1-3.Standardization of laboratory automation systems:transportation methods and racks

In recent years, the standardization of laboratory automation systems, especially analytical software, has gained momentum throughout the world. This is due to one milestone in this field. In 1956, AutoAnalyzer was introduced by, the American company, Technicon Corporation and it immediately made the automation of clinical chemistry tests advance then that of hematological tests [1]. In order to determine the advantages and disadvantages associated with AutoAnalyzer, young researchers who assumed the leadership in the field of clinical chemistry investigated the analyzer from various angles.

This is not a criticism of Technicon Corporation, but the original idea behind automation that Technicon Corporation developed was to standardize simple procedures and then mechanically process them. Muller, with the help of Andres Ferrari who was the head of Technicon's R&D department, introduced AutoAnalyzer in 1958, and he made the following comments [2]:

"The AutoAnalyzer is a completely automatic system which records the level of concentration of a given component in the test solution against a known concentration of that component in a standard control solution. No gravimetric or volumetric measurement is involved: Rather it is a matter of continuous plotting of ratios (the concentration of the sought material in the unknown against its known concentration ratio in the standard control). For that reason it is never necessary to bring a reaction to completion."
In other words, it is not necessary for coloration to be completed in reaction solutions that reach a colorimeter as long as the state of the reaction for an unknown and that for the standard are equal.

This was based on the following assumptions: 1) development of the color is time dependent, and the rate of reaction is concentration dependent, and 2) even though the compositions of a standard solution and a sample may differ, as long as the concentrations of the component interested are the same, the rate of reaction will also be the same. These assumptions caused many problems later on.

Furthermore, since AutoAnalyzer first employed sample cups and an auto-sampler, the only thing a laboratory technologist needed to do was to place appropriate amounts of a specimen and standards in each sample cup. Up to eight tigon tubes with a different thickness were drawn at the same speed to pump the specimen and reagent. The necessary amount of reagents was determined in the ratio to the specimens. For example, if 0.1 ml of a specimen and 0.5 ml of a reagent were required, two tubes with a diameter ratio of 1 : 5 were used. This was the third assumption: as long as the ratio of a reagent and a target component can be maintained, the stoichiometry can also be maintained. In other words, this assumption assumes that, as long as the ratio of a specimen and a reagent is maintained, the absorbance developed by a 0.1-ml and a 100-ml of total solution would be the same.

In addition, the instrument that was introduced for the measurements of urea and blood sugar [3] adapted a dialysis membrane called dialyzer used from the earliest model. It was also assumed that the dialysis effects of standard solutions and specimens were the same, which further complicated Assumption 1 written above.

Since the objective of the present article is not to argue on these problems, I will not touch them on further. However, it is clear that they later necessitated frequent revisions, as surf of beats the shore, to the various software and hardware that comprise laboratory automation systems.

Standardization was achieved in all aspects of the Type 1 AutoAnalyzer that was introduced in 1956. Repetitive procedures were established by excessively simplifying and standardizing all include sampling cups, sampling mechanisms, reaction processes and output recordings. For another example, AutoAnalyzer was equipped with an oil bath that could regulate the temperature within 0.1 °C (constant temperature bath). Reaction temperatures and heating time could easily be adjusted using a temperature controller with a thermometer applied. The reaction time was easily adjusted too by changing the length of a coil. Therefore, tests could be performed using the same machine at any temperature. In his 1958 report, Muller made an extremely important comment on the temperature regulation of constant temperature baths: he proposed a "reaction temperature at 37°C" that states enzymatic reactions should be performed at 37 ± 0.1°C. The temperature of enzymatic reactions is still under debate today. Oddly enough, this debate was not initiated in the field of academics.

The reaction temperature for the LDH activity measurement that was introduced in 1958 followed Muller's footsteps by setting the temperature at 37.0 ± 0.05°C [4]. However, when the 14th American Association of Clinical Chemists was held at Santa Monica in 1962, the focus of the automated method for the measurement of transaminase activity was clearly on efficiency of the laboratory performance [5]. You can say, even though enzymatic activities were measured at 25°C at that time, since enzymatic reactions were slow at this temperature, the efficiency of transaminase analysis was low. As a result, since it was determined that the quantitative relationship was maintained at 50°C, the reaction temperature could be raised. Pyruvic acid and oxalacetic acid were used as the standards for that analysis. To convert test results that were obtained at 50°C to those obtained at 25°C, a coefficient of 2.5, that was experimentally obtained, was used. This is the beginning of the concept, "You can do laboratory tests by an automated analyzer, if you want to finish it fastest".

