Instrument Error Codes and Diagnostic Serendipity

A 60-year-old female patient presented to a family physician with recent symptoms of fatigue, but no prior medical problems. During that initial visit, the physician ordered laboratory evaluations for anemia (complete blood count, iron, and vitamin B12), thyroid status, and renal function (Table 1). The ordered test results were not remarkable except for an unexplained mild anemia: hemoglobin 9.1 g/dL (91 g/L); hyponatremia: Na 127 mmol/L; and hypoalbuminemia: albumin 2.1 g/dL (21 g/L). Selected results of laboratory testing performed after initial evaluation of the patient with age- and sex-dependent reference intervals.a Results for glucose, urea, magnesium, phosphate, alkaline phosphatase, alanine transaminase, aspartate transaminase, γ glutamyltransferase, and lipase were within reference intervals and are not shown. Selected results of laboratory testing performed after initial evaluation of the patient with age- and sex-dependent reference intervals.a Results for glucose, urea, magnesium, phosphate, alkaline phosphatase, alanine transaminase, aspartate transaminase, γ glutamyltransferase, and lipase were within reference intervals and are not shown. Results for vitamin B12 could not be provided by the laboratory. During the analysis, a series of instrument error codes were observed by staff and the instrument stopped. The error codes were triggered after incubation of plasma with reagents for measurement of vitamin B12 on an E601 analyzer (Roche Diagnostics). The error codes appeared in the software and computer that operates the E601 analyzer and were recorded by the middleware (Cobas I.T. middleware 1.05.02) but was not communicated to the laboratory information system). The 4 error codes indicated a shortage of auxiliary reagent, an abnormal measuring cell condition, an abnormal sipper pipette movement, and a tip/cup pickup error. During investigation, staff observed that as an automated pipette probe attempted to transfer the reaction mixture into a measuring cell, the disposable reaction cup became attached to the end of the probe, and the reaction cup was pulled out of the rack as the probe rotated (Fig. 1). It appeared that the mixture of reagents and patient plasma had formed a gel or clot and the light-weight reaction cup was pulled from its rack by the gel as the pipette probe changed position, such that vitamin B12 analysis could not be completed. As the automated pipette aspirated the fluid and rotated position, the reaction cup was pulled from its rack triggering instrument error codes. The reagent-specimen mixture appeared to be attached to the pipette by a gel or a clot. Inspection of the primary patient specimen revealed there was no problem with the specimen by visual inspection. Portions of this specimen had already been dispensed for 19 chemistry tests and one other immunoassay test without triggering error codes. Two additional attempts were made to analyze vitamin B12 on this specimen, but each time, the reaction cups were pulled from the rack by the pipette probe triggering identical instrument error codes. The supervisor and laboratory director were alerted to the problem for investigation. In this case scenario, there was no visually obvious problem with the plasma specimen. The vitamin B12 test package insert listed an unusual warning that the protein concentration of specimens should not exceed 16 g/dL (160 g/L) (Vitamin B12 II package insert, 2016-10. V 2.0 English, Roche Diagnostics). That warning prompted evaluation of the plasma total protein concentration, which was found to be 16.2 g/dL (162 g/L) [reference interval 6–8 g/dL (60–80 g/L)]. The albumin level was 2.1 g/dL (21 g/L), which implied that approximately 14 g/dL (140 g/L) of unidentified globulins were in the specimen, and this result could explain why the specimen clotted or gelled when contacting the reagent. The family physician that ordered the vitamin B12 test was contacted by the laboratory medical director to explain that during analysis of vitamin B12, unusual instrument errors codes were triggered. The subsequent investigation of the error codes by the laboratory revealed the presence of a high concentration of unidentified plasma globulins, possibly associated with a pathologic condition. The laboratory cancelled the vitamin B12 test because of an unknown interference, and the family physician formally requested analysis of the plasma for total protein. The patient was recalled the next day for evaluation of serum protein electrophoresis that revealed a band of 8.6 g/dL (86 g/L) of monoclonal IgG λ. The patient was referred to the Saskatoon Cancer Centre for diagnosis and treatment of multiple myeloma. It should be noted that the low sodium concentration listed in Table 1 was generated by an indirect electrode method (C501 Analyzer, Roche Diagnostics), a method for which pseudohyponatremia has been reported in patients with increased levels