The Paperless Analytical Lab

Today’s laboratories are besieged by demands for improved efficiency, increased productivity, improved data quality, immediate access to data and tighter cost control In addition, increasingly sophisticated laboratory instrumentation requires the day-to-day management of floods of analytical information. The traditional paper-intensive management systems found in today’s laboratories cannot address these demands or efficiently manage the volume of data produced.

For today’s analytical laboratories to gain and maintain a competitive advantage, they must use integrated electronic systems. The paperless laboratory is a vision of the analytical laboratory of the near future. It features flexible electronic systems designed to manage, review, store and report analytical information more effectively.

The paperless laboratory of the future is radically different from the typical paper-burdened analytical lab of today. Under the current system, samples and records typically are generated from one of three sources: a product/material release system (new products), a stability system (concerned with shelf life), or a sample request system (requests for analysis). These databases sometimes reside on three separate computer mainframes. Records and samples are delivered manually to a central sample distribution center, where technicians sort them and deliver them to the appropriate laboratories for testing. At the labs, technicians enter the samples manually into individual laboratory sample tracking systems, and then schedule it for testing.

The sample scheduling process is also done manually. Technicians schedule samples for testing based on the volume of work, critical sample lists identifying those lots requiring immediate testing, lot turnover rate objectives, and the resources available at the time. When a batch of samples lists scheduled for testing, laboratory personnel initiate a paper primary record. These records include sample identifications transcribed manually from the paper transport record, procedural information, reference and performance standard data, sample weights or volumes, dilution factors, instrument identification and calibration data and so forth. The greater part of a day can be spent preparing the samples and instruments for the analysis. Typically, workers collect the raw data and store it overnight on some type of computer. The next morning, they enter information from the paper primary record and the appropriate data file into a calculation program. They then transcribe the final calculated result onto the paper transport record.

At this point, workers collate all of the paper primary documentation and procedures into a packet and deliver it to a record checker, who double-checks all transcriptions and calculations in this primary data package and ensures that the test was run according to procedure and that all performance measurements were within required limits. Once these checks are complete, the records are approved and the sample is logged out of the laboratory sample tracking system.

The paper primary records are stored in the laboratory for later reference and eventually are sent to a record storage facility. The paper transport record is sent to data entry personnel for keypunching into the appropriate database. A second person normally is required to verify the accuracy of this data entry step. To complete the process, each month the data recorded on these systems can be stripped and combined with other statistical information (for example, a time accounting system) to produce a retrospective look at laboratory performance.

These manual systems are tedious, labor-intensive, and very susceptible to transcription errors and can lead to a reduction in overall laboratory productivity. They also can contribute to excessive production times and the inefficient flow of information.

The Paperless Lab

For today’s laboratories to gain the competitive advantage, they must develop integrated electronic systems to manage, review, store and report analytical information. These systems must include electronic data transfer, interactive sample tracking, automated identification systems, computerized sample scheduling, electronic worksheets, and automated data review and release.

Computerized sample tracking and data management systems in the paperless laboratory will replace manual tracking and scheduling systems. These new systems will provide instantaneous status reports on sample lots for the purpose of planning, scheduling and expediting lot release.

The three databases requesting laboratory services—the sample distribution/information center, product/material planning system, and quality assurance information system—will be linked electronically. This will allow for better projections of workloads, both for the long- and short-term.

A computerized sampie priority network will consolidate the information from production and planning schedules, the cost of inventory, efficiencies of sample grouping, and sample target dates to optimize lot turnover time. Once this information has been centralized, it will be linked to a system that will create daily work schedules for each laboratory and analyst. Once samples are received and logged into the sample distribution/information center or database, the system will automatically assign a due date based on established priorities. It will then generate bar code labels and affix them to the laboratory samples. That can then obtain sample information by querying the database, instead of making phone calls to telephone answering machines.

Robots will deliver samples to the appropriate laboratory for testing. Here, a scan of the bar code label will change the sample status to “received awaiting testing.” Computerized sample scheduling will then begin. The system will generate daily work schedules automatically for each laboratory and analyst. It will schedule samples for analysis based on the due date assigned by the sample priority network, resource information and workload forecasts from other laboratories. Transferring samples or manpower more efficiently will be enhanced by networking laboratory sample databases.

When a sample is ready for testing, the system will begin an electronic primary worksheet. A query to the database will display the sample status updated to “in test.” Procedural information will be available electronically. Many common laboratory operations (such as sample, standard and instrument identification) can be represented by a sequence of bar code scans, eliminating the need for keyboard entry. The use of bar codes at all stages of the process will greatly reduce errors and make the system easy to use.

The system will use balances linked to computers to transfer sample and standards weights to the electronic worksheet. As data are transmitted to this primary record, the automatic data review process will begin. The computer will compare all information transmitted with previously established limits, and will flag data outside of limits with a warning message or issue an exception report. Analysts will have this information immediately instead of the next day.

Robots and instruments will be interfaced to the computer to the maximum feasible extent. The automated data review process will be activated as data are being collected. Once all the data are collected and checked, calculation programs will begin to generate the final result automatically. The computer will flag for review any results or performance parameters that are outside of established targets. Artificial intelligence systems could be used to assess the acceptability of assay results or performance parameters. The completed analyses will be transferred automatically to a reviewer worklist. The computer will now update the sample status to “waiting for review.”

The net result is a system in which human judgment and data review are substantially replaced by a computerized system of review by exception. (Review by exception is the process of evaluating only those parameters that exceed established limits.) The benefits of computerized review include consistent application of quality standards, reduction in laboratory personnel, and shorter turnaround times.

Automated data release will allow the safe and immediate transfer of analytical information. Human error will be eliminated and sample lots can be released much more rapidly. These integrated electronic systems will free laboratory analysts to do what they do best—analysis—and thereby contribute to improving the quality and quantity of their output.

Upon electronic approval, the system will log the sample out of the laboratory automatically and transmit the data electronically to the appropriate database. This will completely eliminate manual data entry and verification. Laboratory primary data can now be archived electronically, instead of boxed up and sent to record storage. Laboratory performance information will now be available for both prospective and retrospective evaluations.

The Upjohn Co. is currently implementing systems for a paperless laboratory in its control labs. Called the Automated Laboratory Systems (ALS), the program is an integrated laboratory automation system linking the elements described above.

These links will be established using a blend of software developed by Upjohn and two commercially available laboratory information management systems—the Digital LIMS/Sample Management and the Varian LIMS/Data Management software packages. The Varian LIMS/DM system is designed to handle data and sample information obtained by manual entry or by instrument interfaces for a single laboratory. The Digital LIMS/SM is designed for sample management activities across many laboratories. The LIMS/SM package is designed to link up with the LIMS/DM package and with external sample databases (such as, stability, product release, request for sample, and so forth).

There are several potential obstacles to making the conversion to a paperless laboratory. The goal of computerization is to provide services beyond those currently available, not merely to provide old tools in a different way. An extensive analysis of where and how analytical information is used in the organization served by the laboratory is a necessary first step. Without such an analysis, there is a good chance computerization will create a system that will only generate more paper faster.

Another major obstacle is minimizing the disruption caused by bringing new technology into an organization. Conversion to a paperless system could fail not because of technological weaknesses, but due to an inadequate analysis or misunderstanding of the interface between workers and the machines. Insensitivity to user needs and motivations, lack of attention to political and personal considerations, or lack of proper worker education can lead to a disastrous implementation. Key issues such as system planning, training, education, software support, system acceptance and growth must be explored to ensure a successful conversion.