The problem of analytic temperature is still a mystery; Leonard T. Skeggs, who invented the AutoAnalyzer, and Edwin S. Whitehead, who founded Technicon Corporation have passed away, and since Technicon has been taken over, the bulk of old experimental data has been missing. The reason why I want to find the old data is that the measurement temperature was determined by the AutoAnalyzer itself. Nonetheless, it is clear that Type I AutoAnalyzer was equipped with an oil bath using diethylene glycol in it, and the recommended temperature of enzymatic reactions at was 37.0°C at that moment. However, as the proper enzymatic reaction temperature was debated from various angles, the situation become increasingly complicated. This point will be mentioned briefly later.

In 1963, an experimental 8-channel multiple AutoAnalyzer was developed and introduced at the 5th International Congress on Clinical Chemistry in Detroit [6,7]. This machine is important in the history of auto analyzers in that it was the first machine to perform multiple analyses for one specimen. In the instrument, analog data was incorporated by multiple channels and recorded chronologically on a chart. Since AutoAnalyzers were not digitalized in its earlier models, many attachments were invented. The same devices used for the original AutoAnalyzer were used also for the multiple AutoAnalyzer. Technicon constructed multiple AutoAnalyzers by combining analyzers that could function independently to increase test items. This apparatus introduced the concept of simultaneous multiple analyses to the field of auto analyzers.

To make this type of analyzer commercially available, Technicon needed to improve its Type I AutoAnalyzer. In 1970, the Type II AutoAnalyzer was introduced, and the analyzer was also exported to Japan, with the first unit being installed at Kanto Teishin Hospital on September 30, 1971. Soon after, a 12-channel analyzer performed with very small volume of specimen was developed and introduced as SMA12/micro. To return to the measurement temperature issue for a moment, even though the type II also used an oil bath for maintaining the reaction temperature, the bath was changed from a circular shape to a rectangular one, and the oil bath was smaller in the new analyzer. Temperature control was conducted at the base of the oil bath using a patented circuit regulated by a micro chip. The temperature of the oil bath was set at 37.5°C, and it was not adjustable. As mentioned earlier, it is not clear why the oil bath temperature was changed from 37.0°C to 37.5°C. According to a 1974 Technicon brochure, the reaction temperature of transaminase assay was set at 37.5°C. Since the same heating bath was used for other tests, the same reaction temperature was maintained while performing various serum enzymatic activity measurements.

In 1972, Technicon developed SMAC-1 AutoAnalyzer and then began shipping the unit in 1974. The SMAC-1 was first installed at Kanazawa Medical University. The same devices used in the Type II were also used in the SMAC-1, and for the first time, the whole control was computer assisted. The SMAC-1 had 20 channels, and was capable of processing 150 specimens per hour. Again to mention about the reaction temperature one employee told me that the setting of the oil bath of the SMAC-1 was 37.0°C, and the technical data sheet does not explain the 0.5°C decrease. It was funny the same heating method was used in the type II and the SMAC-1, the temperature of the bath was 0.5°C lower in the SMAC-1.

The period from 1963 to 1970 marked the beginning of simultaneous multiple automated analyzers. At this time, clinical laboratories began to demand the mechanization of laboratory operations. After World War II, advanced technology that had been developed by the American military was also applied to clinical laboratories, and the provision of various all kinds of data was enabled. The bulk supply of enzymes such as glucose oxidase changed laboratory analyses. Many laboratories began to receive specimens in numbers that far exceeded their capacity, and demand for increasingly prompt and precise laboratory tests grew further. In 1963, Skeggs explained the origin of multiple auto analyzers as follows,
"Several years ago, a fundamental decision was made. It was decided that it would be far easier and better to do all tests on all samples rather that sort them out according to the requisition slips. It was anticipated that the additional analytical data might be very useful. Experience has shown, through the work of Thiers and others, that this is certainly true. At any rate, we set about the construction of a machine to accomplish this purpose [8]."
This important concept of "profile" was verified by Thiers and colleagues [9]. The idea of displaying all patient data on a single sheet of paper was readily accepted throughout the world.

