| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1 University of Pennsylvania Medical Center, 3400 Spruce St., Philadelphia, PA 19014-4283
 |
Introduction |
|---|
 Top Introduction Summaries of Papers Addressing... |
|---|
It is perhaps easier for hospital administrations to assess the productivity of their clinical laboratories than of most other hospital services. The number of tests, the number of staff, and the cost of running the service as determined by the supply and salary budgets can be readily quantified. Furthermore, these factors can be bench-marked against the performance of other institutions. However, clinical laboratories also have to contend with the absurd concept of the "billed test" beloved by the federal government, insurance carriers, and consulting companies lacking laboratory expertise. The "billed" test assigns equal weight to a multitest outpatient panel as it does to a dipstick urinalysis or to an elaborate genetic test that is labor-intensive and may take days to complete. This ridiculous concept makes comparisons of productivity between institutions impossible. Indeed, the billed test concept hides increases in productivity because one billed outpatient test may generate as much work as 12 inpatient tests. Successful efforts by hospitals to reduce their inpatient testing, because of non-reimbursability, then mask any increase in revenue-generating outpatient tests. This dual objective of reducing unnecessary inpatient testing and capitalizing on the potential for outpatient revenue has become a major charge for the responsible clinical laboratory director.
Clinical laboratories everywhere have been faced with the challenge of doing more tests at less cost, i.e., boosting their productivity. Many laboratories have reached the point at which it is impossible to increase productivity using the equipment that they have. Although each generation of "automated" analyzers usually provides some improvement in throughput and turnaround time for results, they do not have the ability to make the quantum improvements that are a prerequisite to significantly improving productivity. This has led to the concept of "total laboratory automation", as much a misnomer as "automation" is for a single laboratory instrument. Total laboratory automation goes beyond the automation of analyses but includes automation of much of the important hitherto labor-intensive manual preanalytical phase in the process. The concept was conceived in Japan and has been widely accepted there, so that many large Japanese hospitals now include robotized specimen processing and delivery systems. In the United States, only a very small proportion of even the largest hospital and reference laboratories have installed such systems. Clearly, many laboratory directors have been waiting to learn of the success, or otherwise, of the automated systems in daily operation before they, too, embark on such a major investment. Many also remain uncertain as to whether maximum centralization, as represented by total laboratory automation, is to be preferred over maximum decentralization, as represented by point-of-care testing.
The 1999 Clinical Chemistry Forum was designed to present the arguments as to why a fresh approach to laboratory testing was needed and to detail the steps necessary to make the decision whether to commit to total laboratory automation and how to identify the steps involved in a successful installation. The presentations began, appropriately, with discussions of alternative approaches to coping with rapidly escalating workloads. These included total laboratory automation for both individual hospitals and for networks of hospitals. Within the laboratory, alternative approaches were presented, including the use of modular components and automation of selected fixed tasks.
The topics covered included a discussion of the components of the necessary overall planning process by a senior administrator from an integrated health system. Another paper dealt with the internal marketing of the concept by the laboratory to the administration and medical staff who would have a major, and vested, interest in the successful operation of a new system. Two of the critical areas that can make or break a robotic system are the layout of the facility with its attendant demands, which involves providing an appropriate environment for both the operators and the analytical systems, and the design and implementation of a superior information system. The latter is essential for capitalizing on the rapid generation of test results. The planning for an automated laboratory entails much more than the operation of the system once it is installed. One of the difficulties in many laboratories is maintaining the daily processing and testing of specimens while a large part of the laboratory’s space is taken out of service during construction. An especially difficult area to manage is ensuring the loyalty and productivity of staff. This is particularly true when they are aware that one of the objectives of installing a robotized laboratory is to reduce labor costs, which must inevitably impact some of the staff whose goodwill and cooperation are essential. This also is essential during all of the steps before the successful introduction of routine operation of the system on a daily basis. A majority of the forum papers are presented here in their full-length form. Four other papers are summarized below that address key problems in working toward an automated laboratory.
