in this system, the LIS is linked to the STS and the STS is linked to analyzers.
Nearly the same request information that is exchanged between the LIS and analyzers is exchanged between the LIS and STS, so the same communication procedure can be used.

Since commercially available analyzers were designed to be integrated with STSs, communication between STSs and analyzers was unnecessary. Nevertheless, many began to see the benefit of separating analyzers and STSs and standardizing the communication procedure between them so that analyzers can be connected to a STS regardless of its physical configuration (either rack or holder method). Consequently, the standardization of mechanical and communication protocol between STSs and analyzers has been extensively investigated.
This communication interface controls the timing of specimen ID reading and specimen sampling, and this function can be executed from a remote location. The specifications for the timing of specimen sampling must be rigidly regulated with respect to time.

STSs deliver specimens to analyzers, as well as send and receive information such as the synchronization of sampling and the status of tests. Furthermore, in order for an STS to control specimens with greater efficiency, it became necessary for STSs to handle requests and test results. In addition, the demand for a system that can plan and perform optimal tests by assessing information such as residual amounts of reagents or analysis times continued to grow. The system that performs these functions is the LAS.
With the system configuration shown in figure 3, there is a lot of necessary information transfer, and the communication between the LIS, LAS and analyzers is complicated. Furthermore, the distribution of the responsibilities of LIS and LAS also becomes excessively complicated in some cases. Figure 4 shows an integrated system that reorganizes all components more efficiently.

If STSs could be controlled by utilizing data managed by LISs (e.g., test requests and results), then many useful applications can be constructed. For example, as shown in figure 5, when STAT specimens arrive, analyzers may be in sleep mode or performing maintenance procedures (e.g., cell rinsing). Nonetheless, since a certain amount of time is usually required before a test is performed (e.g., centrifugation or transportation), regular maintenance such as reaction solution exchange, calibration and warming-up can be temporarily suspended. In other words, since the LIS controls the flow of specimens, if an LAS can regulate this flow, it can send commands to have the necessary analyzers to be ready for tests from a remote control. Subsequently, when specimens arrive, the necessary tests can be performed promptly.

In most existing systems, reruns are performed inside analyzers. Nonetheless, when this happens, specimens must be retained by the analyzers until the results of tests are obtained.

This problem can be resolved by letting an LAS have total management of automatic rerun17). Furthermore, by performing automatic diagnosis based on test results of the other analyzers with the help of a diagnosis support system, when technical errors are suspected or when additional tests are necessary, automatic reruns can be automatically performed. Figure 6 shows the overview of a system configuration that can handle this type of rerun. In this system, it appears that one giant auto analyzer is connected to the LIS.

When considering the standardization of international communication procedures, it is important to consider what specifications are required in this field, and the future direction of LASs.
For example, information transfer synchronized with specimen sampling is necessary between an LAS and analyzers. At present in 1997, information transfer is carried out in four-to five-second cycles in the fastest analyzers, so it is necessary to have the next specimen ready and send relevant information such as specimen ID, dilution and requests during this time. Furthermore, by utilizing this communication path, test results, remote control and quality control must be completed. However, this is impossible with the current communication specifications used for clinical laboratories. In addition, even though information transfer is achieved via LAN or serial bus, the traffic caused by other analyzers must be taken into account. Therefore, more time-responsive communication protocols are needed.
Furthermore, to achieve international standardization, the Double Bytes Code System (DBCS) that is most frequently used in Asia, such as Kanji codes, should be investigated. In addition, it will be necessary to standardize the transmission format for waveforms or image information such as protein electrophoresis or reaction curves.

Trends in the current standardization of information and communication in clinical laboratories and the inter-system topology of LISs, LASs and analyzers were the mainly topics of this article. If these components were to be compared to body parts, then LIS would be the brain, and LAS would constitute organs such as the eyes, hands and legs that enable intelligent functioning of the system. Information and communication systemically manages these organs, so as long as communication procedures are not standardized, the hands and the legs can not function in unison. Therefore, the standardization of communication procedures is important, and the standardization of the specifications at the application level is especially important.
The NCCLS established five working groups to draft a set of international standards. One of the five groups is investigating communication specifications between LASs and analyzers. Once specifications are established, it will be possible to connect analyzers more easily at lower costs, thus making it possible to construct a large LAS from small analyzers to suit the needs of every laboratory. When this system is widely installed, medical technologists can concentrate on laboratory services that they do not have time to perform now, explore laboratory operations that can not be automated, and spend more time assisting diagnoses or developing knowledge of LASs.
Rather than discuss the technical aspect of communication specifications in detail, this article presents an overview of and summarizer trends in communication specifications. Please refer to the documents listed in the references for details.