The instrumentation has changed quickly from automated analyses to multiple automated analyses. During this period, the Robot Chemist was introduced (first developed by Warner-Chilcott Laboratories, Instruments Division, and revised greatly in 1967, when the company was renamed AO Instrument Co.). This instrument was the forerunner of discrete analyzers. One clinical trial describes the Robot Chemist in 1967 as follows [10]: specimens were placed on racks that moved along analyzers on a conveyer belt. At the sampling point, the analyzer would swing out a sampling nozzle and transfer the necessary amount (ex., 0.1 ml alkali phosphates) from a sampling cup to a reaction tube. The location of this test tube was defined as position 1. A total of 52 reaction tubes were set on a circular turntable in a constant temperature bath, at 37°C, and the table turned clockwise one position per minute. If alkaline phosphates were to be assayed, then a substrate solution was placed in a reaction tube in position 45 so that the substrate solution could be pre-heated to 37°C until the tube arrived at position 1, which took 8 minutes. At position 12, a reaction-stop-solution was added, and the color developed was measured by a spectrophotometer. Based on the absorbance of a standard, the concentration of the specimen was calculated and digitalized with its absorbance using a software program, called Data scribe, developed by Warner-Chilcott. In addition, used reaction tubes were washed for use in the next test [10]. This mechanism is still being utilized in many automated analyzers, for an instance Hitachi product. Analyzers that use liquid reagents are called Wet-Chemistry, and those that use a film are referred to as Dry-Chemistry [11]. Consequently, the Robot Chemist was mechanically different from Technicon AutoAnalyzer, which provided clinical analyzers with some alternatives:

· Specimen transportation

AutoAnalyzer: turntable
Robot Chemist: horizontal conveyer

· Specimen allocation to reaction tubes

AutoAnalyzer: continuous flow method
Robot Chemist: discrete (none-continuous, independent) method

· Specimen processing

AutoAnalyzer: deproteinization is possible with dialysis
Robot Chemist: deproteinization can not be performed automatically

· Test result output

AutoAnalyzer: analog output to a recorder
Robot Chemist: digital processing by a calculation circuit

Meanwhile, instruments other than AutoAnalyzers were being developed. Since the early part of the 1960's, Anderson at Oak Ridge studied cell fractionation by the concentration gradient centrifugal fractionation method. To measure the enzymatic activities of fractions in centrifuge tubes directly, Anderson thought that either a target middle layer could be transferred specifically to another tube using centrifugal force [12] or enzymatic activities could be measured during centrifugation [13], and analyzers were constructed based on these ideas [14]. The basic purpose of these machines was to accelerate the performance in processing analysis. According to Anderson, like trains that travel from one station to the next, both continuous flow and discrete methods are insufficient for achieving fast analyses [15]. Three types of analyzers were introduced commercially: CentrifiChem (Union Carbide Research Institute), Gemsaec (ElectroNucleonics) and RotoChem (Aminco). RotoChem was initially introduced with 15 cuvettes, and later models were improved to accommodate 16, 30 or 42 cuvettes [16], and these analyzers were capable of performing either 15 different tests at once or test 15 specimens simultaneously. CentrifiChem was designed to handle 30 cuvettes [17]. At an AACC meeting in 1970, an interesting comment was made: "since humans are now traveling to the moon, it will also be possible to perform clinical laboratory tests there" [18]. The main disadvantage of the flow method was carry-over, and the discrete method was designed to resolve this problem. However, accuracy could not be improved drastically, and it was predicted that superior systems would be developed soon. Nevertheless, Gemsaec, which utilized centrifugal force, was touted as an analyzer that could be used even in outer space.

In the 1960's, AutoAnalyzer and RobotChemist were exported to Japan, where they were utilized in many laboratories. One of the reasons for this was that the Hospital Division of the Ministry of Health and Welfare encouraged improved efficiency and the integration of laboratory tests [19]. In 1968, the first domestic auto analyzer was manufactured by Hitachi Ltd. [20]: a total of 45 cuvettes were attached to a chain conveyer in a single line, and the conveyer was rotated from top to bottom. Cuvettes were placed so that the openings faced upward and the bottoms downward on the conveyer, and measurements were taken during this time. Then, when the cuvettes reach the end of the conveyer, they were positioned upside-down to be washed and dried so that they could be used for the next series of tests. A clinical trial was performed on this analyzer, but no complaints were received regarding the performance of the instrument [21]. The basic mechanism was followed as the tests were to be conducted manually. Cuvettes were arranged on a circular disc in the Robot Chemist but were rotated vertically in Hitachi's first analyzer. Only Hitachi 400 and 500 models employed this vertical movement. Cuvettes were arranged on a circular disc in Hitachi 700 and subsequent models.

