Melding brain with brawn
Much of the knowledge transferred into the LIS resides in the laboratory staff and the various consultants that help to set the standards the lab uses for delta checking and critical values. An essential facet of managing a lab today is working with an LIS vendor. CCH uses LabWeb LIS software, and we work with the CCH vendor, Psyche Systems, to address new areas of functionality and continually advance the system in the appropriate directions.
One way to accomplish this is to work through a series of what-if scenarios similar to the case studies used in personnel competency testing. Often, an LIS vendor has trained medical technologists on staff, experienced in various areas of product development, testing and training processes, and a flexible, customizable system. My counterparts at the LIS vendor work with me to extend the functionality of the rules system into new areas, such as instituting flags beyond the reference ranges.
Another way is to ensure that laboratory staff is exposed to an active continuing education (CE) program. At CCH, a lab education coordinator alerts staff to new in-house materials. Our hospital requires staff to complete 15 CE hours yearly, and tests competency annually, as well. Testing staff members critical-thinking skills can help them better assess a real-life patient scenario.
Reviewing manual procedures
Before the LIS can help enforce proper procedure, a review of the manual processes responsible for inputting the correct data into the systems is required. The blood bank is one area of the lab where it is essential to check, recheck and triple check specimen identification, processing and resulting. Accuracy of patient identification when taking blood samples was one of six patient-safety goals established last year by the Joint Commission on Accreditation of Healthcare Organizations.
Ensuring accuracy in specimen collection and processing begins with an important system performance check, a simple cross-reference of the patients blood type with existing information that resides in the directory or card file to ensure that incompatible blood types are not mixed. A flag is raised if an attempt is made to cross-match units of a different ABO group. Such a simple check could prevent a potentially very serious and costly error.
For the year ending in September 2002, the Food and Drug Administration reported 85 transfusion-related deaths of recipients nationally and 10 of donors. Of these recipients, 14 deaths were attributed to blood incompatibility caused by hospital errors. Balancing human expertise and computer-aided checks and balances can provide a source of security if implemented properly.
Building rules-based analysis
While the instruments have built-in linearity checks, this system of checks and balances is dependent on the rules set in the LIS and the specific triggers built into the test menu. These are determined by a talented team: our pathologist medical director, attending medical staff, laboratory staff and consultants. LIS-supported, rules-based analysis is a powerful tool in minimizing errors in test resulting.
CCHs LIS allows us to set up rules for both resulting and ordering tests. It can be configured to prompt users according to test preparation and transport requirements, or to remind techs that for certain tests informed consent must be obtained prior to collection. Based on information specified by human expertise, the information system can be configured to alert the user when additional testing is required. If a result is elevated or abnormal, the computer will instruct the user to perform confirmatory testing. The software can even perform reflex ordering of those confirmatory tests.
The delta and critical value limits set for tests, used by all areas of the lab, are the first lines of defense from the standpoint of treatment and response. These limits have established ranges, automatically flagging results that fall outside of them. These limits help identify significant changes in test results generally within a 24- to 48-hour period on repeated tests that, under normal circumstances, would not occur, and raise a flag for a technologists verification of patient status. Typically, when a critical value is reached, the technologist must immediately alert the appropriate medical personnel who can coordinate any specific orders for the patient in response to the critical value. Good documentation of critical-result handling is part of the test result, provides a record of how the result was handled and minimizes potentially egregious errors.
Hematology
Delta failures are fairly frequent in hematology (i.e. a change in white count, or a drop in hemoglobin) and can be valuable in flagging platelet-count problems in addition to which, here, they are most valuable in the first 24-hour time frame. Recently, such delta checks helped CCH identify hemolytic uremic syndrome (Ecoli 0157) in a child admitted to the emergency room. Initial baseline complete blood cell count and chemistries had not shown any apparent abnormality, but repeated tests after eight hours showed a significant drop in hemoglobin and platelets, which raised a critical value flag, along with a marked shift in blood urea nitrogen and creatnine, which raised a delta flag. These checks prompted the lab staff to alert the childs doctor who was then able to treat the child with dialysis.
