Common laboratory errors by discipline that can be incorporated CDS tool development
Introduction
As the laboratory results play an important role in diagnostic decision-making process and the usage of clinical decision support(CDS) tool is gaining its popularity, it would be helpful to recognize common laboratory errors and incorporate them when designing and developing CDS tools for computerized provider order entry(CPOE). Considering well-designed CDS tools can yield appropriate treatments in response to abnormal lab results, the opposite may be true. If results are inaccurate, it may lead to inappropriate treatments. Therefore, developing CDS tools that can prevent and detect common lab errors can add a value to utilizing CPOE and EHR system to its potential.
The majority of errors occur in the pre-analytical phase, and bleed into analytical and post-analytical phase. Not only do they lead to a waste of time and resources but also possibly cause a delay or harm in patient care if not caught in a timely manner. Therefore, it is important to catch them in analytical and post-analytical phases prior to clinical decision-making process. Ideally, both laboratory personnel and the treatment team have an insight and knowledge to recognize those errors. Utility of CDS tools can reduce errors if developed in such a way that would alert clinicians and prompt them to recollect and reorder and lab personnel to self-check before reporting.
There are three representative pre-analytical, analytical, post-analytical critical errors throughout the disciplines that have the similar consequences. 1. For pre-analytical phase, un/mislabeled samples. 2. For analytical phase, sample/aliquot/sub-plates mix-ups. 3. For post-analytical, misinterpretation, manual transcribing/calculation error and failure to recognize and communicate critical values especially during the software downtime/malfunction. Below are descriptions of errors that have different originations and characteristics and listed by discipline.
Discipline
1. Chemistry
1) Contamination: The sample is most commonly mixed with TPN. Often IV hasn’t been turned off and drawn from the same side of arm and/or the line hasn’t been flushed enough. It is often indicated by unusually high potassium and glucose values.
2) Dilution: The sample is often diluted with saline and fails delta check if a lab has applicable protocols. Testing personnel(i.e.MT and MLT) may request redraw upon comparing with previous values on relatively stable analytes such as calcium or total protein.
3) Wrong Order of draw: When a hematology(lavender top) or coagulation(blue top) tube is collected prior to chemistry tube(e.g.gold top or green top tube), calcium is often critically low and/or potassium is critically high due to chelating.
4) Drug level: There are occasions where peak and trough draws are switched. Precautions such as careful review of the values(i.e.trough value should be lower than peak) and confirming with collecting personnel, if any questions, should be practiced prior to reporting. It’s one of common post analytical errors especially with therapeutic drug monitoring.
5) Testing personnel error: With results that are above or below linearity, instruments usually have a flag or alpha response, not typical numeric values, because the result surpassed its reading range. Without paying close attention, misreading might occur between below or above linearity. The consequences can be severe if reported erroneously especially with drug values. Repeating the test to rule out errors from insufficient sample/reagent volume/bubbles and secondary check of the flag might be useful prior to reporting to prevent error.
6) Dilution error: Analyzers lack an ability to perform on board dilution for each analyte and often diluting capacity is limited. Dilution can be a source of many errors-testing personnel may choose a wrong diluent. They may erroneously perform dilution or miscalculate total dilution. Since the testing volume tends to be limited (within 1mL) and dilutions are performed with pipettes, there is a high chance of not mixing the sample/diluent mixture thoroughly or introducing bubbles. If an instrument aspirates bubbles, it would give an erroneous value. If the dilution is too high, the result is less likely to be accurate-where 1:100 dilution is performed when 1:10 dilution would be appropriate. Often enzymes such as CK, lipase, or amylase have a high value so a close comparison with the previous results might be useful to determining dilution factor.
2. Hematology
1) Contamination: It is easier to recognize in chemistry results than hematology. Results from redraw often remain relatively comparable. They often affect MCV and hemoglobin/hematocrit. Any questionable results should be interpreted concurrent with chemistry results if both are ordered.
2) Dilution: A sudden significant drop in hemoglobin/hematocrit and platelet are indicators of dilution without plausible reasons such as postpartum or operations. Erroneous reporting that doesn’t get caught by labs or clinical care team members may lead to unnecessary blood product transfusion, causing unfavorable outcome.
3) Review error: Blasts/abnormal cells can be hard to recognize among inexperienced testing personnel. Cell differential results should be closely examined in association with hemogram results to avoid error. With white blood cell count of >100x10^9/L, normal differential results should be questioned.