In 1969, reviews on automated instruments for clinical chemistry were published both in Japan and overseas [22, 23]. Due to the introduction of AutoAnalyzers, the amount of information generated by the above-mentioned profile tests increased markedly. Then, in 1966, the implementation of information processing system began [24], and the movement to mechanize a large number of tests contributed to the development of new analytic methods. This in turn placed a greater burden on laboratories to perform more tests, and thus encouraged the development of instruments that could simultaneously perform multiple tests at higher speeds. Furthermore, there was demand to automate all laboratory tests and as each manufacturer tackled this task in a different way, the need for the standardization of laboratory tests soon became apparent. In addition, the quality control of laboratory tests was questioned, and there was a definite shift from simple coexistence to total harmony between man and computers. In the same year, Saito wrote the following:
"Has anyone purchased an automated analyzer, and is now having trouble deciding where to place it? Has anyone installed an automated analyzer, and is now having trouble locating work space for technologists? The idea of automation is to enable laboratorians to oversee more laboratory examinations. We have successfully installed automated analyzers in laboratories without sacrificing a single work desk" [25].
This concept was slightly different from that of Omori and colleagues:
"As long as clinical tests are performed manually, errors are unavoidable, and excepting errors made during analyses, from the time specimens are collected to the time reports are delivered to physicians, there is about a 2% chance of clerical error. To lower this figure further, the entire laboratory operation process must be automated, and this can be achieved only by installing computers and automated analyzers in hospitals" [22].

This is the historical background of the automation of clinical laboratories, which will eventually lead to unmanned operation. Of various automated analyzers, the design of Hitachi's was the most flexible. In the beginning, the M400 sampling method employed a gondola method in which ten specimen racks were rotated from bottom to upwards. Sampling was performed on the specimen rack at the top of the conveyer. A total of 100 specimens could be set on the instrument, and when more than 100 specimens had to be processed, racks that had gone through the entire round were replaced with new ones. In the next model, M500, an infinite number of specimens could be placed using a snake chain. However, Hitachi abandoned the infinite snake chain, and again chose the rack. Ever since the AutoAnalyzer was first developed, Technicon Corporation has utilized turntables (Picture 1), each of which could hold 40 specimens in sample cups. When more than 40 specimens needed to be processed, either another turntable had to be positioned beforehand, or each specimen had to be replaced manually.

Picture 1.AutoAnalyzer by Technicon Corp.

Ideally, it would be very useful if all one had to do is push a start button to run the automated analyzer until completion. It would be a waste of manpower if one laboratorian was simply relegated to replacing specimens. Let's review the units that were developed around 1970 when a series of comprehensive reports announcing the movement toward automation were published.

A variety of specimen handling systems were employed by an automated analyzer that were introduced at this time: the surrounding chain method for the Automatic Analytical System-AC600 (Pye Unicam Ltd.); the endless snake chain method for the Auto Lab (Lars L Jungberg Ltd.); the turntable method for the Mark X (Hycel Ltd.), the Beckman DSA560, the Zymat 340 (Joyce Loebl Ltd.), the type 1 ACA (Olympus Optical Co., Ltd., Japan) [26], and the JCA-6K (JEOL Co., Ltd., Japan) [27]; the ten-specimen-rack method for the Robot Chemist (AO Instrument Ltd.), the model 8900 (LKB Ltd.), and the model 3400 (Gilford Ltd.); the six-specimen-rack method for the IATRON-DKK MC600 (Iatron Co., Ltd. and Denki Kagaku Keiki Co., Ltd., Japan); the 24-specimen-rack (8x3) method for the model 3000 (Olli Ltd.). The ACA, manufactured by du Pont, later became the standard equipment for emergency test, was unique in that the necessary reagent for one analysis was placed in a 'bag', and each specimen was simply placed on the analytical bag [28]. This instrument should be considered solely from the total automation of laboratories.

Besides these analyzers, more analyzers were developed around 1970, but their specimen handling methods can basically be classified into one of the three following categories.

1. Turntable
2. Snake chain (or chain)
3. Rack

Looking back upon 30 years history of automated clinical analyzers, it is very sorry that the coexistence of man and computer was not considered in the earlier days. For the last 25 years, people who worked in laboratories were the ones who had taken it upon themselves to care for and attend to machines.
In the 1980's, handmade total automation systems were independently produced in Germany and Japan.