We believe that the meeting achieved its objective of presenting all of the issues that need to be recognized by a laboratory director before embarking on the very challenging and expensive pathway leading to total laboratory automation. Although this concept has been well accepted in Japan, the small number of installations in the US to date means that those laboratory directors who have installed systems are still pioneers. We are grateful that they were willing to share their experience at the 1999 Clinical Chemistry Forum. In addition, the attendees and the readers of these Proceedings need to recognize the dedication and support given by Jean Rhame and Pamela Nash of the American Association for Clinical Chemistry’s staff, who made the meeting happen.
|
Summaries of Papers Addressing Key Problems in Working toward an Automated Laboratory |
|---|
|
TopIntroductionSummaries of Papers Addressing... |
|---|
Julie A. Fisher, Mount Sinai Medical Center, New York City, discussed selling the concept of a totally automated laboratory to a hospital’s administration and other stakeholders. Successful selling is based on extensive communication and detailed financial and other justifications. There are eight essential elements to successfully selling an automation concept. These are defining goals, assessing needs, obtaining stakeholder buy-in, the decision-making process, vendor selection, the financial planing process, implementation, and metrics. Continuous communication is essential throughout all phases of the project. The wishes of the laboratory must be congruent with those of the administration. The process may be protracted; the cycle between initial concept and routine operation may be as long as 6 years.
The trigger for a laboratory to consider automation usually is pressure to reduce costs and improve its efficiency. Automation has the potential to enhance the economic survival of a laboratory, reduce its operating costs, improve the quality of services, and provide a safer work environment. The need for automation should be assessed in the context of whether the institution is planning to expand or to just cut costs. Every ramification must be considered. For example, contractual arrangements with unions must be taken into account. This will become particularly important when the system is fully implemented because contracts may determine who may or may not be laid off. Additionally, needs for upgrading or changing the laboratory information system and analytical instruments must be assessed.
Stakeholder buy-in.
A successful automation project depends on
stakeholder buy-in. The stakeholders include the laboratory staff, the
hospital administration and Board of Trustees, and hospital physicians.
It is important to communicate to each of the groups what automation
will do for them. Each of these constituencies has different interests
and concerns. The laboratory staff are most concerned about job
security, but it is important to let them know that automation is a
tool to help them perform their jobs differently, and perhaps better.
For the administration and Board of Trustees, the focus needs to be on
the financial bottom line, with emphases on the opportunity for both
revenue enhancement and expense reduction. Other selling points for the
administration can include the potential to perform tests for
other hospitals and develop group purchasing arrangements with other
hospitals for which laboratory services can be provided. Physicians are
primarily concerned with turnaround times of test results as well as
enhanced information.
Financial planning.
The financial planning process requires
projections of revenue and expenses. A break-even analysis is essential
and must demonstrate that automation will reduce costs and/or enhance
revenue. Various approaches may be used. A traditional return on
investment (ROI) analysis relates net income to investment capital. The
formula for calculating a ROI may be refined to take into account sales
as well, as in a DuPont analysis. This approach recognizes that it
might not be beneficial to tie up assets, thereby lowering
profitability. The same formula can be used for an expense analysis by
keeping sales constant. The net profit margin increases with a
reduction in expenses, and with automation, the key expense reduction
is in labor. Technical productivity can be calculated by dividing the
number of tests performed by the total number of paid full-time
employees or equivalents (FTEs). The calculation of labor savings
should take into account how the number of employees will be reduced.
With layoffs, there often will be severance and/or retraining expenses
to equip the laid-off employees for other jobs. Different laboratory
areas will be affected differently. Thus, the laboratories in which
automation will be implemented will be more impacted than others. For
each laboratory area, a separate projection of staffing needs to be
done.
Beyond a ROI.
Recently, there has been a trend away from
justifying automation solely on an ROI analysis because not all of the
benefits can be quantified in financial terms. Automation provides
added value through improved efficiency coupled with reduction in
processing errors, improved turnaround times, automated repeat and
reflex testing, enhanced safety, and improved specimen tracking.
Maintaining goodwill.