4) Failure to detect a microfibrin: Analyzers don’t always flag microfibrins, resulting verification event without further workup. It is very common to see a microfibrin in a pediatric lavender tube due to the nature of collection and difficulty to mix blood properly in a smaller container. It significantly affects a platelet count, therefore, extra caution should be practiced when a sample collected in a pediatric tube prior reporting. It may be beneficial for a clinical team to recollect a sample if the result is questionable before developing treatment plans.
3. Coagulation
1) Improper fill: A blue top tube used for coagulation requires a discard tube to achieve a proper volume. It cannot be undefiled or overfilled.
2) Wrong order of draw:PT and aPTT result may be erroneously prolonged if EDTA tube is drawn before the blue top tube.
3) Contamination: A sample is often collected from the same side where patient is receiving fluid and it often leads to the erroneous critical result if not collected from a different side or flushed long enough. Often aPTT is critically low if contaminated with heparin.
4) High hematocrit: When hematocrit is >55%, it has a dilutional effect. If a patient has a history of high hematocrit, the anticoagulant volume needs adjusted prior to collection for the accurate result. Before recollection, the sample should be retested after thorough mixing to avoid erroneous reading from hemoconcentration.
4. Blood bank
1) Review error: Antigram cross out technique may be deficient (e.g.not ruling out cells homozygously when indicated), resulting in overlooking a clinically significant antibody.
2) Procedure error: When manual techniques are involved, the steps may not be followed correctly(i.e. wrong incubation time, wrong centrifuge time, improper washing technique, erroneous reading of agglutination). It might lead to failure to detect an antibody or antigen, possibly resulting in a transfusion of incompatible blood products.
3) Storage issue: Blood product wastage has been a common issue, especially in massive transfusion protocol. Blood products require different storage conditions(i.e. temperature, constant rotational requirement for platelet)and they often get mixed together during MTP, not suitable for transfusion or reuse. Raising awareness for the proper blood product storage can reduce wastage and guard safe inventory level for emergent situations.
4) Failure to check patient’s history: Bone marrow transplant patient may have discrepant ABO types after the procedure. Certain drugs such as daratumumab can cause false panagglutination, acting like a warm autoantibody. A very high MCHC result might indicate cold agglutination, causing ABORH type discrepancy. A clinically significant Kidd antibody is infamous for showing dosage-may not show up in each antibody screen, leading to failure to transfuse antigen negative units. With patient’s history at hand, workup required to find a compatible blood product can be significantly reduced.
5.Dilution: When the sample is significantly diluted, the level of antibodies in plasma/serum might not be enough to be detected. Collecting personnel needs to ensure proper sample collection. Testing personnel needs to visibly check the sample quality (e.g.color of plasma, ratio of packed red cell and plasma upon centrifugation). When the sample is severely hemolyzed, it is difficult to determine a true positive antibody screen result.
5.Microbiology
1) Limitation of gram stain technique: Since developed 1882, gram stain technique has played an essential role in identifying gram negative bacteria from gram positive that guides antibiotic treatments. However, it has its limits especially with poor staining techniques, bacteria that don’t take up stains very well due to age or its nature. Checking the history can be helpful if the result is questionable. if the patient has an offending organism in urine culture, it might be the same organism that is causing sepsis.
2) Improper technique: When testing susceptibility, some manual methods such as Kirby-Bauer may be required. Not creating an even bacterial lawn(i.e.same concentration) for antibiotic disks to sit on may lead to a misleading, unreliable disk diffusion reading. If susceptibility results are questionable and manual methods are used for testing, it might be beneficial to retest for confirmation.
3) Interpretation error: Usually not every organism on a culture is identified and susceptibility is tested, especially with cultures from non-sterile sources. If an organism doesn’t have a typical morphology for a possible pathogen for that specific source, it generally doesn’t warrant further workup. However, some strains of an organism lack typical morphologies and those are often mistaken or missed (e.g. non hemolytic GBS, non hemolytic Staphylococcus aureus). If the patient has a recurrent infection and only normal flora are found in the culture, the offending organism might have a non-typical colony morphology that is being passed as normal flora or it could be one of normal flora acting as an opportunistic pathogen.
References
Whitehead NS, Williams L, et al. The Effect of Laboratory Test–Based Clinical Decision Support Tools on Medication Errors and Adverse Drug Events: A Laboratory Medicine Best Practices Systematic Review. Academic.oup.com. https://academic.oup.com/jalm/article/3/6/1035/5603075 Published May 1, 2019. Accessed April 26, 2022.
Submitted by (Gina Lee)