The clinical laboratory at the Grosshadern Hospital in Munich was designed by Knedel et al., and the laboratory was operated like a plant [29]. In this system, each specimen is placed in a canister and sent by air shooter to the laboratory. A clinical technologist takes the specimen and a requisition out, and the order is read by a marked-card reader. The specimen is centrifuged, then it is set on an automated sorter. Depending on the types of tests requested, serum is aliquoted in a sample cup, and at the same time, an identification tag is attached to the cup (a hole is made in the plastic by heat) to be transported to a designated analyzer via a snake chain. Several branching chains are utilized to carry specimens to the analyzers responsible. This laboratory has installed many analyzers, and technologists remove chains from the sorter as necessary, and set specimens on the analyzer, which all fit the same snake chains. The basic purpose of complete automation was to identify and process daughter-specimens. Complete automation could not be achieved using this system since specimens that are connected to snake chains must be set manually on analyzers.

In 1981, a new type of laboratory was developed in Kochi [30]. Blood samples are either collected at the collection center or delivered from wards, after centrifuging each sample if necessary, serum is classified manually among ten specimen racks (Hitachi Koki Co., Ltd., Japan) and then sent to analyzers by eight belt lines. Each instrument receives the racks and analyzes each specimen. Specimens are managed by means of bar codes. Even though specimens are sorted manually, unlike in Knedel's laboratory, the subsequent operation is completely automated.

These two total automated laboratory systems that were developed on the different lands opened the door for the laboratories on the next stage. Specimens are transported directly from patients to a central laboratory, and then assigned to a designated analyzer. Once specimens are delivered to the analyzer and are identified again before the necessary tests are determined, and after the requested tests performed, the test results are delivered to a laboratory information system. If the flow of tests were consisted of the flow of specimen and information (electronic signal), then they should be transported via totally different routes. What is truly utilized is the information from a specimen, so it must be free from the existence of specimen itself. The issue of the distance between blood collection site (patients) and analyzers will be discussed at some other time, but as blood samples must be treated as specimens, they must be managed and transported. In centralized laboratories, gathered specimens must be sorted for delivery to the appropriate analyzers. For this purpose, the automation of laboratory transportation system was developed, and in the last 25 years, two methods have emerged: the snake chain and the rack (not mention how many numbers of specimen carried by a rack).

Members of the Committee on Analytical Systems of the Japan Society of Clinical Chemistry (JSCC) recognized the field of clinical chemistry had arrived at a new era and understood appropriate standardization acts must be taken. Professor Masahide Sasaki at Kochi Medical School is eminent in the field of clinical chemistry, and the committee appointed him a chairholder. They found that the quality of the laboratory was based on identifying specimens and delivering them to the appropriate analyzers, and this could be achieved faster and should be more dependent on instrumentation. They saw if there were any disadvantages there might be a problem on the specimen transportation systems. Specimen transportation systems relieve workers from tough works associated with biohazards and manual labors. Furthermore, they emphasized that in order for analyzers and other instruments to be connected to a specimen transportation system, it would be necessary to promptly standardize the procedures and mechanisms in which samplers handle specimens.

With the cooperation of many manufacturers, the results of their consideration were introduced as the specifications of the JSCC. This is the five-specimen-rack. Rack is 100 mm long and 25 mm wide. Whether or not specimens are in the same rack or in the different racks, the distance between any two successive specimens is 20 mm. Details will be discussed later.

Soon after this, the Area Committee on Automation of the NCCLS began discussing standardization. Currently (July 1997), under the leadership of Professor Sasaki, the JSCC committee is actively working its specifications adopted as international one. Even though the details of the international specifications have yet to be finalized, it is clear that some type of international specifications will be established in the near future. As a result, all instruments and units that are involved in laboratories will be designed and manufactured according to one set of specifications. To achieve complete automation, it is important that all instruments be compatible with any specimen transportation system, regardless of where they were manufactured, and it is also necessary for information processing systems to be able to exchange data freely.

The discussion on standardization on the automated analyzers, specially on the specimen handling systems, is now complete, and in the remaining pages, I would like to present in detail the history of JSCC 'rack' standardization. This may appear meaningless to some, since these records will likely be lost in the future, it is important to preserve information on those who were involved in this acts and the contributions they made to this topic so future researchers can have information on the concept's origin.

At the Academic Communication meeting of JSCC held on July 22, 1993, the immediate past chairholder of the Committee on Analytical Systems Dr. Okuda (a professor at Osaka City University at the moment) commended Dr. Sasaki (a professor at Kochi Medical School) as the new chairholder. This meeting was one of a series held during the 13th Summer Seminar of JSCC (executive committee: Dr. Yoshiaki Katayama, National Cardiovascular Center) in Mino-o, Osaka. On July 23, 1993, the Committee on Analytical Systems was held to examine past issues and form a new plan under the leadership of the new chairholder. The fact that Dr. Sasaki was agreed as the chairholder was the most important move toward internationalization. The 14th Summer Seminar was held in Shinshu (executive committee: Dr. Shozo Nomoto, a professor of medicine at Shinshu University at that moment) on July 7, 1994, and the following programs were presented under the conduct of the Committee.