The active participation of stakeholders
in the planning process enhances the laboratory’s ability to sell the
concept. Thus, an overall executive committee derives benefits when
supported by laboratory management with information systems and
instrumentation teams. It is advantageous to enlist stakeholders in
vendor selection because acceptance of the system is critically
dependent on the their involvement. The more people involved in
different aspects of the planning process, the greater the probability
of acceptance. Even during the implementation phase, it is important to
involve the stakeholders, especially the staff who will be directly
affected by the system. During the installation and after the system
becomes operational, it is important to continue to communicate to the
stakeholders. Information that should be communicated includes actual
performance compared with projections, especially with regard to
revenue projections and/or expense reductions, the quality of service,
and whether a safer environment has been created.
Patricia Abbott, Hospital of the University of Pennsylvania (HUP), Philadelphia, discussed the practical aspects of creating a robotized laboratory. Because acceptance of laboratory automation by a hospital’s administration is, to a great extent, dependent on perceived financial benefits, an accurate estimate of the number of employees needed to operate the system is required. The greatest financial returns are likely to arise from reduced labor costs. Unfortunately, the estimate of the number of staff needed to operate a robotized laboratory must be made before the laboratory has any experience with the system or its impact.
One of the first steps in the planning process is to decide which tests will be performed in the automated laboratory and which will be performed elsewhere. This decision requires not only an analysis of which tests are performed at each existing bench station but the proportion of tests requested stat vs routine per shift, the number of tests per shift, and the number of technologists working on each shift on each day of the week. With automation, it becomes feasible to combine the stat and routine workbenches for the high-volume tests, but for precise planning of staffing needs, the time of receipt of specimens in the laboratory must be considered. It is also necessary to consider physician needs in deciding which instruments should be interfaced with the robotized track and to assess whether greater operational efficiencies can be derived through modifying the test menus on different analytical instruments.
Selection of staff.
From the start of the planning process, it
is essential to assess the skills and interests of the laboratory
staff. This is especially important if the laboratory had been
previously oriented to a federation of separate specialty laboratories
because there may be a need for extensive cross-training of existing
employees. To capitalize on the skill-mix of the staff in the
laboratory at HUP, it was decided early to staff the automated
laboratory with a core staff who would be trained to operate all of the
instruments in the facility, and who would be supplemented by
additional staff from the specialty areas working in their areas of
expertise. By this approach, the laboratory would, for example, be able
to capitalize on the morphological skills of certain of the hematology
technologists who would be expected to operate only the cell counters
in the automated laboratory without having to become proficient in
operating chemistry analyzers. Likewise, technologists who had
previously been assigned to the toxicology, endocrine, and general
chemistry laboratories would not have to master the skills needed to
operate cell counters.
After a program was developed to assess the competencies of the candidates for working in the automated laboratory, it was "test-driven" on a small number of staff. The program had two purposes: for the technologists to assess their own adaptability to new technology, and for management to assess each individual’s potential for a successful transition to a cross-trained environment with new technology. Cross-training for the individuals selected to work in the automated laboratory was customized around each individual’s existing skills. The cross-training program incorporated not only training on unfamiliar instruments but also didactic instruction on the clinical significance of the assays that were new to them and interpretation of the results of these tests. Again, before all technologists were exposed to the program, it was pretested on a few selected staff and modified by their feedback before it was rolled out to all the staff.
The development of the automated laboratory required the laboratory to honor previous turnaround time commitments to the intensive care units, the transplant clinics, and the in vitro fertilization program, as well as adapting from a batch processing mode to a continuous mode of operation. Additional benefits arose through implementing autoverfication of test results wherever possible to maximize the efficiency of testing permitted by the automated laboratory. As a consequence, the median turnaround times for complete blood counts (CBCs) and high-volume chemistry tests between acknowledgment in the laboratory information system of the receipt of a specimen and its test results being accessible to clinicians is now 18 min for CBCs and 48 min for the chemistry tests.
Technical staff involvement.
It is important to have a
committee of technologists to look at all facets of work life,
including weekend/holiday policies and changes in work schedules. A
representative of the Human Resources Department became a member of the
planning committee once the decision to proceed had been made to ensure
that the interests of all of the technical staff were safeguarded. The
planning committee has been maintained after the system became
operational to review operations.