1) Interface committee, Japan Society of Clinical Laboratory Automation: Regarding the standardization of analyzers and computer interfaces. Japan Society of Clinical Laboratory
Automation Magazine, 6 (1): 4-6, 1981.
2) Interface committee, Japan Society of Clinical Laboratory Automation: Bit serial interface standardization plan. Japan Society of Clinical Laboratory Automation Magazine, 6 (1): 7-10,
1981.
3)Interface committee, Japan Society of Clinical Laboratory Automation: Regarding the standardization of analyzers and computer interfaces (part II). Japan Society of Clinical
Laboratory Automation Magazine, 7 (1): 1-4, 1982.
4) Interface committee, Japan Society of Clinical Laboratory Automation: Regarding the standardization of analyzers and computer interfaces (part III). Japan Society of Clinical
Laboratory Automation Magazine, 8 (1): 53-55, 1983.
5) ASTM, E-1381-91: Specification for Low-Level Protocols to Transfer Messages Between Clinical Laboratory Instruments and Computer Systems, 1991.
6) ASTM, E-1394-91: Standard Specification for Transferring Information Between Clinical Laboratory Instruments and Computer Systems, 1991.
7) Kataoka, H., et al.: Proposal for standardizing communication specifications in analyzers matching specimen transportation systems. Japan Society of Clinical Laboratory Automation
Magazine, 21 (2): 154-158, 1996.
8) Kataoka, H.: Transportation systems and CLIS. CLIS PRIMER '95-99 A&T, (4): 3-14, 1995.
9) The Committee on Analytical Systems, Japan Society of Clinical Chemistry: A standard specification plan for the communication between analyzers and transportation systems.
Chapter 1. Summary of specifications. Japan Society of Clinical Chemistry Magazine.Vol.26, Supp.2, 105b-113b, 1997.
10) The Committee on Analytical Systems, Japan Society of Clinical Chemistry: A standard specification plan for the communication between analyzers and transportation systems.
Chapter 2. RS232C serial communication specifications. Japan Society of Clinical Chemistry Magazine.Vol.26, Supp.2, 115b-132b, 1997.
11) The Committee on Analytical Systems, Japan Society of Clinical Chemistry: A standard specification plan for the communication between analyzers and transportation systems.
Chapter 3. Specifications for High-Level Protocol Layers. Japan Society of Clinical Chemistry Magazine. Vol.26, Supp.2, 133b-142b, 1997.
12) The Committee on Analytical Systems, Japan Society of Clinical Chemistry: A standard specification plan for the communication between analyzers and transportation systems.
Chapter 4. Specifications for application protocol. Japan Society of Clinical Chemistry Magazine.Vol.26, Supp.2, 143b-152b,1997.
13) Tani, S., et al.: Development and investigation of analyzer interfaces with a protocol database. Japan Society of Clinical Laboratory Automation Magazine, 18 (1): 36-42, 1993.
14) Kataoka, H.: Auto analyzer online systems employing a client server method, Kochi (ISSN1340-7937), 23 (3): 54-60, 1993.
15) Kataoka, H., et al.: Development of auto analyzer online systems emphasizing machine compatibility in internal general specifications. Japan Society of Clinical Laboratory Automation Magazine, 19 (4): 466, 1994.
16) Kataoka, H., et al.: Reducing the number of procedures for online connection of analyzers, and problems associated with standard specifications. Clinical Pathology, 45 (6): 564-572,1997.
17) Kataoka, H., et al.: Check system for abnormal test results in automatic rerun systems. Japan Society of Clinical Laboratory Automation Magazine, Shikoku-branch 3 (10): 67-75, 1986.