• Racks used in laboratory automation systems and problem points
  (presented by Dr. Ryo Fushimi, Osaka University).
• Rack management based on a bar code system
  (by Dr. Kanji Morita, Osaka City University).
• Necessary standardization of automatic analysis systems
  (by Dr. Masahide Sasaki, Kochi Medical School)

On January 17, 1994, Professor Nomoto requested that the name of the projects be submitted by the beginning of February. At that time, Professor Sasaki was only scheduled to speak at the seminar, then decided to add as above on February 22. In his presentation "Necessary standardization of automated analytical systems", Professor Sasaki made the following comments:

"In the last five to six years, more and more laboratory automation systems with transportation device are being installed in Japan. Consequently, new analyzers and transportation device utilizing specimen rack is under rapid development, and new instruments equipped with new technology are introduced continuously.

Since the configuration of transportation methods and the shape of racks differ among manufacturers, it is difficult to construct a laboratory out of analyzers with different specifications, thus making it necessary to spend a great deal of time and money to develop an ideal transportation system.

To resolve this problem, manufacturers, users, related companies and academics must come together as one to standardize the specifications of transportation system and analyzers.

Therefore, the Committee on Analytical Systems of JSCC proposes a plan to standardize automated systems. By thoroughly investigating this issue as users and as members of this academic society, we hope that a basic plan for the standardization of laboratory systems that is approved by as many people as possible can be established...."

Chairholder Dr. Sasaki also added that specimen container (test tubes), specimen rack, rubber stopper and transportation mechanism would be investigated at high priority, and he proposed the following concrete steps as part of the effort to achieve the standardization of specimen rack:

1. Regarding the length, width and height of specimen rack
   At present, different specimen racks are made to match various analyzers, and this is the biggest obstacle in achieving the standardization of transportation system. To standardize the design of rack, the length, width and height should be standardized first.
2. The size of test tube holes
   The outer diameter of test tubes is standardized, and the existing rack can be accommodated using sleeves, so the size of specimen tube holes is not a problem.
3. The number of test tubes in a rack
   The rack of current analyzers hold anywhere from one to ten tubes. However, depending on factors such as the design of the analyzers and the laboratory philosophy, the shape of the rack differs from one analyzer to the next. Therefore, to standardize the entire laboratory operation, it is necessary to decide the number of specimen tubes in each rack.
4. Bar code for each specimen in a rack
   As most manufacturers realize it is beneficial from the standpoint of test speed and test result management to be able to process specimens at random, most recent automated analyzers are equipped with a bar code reader. Nonetheless, to read the bar codes that are placed on specimen tubes, it is still necessary to facilitate the reading of the bar codes that are placed on specimen tubes by bar code readers. Since the size and location of specimen bar codes have been standardized by the Japan Society for Clinical Laboratory Automation, it will be necessary to standardize the length and width of the holes on the specimen rack.
5. Miscellaneous
   Racks can be constructed to hold up to a specified number of test tubes, but a rack that can hold any number of sample tubes through the attaching of additional racks (building block method) have already been introduced overseas.

In addition, to give each rack a certain degree of mobility, racks can be connected in a row (chain rack), and it is also possible for users to adjust the length of the racks (1, 5 or 10 tubes in a rack) depending on the type of analyzers used.

Furthermore, it is necessary to standardize the bottom thickness so that the bottom section of the specimen racks will not be caught by the gap between the rack guide and the belt."

So, even at this early stage, Professor Sasaki was aware of potential problems associated with the standardization of laboratory operations based on his experience at Kochi Medical School, and was determined to share his knowledge of these issues with his fellow researchers.
On November 4 and 5, 1994, the 34th JSCC Annual Meeting was held in Tokyo (organizer: Dr. Akiyuki Okubo, the then professor at Tokyo University). On the 5th of November, the Committee on Analytical Systems hold a committee that included representatives from manufacturers. A total of 27 people attended the meeting (13 committee members and the representatives of each prefecture of Japan, and 14 people from 9 different manufacturers). The name of the meeting was "Standard specifications for specimen transport systems," and the first topic on the agenda was specimen rack. During the meeting, chairholder Dr. Sasaki made the following comments:

"There are advantages and disadvantages associated with each type of rack. Even a small difference of 0.5 mm can have considerable impact in some cases. For example, the pitch of the specimen rack made by Hitachi Koki Co., Ltd., Sysmex Corp. and Hitachi Medical Corp. are 19.5 mm, 20 mm and 23 mm, respectively. If the pitch is standardized to 20 mm, then most manufacturers should have to redesign their machine tools, and users would also have to make certain adjustments to their systems. This may be fair in a sense, as everyone would be forced to make some changes."