Because the staff came from several different laboratories, the senior management has worked with the management of the automated laboratory to hone their team-building/coaching skills and has met periodically with the staff on a shift-by-shift basis as well as on a one-to-one basis to ensure that the concerns of the technical staff are heard and addressed. The senior management had anticipated a prolonged shakedown period, although many of the staff had assumed a smoother transition and required frequent reassurance that temporary problems were to be expected. A technologist committee was created as a corrective action team to review and assess problems and recommend solutions.
The ROI for the project at HUP was based on the elimination of several positions. To minimize the impact of downsizing on the permanent staff, temporary staff were hired to fill vacancies for all laboratories, even those not directly affected by the automated laboratory, so that those staff displaced from the automated laboratory could be relocated to another laboratory within the Department. To facilitate recruitment and retention of these temporary employees, full benefits were provided to them. In reality, the number of positions that had to be eliminated was less than had been planned for because of a simultaneous aggressive program to accommodate tests from other hospitals and the Health System’s outreach program.
Beth A. Rockus, Health Network Laboratories, Allentown, PA, described the concept of project management as applied to the creation of a large-scale robotized laboratory. Project management is defined as the application of specific knowledge, skills, tools, and techniques to project activities to meet or exceed stakeholders’ needs and expectations from a project. Thus, it is a coordinated approach to the management of costs, quality, and time. However, it has only recently been applied within healthcare. Project management requires creation of a temporary team to plan, control, and manage people and other resources. One individual is appointed to lead the team and is given responsibility, accountability, and authority to manage the project to its completion. Eight areas of knowledge or function are involved in project management, and the project manager should have familiarity and some expertise in all of them. The primary areas involve the management of scope, time, cost, and quality. These are facilitated by the management of human resources, contracts, risk, and communications.
Areas of project management.
Scope management is concerned
with the definition of the project, the overall planning, and control
of changes. Time management involves definition of the necessary
activities, the sequence of steps, and the timetable for the project.
It is concerned with all aspects of scheduling and requires critical
path diagrams and/or GANTT charts as basic tools. Cost management
encompasses resource planning and cost estimating, budgeting, and
control. Quality management uses all of the tools of total quality
management to ensure that the outcome of the project will meet the
needs of the sponsor of the project. Human resource management ensures
the most effective use of the people involved in the project and
includes team formation and development. Contract management includes
the functions to acquire the supplies and services needed to complete
the project. Risk management is the function of identifying, analyzing,
and responding to risks. The project manager must manage his or her
communication skills so that all of the appropriate people are informed
about the state of the project at the appropriate time in the
appropriate manner, both orally and in written documents.
Phases of a project.
Each project has a life cycle comprising
several phases, which may have different requirements of resources and
outcomes. There is no single correct way to manage a project, but the
simplest approach involves the four phases of concept, development,
implementation, and termination. At the start of the concept phase,
there usually is only a fuzzy definition of a problem, but the
culmination of this phase is the go-ahead for the project. The
development or design phase usually is when the project manager is
assigned to the project and is the critical detailed planning phase.
Simultaneously with planning is the need to develop control procedures
to monitor and manage progress of the project. Two critical
considerations require the involvement of the people who must implement
the project and the recognition that many individuals working on a
project are not working on it full-time. The product of the development
phase is a project plan, which should not be considered rigid. The plan
must identify all the necessary resources and their costs within
± 5–10%. The costs must be appropriately allocated to the individual
work times and must be scrutinized with respect to adherence to the
overall project plan. For large projects, such as for installation of a
robot, costs associated with maintenance should be considered as part
of the overall project. Appropriate models for maintenance costs are
15–18% of the purchase price for software, 8–12% for interfaces,
15–18% for hardware, and 5–10% of the purchase price for the
service agreement for the typical instrument.
For large projects, it is useful to devise a work breakdown structure (WBS), which allows the project manager to identify the component tasks and to allocate resources to them. A WBS is the framework for the schedule and allows estimates of time and cost to be made. Computer software based on critical-path methods is now readily available to identify the longest path through the activities and to determine the earliest completion of the project. In spite of the most careful planning, there is the danger of "scope creep" associated with growth of the project beyond the initial plan. This is not necessarily a problem as long as the project manager controls the process; adjusts the schedule, cost, and quality appropriately; and communicates this to the stakeholders.
