January 7, 2003 Presentation:
Bridging the Pre-Analytic and Post-Analytic Operations
Welcome to the AACC Expert Access Live Online program.
Our topic this month is Bridging the Pre-Analytic and Post-Analytic Operations. The field of lab automation has recently witnessed a surge of new products and concepts as vendors try to meet the demands of laboratory customers looking to gain the advantages of automated sample processing and testing. However, automated solutions for the laboratory remain expensive and may not be the best answer in every facility. Join us today as Leo Serrano, FACHE, CLSup(NCA) , Executive Director of Laboratory Services at West Tennessee Healthcare (Jackson, TN) , answers your questions about evaluating lab processes to determine what degree of automation is right for you, and when an alternative solution may your best bet. View the presentation and direct your questions to this month’s online expert, Mr. Leo Serrano.
AACC would like to extend its most sincere thanks to the Bayer Corporation for making this program possible.
In your opiniuon, what is the future of lab automation and say, handheld devices?
Leo Serrano, FACHE, CLSup(NCA): POC handheld devices are a useful adjunct to a total automated solution. They are not mutually exclusive. If you have a cradle to download the information, it can augment the automation providing a more complete and robust approach to diagnostic work. Additionally, handheld barcode readers/label printers enhance the accuracy of collection/identification of specimens thus making an automated line even more reliable and efficient.
1)In case of highly icteric & lipemic serum sample how to avoid flaws i.e. false low reading due to slight hazy assay system in estimating Billirubin,SGPT. 2)In case of glucose(true glucose estimation is it apt & correct to express the value essentally from whole blood glucose or Plasma glucose. 3)how accurate are the results expressed in serum glucose. Thanks
Dahanu Road,District Thane,Maharashtra
Leo Serrano, FACHE, CLSup(NCA): 1)A sample integrity checker is an important adjunct to automation. Whether visually (by a human) or better yet, automatically (preferably independent of the analyzers); the integrity checker would address the issue of identifying lipemic or icteric specimens. This will identify the sample, whether or not you must take an intervention depends on the analyte, the analyzer and the method. No matter how automated, the clinical judgement of the laboratory professional is still the key to quality lab work. 2) I am not sure of the point of the question here. If you mean the relationship of whole blood glucose to plasma or serum glucose, this is very similar to the issue of whole blood alcohol vs. serum or plasma alcohol levels, it is a matter of volume displacement by formed elements so there is an issue of relative volume. As long as the clinician is aware of the method and the implications of the specimen, they can make the necessary adjustments in treatment or diagnostic considerations
Are there any automated systems that work well with small volume samples as in pediatrics, or neonates? Vijay L. Grey
Leo Serrano, FACHE, CLSup(NCA): There are a number of "micro" sample analyzers on the market. When I think of laboratory automation, I consider the pre-analytic as well as the analytic issues. Micro samples are harder to process but I feel several of the current systems can be adapted to an "all micro sample environment".
Our site is a 600 bed not for profit hospital with an active Outreach Reference lab. Most of our Outreach specimens have to be manually logged in, registered into computer and tests ordered in Cerner Millenium. Outreach is about 40% of work. How would a preanalytical system help our output without adversely affecting turnaround times. Presently chemistry technologists centrifuge and aliquot specimens. Our lab is expecting a growth in the Outreach arena and I cannot afford to lose technologists to cost justify a preanalytical system.
Baton Rouge, LA
Leo Serrano, FACHE, CLSup(NCA): The key to selecting automation is to first evaluate your processes. You do not justify automation solely by eliminating technical FTEs. The first thing is to totally evaluate the mechanism you use for your outreach. Can they order electronically at the client site? Is internet or VPN an option for them? If you must use paper requisitions, can they be "read electronically" using a scanner or similar application? Are the specimens labeled using a barcode label at the collection site? If specimens must be ordered and logged in at your site, can you streamline the process? Can you standardize to some level, the types of tubes you receive? Why have techs centrifuge and aliquot- use a pre-analytic module to do that. Do you really need to aliquot or can you better distribute your work to more versatile instrumentation? Laboratory professionals should be used for those tasks that best utilize their scientific skills and leave the repetitive mundane tasks for machines. Use your techs to generate additional revenues by doing new tests and increasing test menus.