The title "Standard specifications for specimen transportation systems" was adopted as the name of the project. However, due to the Hanshin-Awaji Earthquake of 1995, the survey that was planned to be conducted after the meeting was postponed, and questionnaires were not distributed till March. Then, based on the results of the survey, the "Kochi Plan" was formed.

When the 9th Spring Seminar of the Japan Society of Clinical Laboratory Automation (seminar director; the late Professor Shiro Uesugi, Akita University) was held in Akita, committee chairholder Dr. Sasaki discussed the results of the survey and displayed a plastic model that was created according to the Kochi Plan: width 25 mm, height 65 mm, pitch 20 mm, and length 100 mm.

At this time, the committee proposed to adopt the Kochi Plan as the plan of the Committee, and it was determined that the draft that was prepared on May 9, 1995 would be evaluated before the 15th Summer Seminar (seminar director: Professor Isao Kobayashi, Gunma University) that was to be held in Kusatsu-Onsen on July 26 and 27, 1995. The draft was presented to the members of the society, and openly discussed. In addition, on the 27th, the Committee and the representatives of each prefecture held a closed meeting to investigate possible future plans. A total of 13 people attended the meeting, and in addition to the model that was presented with the Kochi Plan, another model with a rack ID was introduced.

The draft was then treated as a tentative plan, and when the 27th Japan Society of Clinical Laboratory Automation Annual Meeting was held in Kobe on September 14th, 1995 (organizer: Dr. Yasuhiro Ohba, a professor at Kinki University then), another expanded meeting was held with a total of 27 people, including 16 representatives from 10 manufacturers. The expanded meeting was held due to strong demand from the manufacturers, who wished to reevaluate the original draft, and the following comments were made.

1. "How about the strength of the rack? In particular, when
   racks are centrifuged at 2,500 - 3,000 G."
2. ."We are currently using 23-mm pitch because some caps are oversized for 16-mm tubes. If the pitch is standardized at 20 mm by JSCC, we are concerned about the caps disturbing each other."
3. ."We would like to determine if the new plan is compatible with the existing rack to avoid placing an excessive burden on users."
4. If the width of our rack is increased, the safety of our system will be compromised at storing racks.
5. Could you encourage other organizations to evaluate the present specifications?

The committee's position on these comments was as follows: the strength of racks can be ensured by selecting an appropriate raw material. The pitch of sample racks, and specimen tubes and caps will be standardized. We are talking about rather than continuing to utilize existing racks that will be produced in the future. From a certain year the new specifications must be employed. If users insist on utilizing former systems, then it would cost them higher. Also, related organizations have responded that if the draft is officially adopted by the society, then they would respond accordingly.

Other topics were also discussed, and the present plan was approved unanimously. The plan was then presented to the board of directors on December 25, 1995, approved at the first meeting of 1996, and published as the specifications of JSCC in their publication "Japan Journal of Clinical Chemistry." When the 36th JSCC Annual Meeting (organizer: Dr. Shoji Kume, a professor at Yamanashi Medical Collage) was held in Kofu, Yamanashi on October 24, 1996, the correspondences on the questionnaires that were sent to related societies and organizations to ascertain their opinions on the specifications were evaluated, and it was determined that the specifications would be effective without any revisions or addendum on the proposed standard.