A second potential problem is poor risk assessment and failure to develop risk mitigation strategies. Risk management must be a continuous activity, and one of the most effective tools to manage risk is through cause-and-effect diagrams. Response to risk can be categorized according to whether the risk can be eliminated, mitigated where the probability of occurrence is reduced, or accepted. The latter requires the development of a contingency plan.
When the objectives of the project are met, it is terminated and its product transferred to an operations team.
Elkin Simson, Mount Sinai Medical Center, New York City, discussed the critical role of a laboratory information system (LIS) in an automated laboratory. Laboratory automation involves much more than a robotic system within a laboratory. The LIS in an automated laboratory is involved in both analytical and "peri-analytical" processes. The latter includes both preanalytical processes, such as the processing of physicians’ orders and specimen accessioning, and postanalytical processes, such as result verification and report generation. The LIS provides patient data to the analyzers, facilitates quality control, and receives, verifies, and stores results and transmits them to the requesting physician. In an automated laboratory, the functions of the LIS must be integrated with the functions of the robotic central processing unit (CPU) and the robotic controllers. The LIS identifies each specimen placed on the robotic system and instructs the robotic process controllers where to send the specimen, and to track the specimen and its aliquots wherever they might be within the system. It receives and stores data from the robotic system regarding the quality of each primary specimen and the residual volume of specimen in the aliquot tubes so that specimens may be retested as needed.
Operational concerns.
It is desirable for tests to be ordered
directly into the LIS. Not only does this reduce errors, it also has
the potential to improve in-laboratory turnaround time. Printing of
bar-coded labels close to the patient’s bedside also reduces errors
and facilitates testing. Furthermore, it enhances patient
identification and provides an accurate time of specimen collection.
Within the laboratory, batch downloading from the LIS to the robotic
CPU facilitates information transfer and minimally impacts the speed
and system capacity. However, such an approach requires clearing or
deleting of specimens for which tests were ordered but not placed on
the robotic track. Real-time querying of the LIS provides considerable
flexibility in testing and enables reflex specimen testing. Automated
sorting of different types of specimens on the robot track adds
complexity but removes the need to manually sort specimens. Automated
sorting adds cost and may also create a rate-limiting factor in the
testing process because all specimens must pass through a single point.
An automated laboratory is critically dependent on a LIS and its vendor. Back-up hardware, software, and system procedures should be in place to limit downtime attributable to failure of some part of the system. An uninterrupted power supply (UPS), backed by an emergency power generator, is essential to minimize the impact of surges or drops in voltage. The LIS should have a minimum of two CPUs that usually share the workload, but with each one capable of handling the entire workload. Constantly updated mirror databases are also needed to provide backup in case one become corrupted or fails. Extra functional ports to and from the CPU to peripherals and instrument and other interfaces should be available for rapid switching without shutting the system down if one or more connections fail. Redundant printers and terminals should also be available.
Back-up in the event of failure.
Each night the system
database should be backed up to magnetic tape. This should be done in
the "background", at the same time permitting normal operation of
the LIS in the foreground. All changes and upgrades of the LIS must be
adequately documented with dates and details of the changes. Before a
change is made, it should be tested extensively in the
"test" portion of the LIS before being upgraded to the
"live" part of the LIS. When a problem occurs with the LIS, the
laboratory LIS staff should attempt to fix it, but then should enlist
the LIS vendor for assistance if necessary. The same procedure should
be followed if a complete failure occurs. The LIS staff should provide
the laboratory staff with an estimate of the likely downtime so that
alternative procedures may be implemented. In the event of a complete
failure, the medical staff must also be notified. When LIS
function is restored, this should be communicated to all parties in the
same manner that the failure was communicated.
The papers summarized above, when taken together with the full-length papers that follow, will provide the reader an excellent overview of the status of current thinking and applications with regard to laboratory automation.
|
Footnotes |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