We just recently automated the laboratory and now am in the process of re-analyzing the workflow and FTE requirements...how does one assess on a numbers/quantitative level, not just qualitative/observational, workflow and productivity of each tech to decided whether automation can indeed allow for improved efficiency and perhaps decrease the number of FTEs needed to run that portion of the lab... Thanks Wes Kim MD
Leo Serrano, FACHE, CLSup(NCA): When you choose to do a workflow analysis, it should be a measurable and quantitative function, not just an observational function. There are a number of excellent programs available. For example, we had Sysmex do a workflow analysis for our hematology section. They provided us with quantitative data that validated the independent analysis we had performed ourselves. Currently, we are implementing the Ortho Clinical Diagnostics "Process Excellence" program. It uses "LEAN, 6 Sigma and Design Excellence" programs to evaluate workflow and eliminate waste, duplication and errors. In our case, the lab has become the "mentor" for process review for the rest of the Healthcare System. I would refer you to Sandy Hood at Grant-Riverside Hospital in Columbus for additional information.
We are beginning the process of picking new chemistry/IA/TDM platforms and are also considering automation. The new instrumentation seems to connect together to the point that a single tube, loaded once onto the system of instruments, will sample for 70% of the labs testing. The new instrument systems seem to be doing the automation. With 70% of our samples arriving processed, the only good I can see of automation is a end product racking system. Do you disagree? If you are getting new instrumentation and want automation, which do you do first?
Albuquerque, New Mexico
Leo Serrano, FACHE, CLSup(NCA): Which came first, the chicken or the egg? You are correct in noting that many systems today can perform 70% or more of the tests from a single sample. That is a form of automation. Track movement of samples from one analyzer to another is not the only form of automation. I find that the pre-analytical function can provide a great deal of savings (centrifugation, sorting, etc) and then a specimen can be "sneaker-netted" to the analyzer network. That is why I advocate "open" systems.
did you automate? with what?
Leo Serrano, FACHE, CLSup(NCA): We did several workflow studies before automating. In our case, automation is an ongoing and evolving process. We have a Sysmex HST 302XE system for hematology processing. We have an Ortho/LabInterlink System for chemistry, immuno-chemistry and coagulation. These three areas handle their samples in the same manner (centrifugation)thus lend themselves to automation. Our analyzers are OCD Vitros 950s and Vitros ECi as well as a sysmex CST 100 coagulation module. All of this is evolving. We use the Lab Master Organizer (Backend) to sort and archive. It sorts for other departments as well as sendouts and also archives the specimens for retrieval as needed.
From your presentation it sounds like labs would be better off doing work flow analysis and changing to a more efficient process as a result of the analysis regardless of whether they decide to automate part of that process or not. I’ve heard that people are expensive and automation is a money saver. It almost sounds like you are saying workflow analysis is the money saver. Do I have this right? If so can you provide some real lab examples? If not please correct me.
San Diego, CA
Leo Serrano, FACHE, CLSup(NCA): It is imperative to eliminate as much waste and duplication of effort BEFORE investing in automation. At that point, you can invest WISELY in automation that will benefit your organization the most. People are expensive and cost more each year. Automation is likewise expensive. It simply can be depreciated over time. If you automate without first learning where you need the help, you are spending money un-necessarily. We changed our workflow patterns THEN automated the areas that were "people intensive" so that we were able to re-allocate the staff to generate other sources of revenue. In the process of automating after a workflow analysis, we reallocated 11 FTEs in our lab and added new services. Our volumes increased by over 30% with the same levels of staffing. The staff simply were doing other tasks.
What is "sneaker-netting"?
San Diego, CA
Leo Serrano, FACHE, CLSup(NCA): "Sneaker-netting" is the term we use for walking a specimen or any item from one location to another.
Can you give me a range for low volume laboratory vs Moderate or High? This will help us give a fit for our lab .
Leo Serrano, FACHE, CLSup(NCA): These are subjective. Moderate volume is 1000 samples per day, High volume is 3000 or more samples per day. This is just my opinion. Others may differ.
Can you define the key preanalytical, analytical and post analytical steps and percentage of labor found within a typical hospital laboratory.
Leo Serrano, FACHE, CLSup(NCA): In our work, we found that approximately 60% of the labor is in the pre-analytical phase of testing. This varies with your definition of "preanalytic". Likewise, depending on your definition of post-analytic, it can represent between 20 and 30% of the labor. The actual testing on an analyzer is between 10 and 20% . This will vary slightly from one lab to another.