1. Skeggs LT: An automatic method for colorimetric analysis. Clin Chem, 2:241(abstr.) , 1956
2. Muller RH: Automatic colorimetric analyzer eliminates need for many analytical procedures. Anal Chem, 30: 53A - 56A, 1958
3. Marsh WH, Fingerhut B, Kirsch E: Determination of urea nitrogen with the diacetyl method and an automatic dialyzing apparatus. Am J Clin Path, 28:681 - 688, 1957
4. Blaedel WJ, Hicks GP: Analytical applications of the continuous measurement of reaction rate: lactic dehydrogenase in blood serum. Anal Biochem, 4:476 - 488, 1962
5. Schffert RR, Kingsley GR, Getchell G: Automated determination of serum glutamic oxalacetic and glutamic pyruvic transaminase. Clin Chem, 10: 519 -532, 1964
6. Skeggs LT: Multiple automatic sequential analysis. I. Theoretical considerations. Clin Chem, 9: 442-443, 1963
7. Skeggs LT: Multiple automatic sequential analysis. II. Construction and operation of an eight channel instrument. Clin Chem, 9:443, 1963
8. Skeggs LT, Hochstrasser H: Preliminary evaluation of the SMA-12 survey. Technicon Symposia Vol. 1, pp351- 356, Mediad New York,1967
9. Bryan DJ, Wearne JL, Viau A, Muser AW, Schoonmaker FW, Thiers RE: profile of addmission chemical data by multichannel automation: An evaluative experiment. Automation in Analytical Chemistry Technicon Symposia 1965, Skeggs LT ed., pp 423-426, New York, Mediad, 1966
10. Morgenstern S, Kessler G, Auerbach J, Flor RV, Klein B: An automated p-nitrophenylphosphate serum Alkaline phosphates procedure for the "Robot Chemist". Clin Chem,11: 889-897, 1965
11. Newman HJ, Hilton M, Crawford WC, Kearney R: A new approach to the automation of wet-chemistry analysis. Clin Chem, 12:554, 1966
12. Anderson NG: Analytical techniques for cell fractions. IX. Measurement and transfer of small fluid volumes. Anal Biochem, 23: 207-218, 1968
13. Anderson NG: Analytical techniques for cell fractions. XII. A multiple-cuvet rotor for a new micro analytical system. Anal Biochem, 28: 545-562, 1969
14. Hatcher DW, Anderson NG: GeMSAEC: A new analytical tool for clinical chemistry. Total serum protein with biuret reaction. Am J Clin Path, 52:645-651, 1969
15. Anderson NG: Basic principles of fast analyzers. Am J Clin Path, 53778-785, 1970
16. Burtis AC, Johnson WF, Attrill JE, Scott CD, Cho N, Anderson NG: Increased rate of analysis by use of a 42-cuvet GeMSAEC fast analyzer. Clin Chem, 17: 686-695, 1971
17. Fabiny D, Ertingsgausen G: Automated reaction-rate method for determination of serum ceratinine with CentrifiChem. Clin Chem, 17:686-700,1971
18. Mather A, Roland D, Cooper TG, George B: Discrete-Processor automation -Does it have a future? Clin Chem, 16:529, 1970
19. Ishii, C. and Yoshimura, F.: Blood sugar determination by
AutoAnalyzer. Clin path, 9: 191-195. 1961
20. Yoshida, K. and Ono, N.: Auto analyzers for clinical chemical
tests. Clin Path, 16: 285, 1968
21. Mizuno, E., Ono, K., Kitamura, M. and Sudo, K.: A study on
Hitachi automated analyzers. Society of Clinical Laboratory
Automation Symposia 1965: 45-51, 1969
22. Omori, S. and Kato, M.: State and trend of clinical laboratory
automation. Jpn Clin, 27: 2833-2845, 1969
23. Alpert NL: Automated instruments for clinical chemistry. Clin Chem, 15:1198-1209, 1969
24. Hicks GP, Gieschen MM, Slack WV, Larson FC: Use of small high speed digital computer as a routine laboratory. Clin Chem, 12:555, 1966
25. Sato, M.: Implementation of automated analyzers. Clin Lab, 13:
1226, 1969
26. Yamamoto, T., Takekawa, H. and Sakano, Y.: Experimental
automated analyzers. Society of Clinical Laboratory Automation
Symposia: 2-4, 1970
27. Okukubo, N., Sekiguchi, M., Nagauchi, N., Komabayashi, T. and
Obayashi, T.: A study on JEOL's automated analyzer, JCA-6K:
Part I. Society of Clinical Laboratory Automation Symposia, 6:
107, 1975
28. Neadeau RG: Development of the du Pont automatic clinical analyzer (ACA)system: Part I. Clin Chem, 14:778-779, 1968
29. Sator H, Neumeier D, knedel M: Optimierung der Provenverteilung in einemzentralisierten klinischen-chemischen Institute. J Clin Chem Clin Biochem,19: 1107-1115, 1981
30. Sasaki, M., Sonobe, H., Nishida, M., Goto, Y., Ogura, K.,Okumiya, T. and Kataoka, H.: Development and application of belt line systems: 1. System Overview. Clin Path, 30 (additional
copy): 341, 1982