Chemistry
By contrast, in chemistry, unless a patient is very ill or dehydrated, many chemistry values will remain stable. For most chemistry tests, a 20% change within 24 hours of a previous result raises concerns and indicates a need for further investigation or action, such as repeating the test, checking the patients condition, speaking to the phlebotomist who drew the sample or alerting the physician.
At CCH, if a patient comes in with a normal creatine kinase (CPK) or normal troponin level, which elevate during one of the cardiac profile work-ups performed every six hours, a significant change will be apparent, and the results will help the doctor understand the patients condition. We consider a normal CPK level set at 79 or less in a patient presenting with a suspected cardiac injury. With a result greater than 79, we perform reflex testing.
The response varies according to the type of test flagged in the system. The key to successfully using this information lies in the ability of staff to proactively take whatever that step might be. Such delta checks keep clinicians informed of changes. In making rounds, they may not have time to carefully study all lab reports unless they are looking for something specific. The lab must proactively contact the doctor or investigate delta changes, retesting when warranted.
Microbiology
In microbiology, many guidelines concerning antibiotics, for example, are based on research literature, such as that of the National Committee for Clinical Laboratory Standards, which we use to define the rules. The key function of the rules-based analysis in this area of the laboratory is in the checks and balances that help identify certain organisms present in a particular specimen. The rules help keep track of the criteria associated with various organisms, as well as how the antibiotics react to certain organisms. Appropriate use of antibiotics is crucial in preventing, or at least slowing, the development of resistant strains that can lead to infections that can no longer be treated effectively, such as the methicillin-resistant Staphylococcus aureus.
Enabling standardized antibiotic testing in conformance with established recommendations allows the lab staff to remain current and minimizes the risk of inadvertently creating resistant strains by testing organisms with an inappropriate antibiotic that may suggest approval for use by the doctor. Similar to delta limits in hematology, certain reactions can trigger built-in flags that require a specific follow-up action on the part of the laboratory staff.
Quality assurance processes
Rules-based analysis is not to be confused with quality control (QC). The QC rules are frequently already in the system, but the processing and resulting rules must be, again, determined by the knowledge of the medical staff. QC varies by instrument and test, and works to ensure that the instruments and reagents are running properly. Like most labs, CCH has standard QC practices. We use the QC in the LIS for the chemistry area due to the volume of tests performed (about 47% of our total volume), whereas QC for hematology resides within the instrument itself.
A control that falls outside the set limits must be addressed before any patient results are released. In that sense, the whole QC function is a rules-based system with strict criteria. These are run in conjunction with all patient testing to ensure that an abnormal result is, in fact, a patient test issue and not a test-system issue.
Through our quality assurance processes, CCH has monitored its critical value follow-up documentation since 1990, at which time the hospital achieved an 80% success rate in appropriately responding to and documenting incidences where critical values were met. By the end of 2002, the CCH success rate had reached 99.3%, even as the number of critical values monitored had more than doubled to 2,055. The real benefit of a robust LIS is not only in improving the reporting and turnaround time, but also the accuracy, the follow-through and the quality assurance documentation.
Planning for the future
The expected shortage of highly qualified personnel could force many laboratories to depend on first, elaborate instrumentation and equipment to aid in performing the testing and second, on computer systems to help assure that the test results are optimal. Just as the science of laboratory medicine advances and the laboratory staff continuously learns about new tests and techniques, the LIS system cannot be static. With these major ongoing changes in healthcare, it is the lab managers challenge to blend and adapt knowledge and technology to the increasing complexity of the laboratory environment.
Judy Robertson, MT(ASCP) is the laboratory manager at Columbus Community Hospital in Columbus, NE.
November 2003: Vol. 35, No.
11
© 2003 Nelson Publishing, Inc. All rights reserved.