US Bancorp, Piper Jaffrey presented (in 1999 at AACC automation conference in Philly) a breakdown of a typical budget allocations for the laboratory. 65% Labor 20% Overhead 15% Equipment & Reagents Do you find the same distribution/allocation in your experience?
Leo Serrano, FACHE, CLSup(NCA): Our experience is similar but again, that varies from lab to lab. We have a heavy labor $ component because of the benefits and the premiums we pay our staff. Our overhead is not as high but that is a function of our organization. Clearly, labor is always the largest component of the modern laboratory. The % is the variable in the equation.
In your last slides you refer to the choice of automation as favoring open systems rather than just the analyzer...as a general question what do you define as an open system? How much does analyzer quality and throughput play a role even if it may not be as open a system as another...finally does automation really save money!...
Leo Serrano, FACHE, CLSup(NCA): To me, an open system is one that is independent of the analyzer; that is, will work equally well with any manufacturers analyzer. If the automation requires you to use the same manufacturers analyzer, it is not open. You want to be able to change analyzers without having to change the pre-analytics, delivery system and archival system. It is important to weigh the total solution when making a decision. A fast analyzer that is of inferior quality will not help your cause. An automation system needs to save money by allowing you to use your high $$$ techs to generate new revenues, not perform repetitive, boring tasks that can be automated. In our case, we have demonstrated an ROI in 22 months- from that point on, we were saving money. Our volume has increased over 30% and the people who were re-allocated, are performing testing we used to send off.
For an auto immunity lab, to treat IFA slides, Elisa plates or for instance a Multiplex microplate, would you use separate instruments, or would you prefer to have a complete front end system able, from the same primary tube to dispense sample and reagents into the 3 tests simultaneously ?
Christine Flandre, Biomedical Diagnostics, Paris France
Leo Serrano, FACHE, CLSup(NCA): Because of the cost of front end processing pre-analytic instrumentation, I would look to some of the available robotic pipetting/dispensing platforms available today. So many of them can process slides, tubes and microplates that extensive research is necessary to find the right fit.
We are in the processing of automating our front end operations but are running into issues with our current bar code symbology 39 versus the vendors preference for 128. What are the pros/cons of either symbology and the options or transition issues to convert to another symbology
Leo Serrano, FACHE, CLSup(NCA): This is an issue we find throughout healthcare. We encountered it with several instruments. It amazes me that the food industry can agree on a standard barcode symbology so that you can have a UPC label on food from a large variety of vendors and yet in healthcare, we can't get there. At our facilities, we have standardized the lab on code 39 with check digit. We require all vendors to accept that symbology. We had no trouble with our automation vendors accepting code 39. We acknowledge however, that some vendors will not be able to accommodate that and in that case, they will not be used in our labs. I really have no idea which is a superior symbology, Code 39 was available in the vast majority of the instruments we had so it was a logical choice.
On slide 9 you referred to "sneaker netting" when volumes are modest. What is this? Since I work in a children's hospital volume is always very modest.
Leo Serrano, FACHE, CLSup(NCA): In this slide, I should have clarified that this was not sample volume but number of samples. Sneaker netting is our term for walking specimens from one location to another.
1. What is your opinion about tubes with separation gel (Sarstedt) for immunoanalyses of hormones and tumor markers including storing? 2. Would you prefer on line or off line connection between pre-analytical and analytical phase? 3. What is the temporary trend - aliquoting and parralel analyses of samples or complete analyses from primary tubes?
Brno, Czech Republic
Leo Serrano, FACHE, CLSup(NCA): 1. Gel tubes complicate the pre-analytic and direct sampling aspects of an automated robotic track. As far as storage of specimens for hormones and tumor markers, this is probably not a good idea but it would be dependent on each manufacturers guidelines. 2. We have an on-line connection between the pre-analytical and analytical phases. In our case it is more of a total solution as opposed to partial solution. 3. The decision to aliquot is dependent upon your workflow and layout. We chose not to acquire an aliquotter because we could route our samples to the various areas. The number of aliquots is critical to determining which way to go. We have chosen to do as many analyses as possible from primary tube sampling.
What is the most common method for intra-operative PTH (for parathyroid surgery)and is it performed in or near the OR or in the central lab?
