AUTOMATION FOR DIAGNOSTICS
) |
Figure 1. (click to enlarge) A few of the systems that are a part of the automated testing portfolio of BD Diagnostics (Franklin Lakes, NJ). Updating such a wide range of instruments to address technological developments and consumer expectations poses challenges.
|
In order to successfully meet these challenges, accurate, timely test results must be delivered to physicians and, ultimately, must lead to improved patient outcomes. These microbiology results are equally important to other healthcare professionals, especially those charged with managing healthcare-associated infections (HAI). Infection-control professionals and pharmacists are responsible not only for the outcome of each patient, but also for the continuous examination of epidemiological trends in their facilities.
Today, and even more in the future, a patient’s microbiology test record is derived from a combination of tests—manually performed rapid assays (i.e., traditional gram stain or newer-generation rapid assays), traditional plated media testing or new chromogenic media testing (growth-based testing), automated blood-culture analysis, automated organism identification and antimicrobic-susceptibility testing, and molecular testing. These multiple tests are performed by different technologists and must be collated to provide a complete patient record. More important, the results of one test can affect the interpretation of another. This means that the complete patient testing record must be examined as a whole to provide an accurate, total result to the attending physician. Because of these factors, it is a standard practice for the microbiology supervisor or senior technologist to review every patient result for accuracy. These individuals apply their knowledge of microbiology, generally accepted practices such as
Clinical and Laboratory Standards Institute (CLSI) rules, and hospital policies to complete the testing of each patient. This scenario can lead to inefficiencies, which can have a negative impact on a department already stressed for resources. It also places the burden on the memory and consistency of an individual, which can lead to inaccuracy.
For managers and their staff, this process becomes more complex as new anti-microbic-resistance mechanisms emerge. To gain efficiency and consistency between technologists, managers construct testing algorithms or pathways that must be manually applied in the testing of specimens and the interpretation of these results.
Computerized laboratory information systems (LISs), which primarily provide specimen tracking and individual test result records, do not adequately address the analysis needs of these departments. However, efficiency and consistency can be gained in the lab through the computer-aided analysis of patient data. These computer systems analyze the data from multiple sources, apply logical algorithms and rules, and present these refined results to technologists and physicians. These expert systems must incorporate the latest information from recognized organizations such as CLSI or its international equivalents. The systems must also factor in the multiple specific reporting and quality standards that each facility has in place. These standards take into account factors such as a facility’s antimicrobic formulary, physician-specific practice policies, and quality assurance measures put in place by the microbiology supervisor, pathologist, infection-control department, or pharmacy.
System Development
The BD Diagnostics (Franklin Lakes, NJ) product line offers specimen-collection and transport products. It also includes instruments for sophisticated automated identification and anti-microbial-susceptibility testing, such as the BD Phoenix automated microbiology system and the BD GeneOhm molecular testing systems (see Figure 1). This broad array of products allows BD Diagnostics to apply an expert system that integrates and applies logical analysis to producing combined test results. Only an LIS, or some other middleware solution, would have this same opportunity. Adequate microbiology labs do not have such a solution.
Because of its broad portfolio, BD Diagnostics faces challenges in developing some of its automated testing systems:
- BD Phoenix, for the automated identification and anti-microbic-susceptibility testing of microorganisms.
- BD Bactec instrumented blood culture systems, for the detection of microorganisms in blood culture.
- BD Bactec MGIT 960 mycobacteria culture system, for the identification of mycobacteria and for anti-microbic-susceptibility testing of this class of organism.
- BD GeneOhm MRSA assay, for the molecular detection of anti-microbic-resistance mechanisms such as methicillin-resistant Staphylococcus aureus.
The development of these systems requires significant time and resources. It also requires extensive field validation and regulatory approval in multiple countries. In addition, the BD Diagnostics support staff must receive considerable training on each type of system so that it can, in turn, provide training and support services when installing a new system in a customer’s lab.
Systems are not developed simultaneously; instead, development is a sequential process. In a period of several years, a company may focus only on the development of a particular system (e.g., for mycobacteria detection). Over many years, the company may develop multiple generations of this system.
When each system is developed, a team consisting of multiple departments and functions is formed. One of the engineering functions is software development. This software team is assigned the creation of the software that will operate the instrument, ensure the quality of operation, provide the end-user the interface to the system, perform the analysis of the testing, and provide the reporting of the results and other functions. Once a system is developed and launched to the market, these teams are assigned to new projects. Modifications to the platform are addressed by new teams after many years.
This model of development is also employed by competing systems. Each generation of product has some type of data manager developed alongside the instrument. Because many competitors do not have a broad array of manual assays and automated systems used in the microbiology lab, the opportunity to link the assays does not exist.
Generational System Development
When BD Diagnostics develops a new generation of an automated system, the engineering team develops the internal computer architecture and a set of software to enhance the use and performance of the individual system. In developing the computer hardware, only enough architecture is used to meet the needs of the particular system. By doing so, the team can ensure that the scope of the project is efficiently managed and that no unnecessary hardware costs are built into the system, thus addressing the company’s need to control manufacturing costs and final costs to the lab. Likewise, the software team uses the current generation of software development tools and limits the project to meeting the exact needs of the current system.
While this method of development has proven efficient in the past, it does not address the changing needs of the microbiology lab or the rapid change in the computer industry, nor does it address the changes occurring in HAI management. It also does not consider the customer’s expectation of innovation in testing systems. When a user interface appears dated and no new functionality is provided through software updates, the system quickly becomes dated in the mind of the customer.