Leo Serrano, FACHE, CLSup(NCA): This question is really not related to automation. I would imagine that an immunoassay method is the most common method. Where it is performed is dependent on the lab, the need and the decision of the Medical Director, Clinical Chemist and Medical Staff.
Can you recommend a time period to recoup capital outlay when evaluating the cost effectiveness of a pre-analytical automation system vs the decrease in FTE cost?
Leo Serrano, FACHE, CLSup(NCA): This really depends on the financial needs of the institution. As a rule, our organization likes to see an ROI of 36 months for analyzers or related materials. In some very large capital purchases, the ROI can be longer. For our Automation project, the ROI was 22 months. We like to use FTE avoidance as the measure since we re-assign the "freed" FTEs to other areas of the lab to generate additional revenues.
In Nov. CAP Today, an automation article stated one institution's bad experience with an automation vendor that was unable to develop an interface to their equipment. We are evaluating a vendor who has yet to develop the interface to 2 of our instruments. How concerned should we be about the ability of the vendor to develop interfaces as promised?
Leo Serrano, FACHE, CLSup(NCA): Based on personal experience, I would be concerned. The delays in developing instrument interfaces cost money and affect the ROI for the system. I would suggest that the payment for the systems be conditioned upon delivery of the interfaces for the instruments. We did that and that has been the leveraging point in our dealings with the vendor. If the vendor has your total moneys up front, you have little leverage to make them complete the tasks. Monetary penalties work.
What do you see as the autoamtion challanges that will confront the Laboratory in the next 5 years?
Leo Serrano, FACHE, CLSup(NCA): The ability to automate more complex instrumentation and analyses will help sort out the mix. Making automation more affordable will also be a major effort. It is hard to justify million dollar plus expenses for automation during a difficult economy. As automation matures, the product evolution will address many of our current concerns.
Any suggestions about handling Legal specimens like ETOH and drugs. If the bench flags the the specimens as collected/received, there is no documentation of the actual phlebotomist. Has this been an issue?
Mt Clemens, MI
Leo Serrano, FACHE, CLSup(NCA): Forensic specimens require extensive documentation. We do not handle our forensic specimens in the same manner as our clinical specimens due to the chain of custody issues, etc. We document the actual phlebotomist on all collections (clinical as well as forensic). If you do a significant amount of forensic work, I would suggest doing it in a distinct "forensic" environment.
In 1999 at AACC conference in Philly, data was presented from U.S Bancorp, Piper Jaffrey on the typical laboratory budget allocations are as follows: 65% labor 20% Overhead 15% Equipment and Reagents Obviously, automation can be in terms of 'robotics devices' but I can be in other areas such as IT (autoverification, etc...) Questions: 1) Are the % consistent with your experience? 2) What % of the labor would you attribute to results production (pre/anlytical/post)? 3) What % to result review and reporting?
Leo Serrano, FACHE, CLSup(NCA): As referenced in a previous question, the percentages may vary from place to place. IT is critical. However, you can have autoverification without an automated track system. It can be a part of the LIS or the analyzer.
What impact has lab automation had on your TAT for Chem, IA and Heme tests - and do you attempt to handle stat and routine samples differently?
Leo Serrano, FACHE, CLSup(NCA): Our TATs improved dramatically as a result of the simultaneous implementation of improved workflow and the automation processes. If you control the specimen from collection thru resulting, you have an advantage. Most labs have control inside the lab. We initially intended to handle stats and routines in the same manner. We now find that it depends on how much volume we have in the preanalytic modules. When we aren't processing large amts of outreach work, they can be handled the same. If you have large amts of routine work, it can tie up the centrifuges and delay the stat to some degree. The number of available centrifuges does play a role in how you chose to handle this.
In preparing to install your lab automation what was the single most important issue you analyzed?
Leo Serrano, FACHE, CLSup(NCA): Workflow. Workflow and processes are the single most important area to evaluate before you invest in automation. Automating a bad process is the same as shooting yourself in the foot.
What improvements did you see with the HST?
Ft. Worth, Tx.
Leo Serrano, FACHE, CLSup(NCA): The HST system allowed us to do significantly more work with less dedicated staff. The quality is not an issue with the analyzers, the issue was getting the specimens to the analyzers and stainers. Another issue was in having slides made and stained selectively. It made no sense to have slides made on every case, we only make them on those cases that meet selected criteria. Our TAT improved dramatically with the HST. The processors can load the racks and the techs simply place the racks on the analyzers and review the results or they are autoreleased as appropriate. The system allowed the lab professionals to do what they do best, make decisions about the quality of the results.