Other elements are also lost in this generational-development model. With an individual-system development approach, manufacturers can often overlook the need to integrate results provided by other testing systems. Retaining valuable software expertise and building a strong team are also not addressed by this type of system development. When a new system project is completed, the team formed to develop the system is moved on to multiple new projects or to other positions in the company. Given this environment in which much history and expertise are lost, updating an individual system is difficult.
A New Development Model
With the development of the BD Phoenix system, BD Diagnostics instituted the codevelopment of a new software platform called the BD EpiCenter microbiology data management system. The BD EpiCenter system was designed to address not only the many needs of BD Phoenix system customers, but also the rapidly changing nature of microbiology and the complex problem of inefficiency in new system development. The BD EpiCenter system also provides a platform for addressing other customer needs, such as improved communication with the infection control and pharmacy departments.
) |
Figure 2. (click to enlarge) A screen shot depicting the graphing capabilities of the EpiCenter microbiology data management system.
|
Unlike other system development teams that are formed and reformed as new instrument systems are developed, the BD EpiCenter system software team remains intact, continually developing the software and releasing regular updates. These updates not only refine the management and capabilities of multiple BD Diagnostics systems, but provide added value by incorporating new capabilities in each version (see Figure 2). Each new version improves the efficiency of its end-users and enables them to deliver better patient care. This focus on the BD EpiCenter system improves the development process while providing customers with a much-needed integrated solution.
A recent example of this efficiency can be seen in the development of the latest generation of blood-culture instrumentation by BD Diagnostics. In the previous product generation, pathway hardware and software would need to be included in the design to enable communication between multiple units. Likewise, long-term data storage, report generation, and LIS interface capabilities would need to be designed into the platform.
By integrating the BD EpiCenter system into this development process and reusing the existing infrastructure of the software, BD Diagnostics provides the required capabilities while saving resources and accelerating the product’s time to market. This combination of efficiency and savings not only benefits the manufacturer, but ultimately the customer, who has access to more-advanced instrumentation sooner.
In addition, the end-user receives an instrument with a familiar user interface. This platform continuity allows for better training and support and eliminates the need to maintain the training and service requirements of multiple platforms. Additionally, the BD EpiCenter system has a remote Internet diagnostic capability which enables technicians to better serve customers in real time.
The BD EpiCenter system uses a network architecture that provides new opportunities for customers to integrate and consolidate tests in their laboratories. With traditional system development methods, the system software is fixed to the location of the instrument. The user must return to the instrument to monitor or work with the system. The BD EpiCenter system, in contrast, offers a completely integrated Windows application.
The software can be installed and deployed on multiple systems throughout the laboratory, improving the efficiency of the staff by providing multiple points of access. This is especially true for the microbiology department, where the lab typically is not staffed around the clock. Using this distributed functionality, other departments can be enlisted to monitor critical microbiology tests such as blood cultures when the microbiology department is not staffed.
Timely reaction to a positive blood culture can directly affect patient outcomes. Therefore, the BD EpiCenter system can alert the lab in the event of a positive blood culture and can continually monitor the lab staff’s reaction time to the event. This monitoring provides laboratory management with a quantifiable quality assurance measure that can become a quality practice goal.
) |
Figure 3. (click to enlarge) A screen shot showing the ability to set conditional alerts with the EpiCenter system. An active alert system enables laboratorians to better monitor testing.
|
The software enables staff to construct rules using a patent-pending graphical user interface (see Figure 3). Once in place, these rules automatically produce alerts when specific conditions are met. In version 5.5, the latest release of the BD
EpiCenter system, an optional BD Sentinel software module can send customized messages to desktop computers or to any device used by physicians, infection control staff, or pharmacists that can receive e-mail.
Regularly updating system software, and thus providing totally new functionality, is new to the market. This continual software improvement creates additional value for potential customers. The BD EpiCenter system has demonstrated success with this concept, releasing six versions of the product in a span of about seven years.
Reaching beyond the laboratory
The BD EpiCenter system has enabled BD Diagnostics to meet the needs of hospital staff beyond the laboratory. While the company’s laboratory customers are responsible for providing rapid, accurate results, physicians are responsible for interpreting these data for individual patient care. Other departments, such as infection control and pharmacy, are responsible for monitoring HAI trends. The software’s network architecture provides physicians and staff in these other departments with direct access to patient data that have been refined by the expert systems in real time.
 |
Richard Jefferson is worldwide marketing manager for the BD EpiCenter system at BD Diagnostics (Franklin Lakes, NJ). He
can be reached at richard_jefferson @bd.com. |
The software adds efficiency in the monitoring of infection trends by providing an extensive library of ready-to-use epidemiology reports. These reports enable infection trend information to be monitored and communicated without asking the IT department to help extract and collate the HAI data. And, because the BD EpiCenter system has an export function that is compatible with Microsoft applications including Excel and SQL Windows, these reports can be easily compiled and shared outside the institution.
Conclusion
The BD EpiCenter system innovates the ways in which patient results from different testing systems are consolidated and analyzed in the microbiology laboratory. The positive outcome from this combined testing integration is more-accurate and more-rapid results. The system also enables other departments to become more efficient in their monitoring of infection trends, thus helping to prevent new infections and improving patient safety.
The system’s software architecture and intrinsic capabilities can be leveraged across the development of multiple instrument systems, thereby saving manpower and reducing time to market. This combination of new efficiencies, for laboratories as well as for the product development team, ultimately helps improve care for the patients they both serve.