What are the sample challenges when utilizing an automation system (analytical and processing) and what recommendations might you have for improvement?
Leo Serrano, FACHE, CLSup(NCA): we found that the variety of sample sizes as well as the placement of barcode labels created opportunities for us. We can handle a variety of tube types but gels and large tubes (16x125 mm) created headaches. We also had issues with the micro or bullet tubes. We have limited the types of tube sizes only to the extent that we don't accept the large tubes, gel tubes and micro samples for automated pre-analytical work.
In your opinion, can you name 3 vendors that have the most mature pre-analytical and post-analytical automation system suitable for a 600 bed hospital with work volume of ~2 million tests per year?
Leo Serrano, FACHE, CLSup(NCA): I hate to name vendors since there is no ideal system that fits everyone. I would look for a vendor whose product has been around for a while (several years), who has multiple installations similar to yours and who can meet your needs. My only prejudice is in wanting an open system that is not limited to a single manufacturers instruments.
How will automated centrifugation impact my existing workflow? What factors should be considered when weighing automated centrifugation vs. manual centrifugation. I am concerned that I might create a bottleneck with automated centrifugation?
Leo Serrano, FACHE, CLSup(NCA): The real issues with centrifugation lie in the "lost" time. I define lost time as the time the tubes sit in the centrifuge after it stops and before it is unloaded and the tubes moved to the analyzers. An automated pre-analytic system can significantly eliminate "lost" time and that is a real advantage. In our case, we were equally concerned until we measured the "lost" time and realized that the bottleneck would be eliminated by using the automated centrifugation. I will say that two centrifuges are better than one when it comes to automated centrifugation.
In view of the capital expense of a total automation system, we are investigating a stepwise approach into total automation over a number of years. We would like to first automate our pre-analytic processes. Second would come the stand-alone Chemistry analyzers, then the track to link the Analytic and Pre-analytic systems, and finally the automated archive/storage system. Unfortunately this plan seems to eliminate most current vendors, with less chance of competitive bids. Is this a the preferred way to bring in automation, and should price be a strong consideration in the vendor?
Leo Serrano, FACHE, CLSup(NCA): The method you describe is clearly the most logical. The vendors all want to sell you everything at once. I know that the vendor we use, can provide the step-wise integration you describe. Both the Beckman Power Processor and the OCD EnGen are pre-analytic modules that can be built upon. Price is very much a consideration. Like all new "toys", automation costs more in the beginning but as the products mature, the price comes down. The field is evolving rapidly. I would look at how long the vendors have been around, how many systems they have installed in similar environments. These all are points to consider.
You mentioned you did several workflows prior to deciding on what to implement. Can you clarify if you did them or a diagnostic vendor did it and if so did they all suggest their solutions for implementation.
Leo Serrano, FACHE, CLSup(NCA): I did my own workflow studies and then had manufacturers do their studies. In each case, our results were comparable. In some cases, the vendors suggested their solutions, in others, they made general recommendations and gave us a list of possible solutions. Of course, their solutions were always just a bit better. However, we didn't always use their solutions.
1)What do you see as the major advantages in combining chemistry/IA in the core laboratory? 2) How does this differ with a blended system versus multiple systems combined by a common track interface?
Leo Serrano, FACHE, CLSup(NCA): We got the most "bang for the buck" by combining both Chem and IA in the core lab. We did not find a blended system that met all of our needs so we used multiple systems combined by a common track interface. This depends on what your individual needs are and how you choose to address them. There is no single right way. The right way is what works for you in the most cost efficient, high quality manner.
That wraps up this session of Expert Access Live Online. Thanks to everyone who posted a question for this month’s expert, and thanks also to our expert, Mr. Leo Serrano. Unfortunately, our allotted time frame wasn’t nearly enough to get to all your questions. If you posted a question that didn’t get answered online, Mr. Serrano has graciously offered to field your inquiry if you re-send it to his e-mail address: firstname.lastname@example.org. We hope you found today’s presentation and Q&A segment to be informative and useful. This Expert Access session, and all previous sessions, are archived on our website in order to serve as a continuing source of education.
AACC would like to again thank the Bayer Corporation for making this educational program possible.