of plasma proteins (1). For this patient, prompt investigation of instrument error codes and subsequent consultation rapidly led to an unexpected diagnosis. Software used to operate automated clinical laboratory analyzers often includes a series of alarms or error codes that indicate an abnormality was observed while the system is switched on. Each instrument/software manufacturer or device will have a unique series of alarms, but in general, they can be divided into data alarms (irregular numerical results or conditions) and instrument alarms (irregular system conditions). Alarms can be classified into different categories of response: (a) warning alarms about single specimens, quality control materials, and calibrators where the instrument continues to operate; and (b) stop alarms where individual analyses are interrupted and no results are generated (Cobas® 6000 analyzer series Operator's Manual, Software Version 04-01/04-02, Roche Diagnostics). In some situations, stop alarms trigger an entire instrument to cease operations. Staff operating instruments are alerted to new alarms by visual cues on computer screens and occasionally by audible alarm sounds or flashing lights. The error code usually provides a detailed description of the trigger for the alarm to facilitate troubleshooting. The purpose of the alarms and error codes is to indicate that abnormalities were detected during operation and need to be rectified by the operating staff to assure the quality of the test results. In this manner, the alarms and error codes reduce the risk that erroneous test results are generated by the analytic system. Instrument errors codes are triggered when performance thresholds are met, typically when mechanical, electrical, physical, or chemical events are detected. Inspection of the instrument can often rapidly locate unexpected physical barriers (e.g., a misplaced tube cap or obstructed pipette probe). When a problem is repeated with a specific specimen, the offending tube and specimen fluid should be closely inspected for the presence of fibrin or cryoprecipitate (flocculent protein or a gel), insoluble fibrinogen (e.g., from a freeze-thaw cycle), or abnormal viscosity. The fluid can be rimmed with a thin wooden dowel to confirm that translucent fibers are not present. The presence of oil droplets floating on the surface of the fluid is difficult to detect and they may indicate that a silicone separator gel had deteriorated. In the case situation we described, formation of a clot or gel in the reaction mixture was not expected and the clot prevented aspiration of the fluid, triggering an error code. The clotted reaction tube was pulled from its rack and impeded movement of the pipette, mechanically triggering additional error codes. The case specimen had no visible problems. The problem specimen was warmed to 37 °C to deter possible cryoglobulin formation when being reanalyzed, but that did not prevent the instrument errors. Sera containing high concentrations of monoclonal globulins are known to intermittently interfere with many different analytic methods (1–6), but there is no convenient way to evaluate if monoclonal globulins are present without performing electrophoresis. Specific sources of potential error or analytic interference are usually listed in the package insert that accompanies reagents. For this case scenario, in addition to information about interference due to hemolysis, icterus, or lipemia, the Roche Diagnostics Vitamin B12 II package insert stated the protein concentration of serum or plasma should not exceed 16 g/dL (160 g/L). Alternative sources of information on analytic or pathophysiologic interferences include the following: (i) A public compendium of adverse events related to diagnostic devices is maintained by the US Food and Drug Administration. The MAUDE (Manufacturer and User Facility Device Experience) database can be searched for interferences with specific instruments (7). (ii) The telephone support hotline or Internet communications can be used by organizing networks or user groups organized by the instrument or reagent manufacturer. (iii) Published scientific reports can be searched using PubMed (8) or Web of Science (9) indices, or (iv) clinical laboratory test interference books can be useful (10). Instrument error codes can reduce the risk of false results and help detect interfering materials in patient specimens. Package inserts supplied with reagents are a convenient and valuable source of information specific to the materials used in the laboratory. Serum or plasma specimens from patients with high levels of monoclonal proteins can cause spurious, expected test results with many different methods, reagents, and instruments. The converse is also true but difficult to remember or investigate: when spurious or unexpected test results are observed (with or without instrument error codes), the specimen may contain a high concentration of monoclonal immunoglobulin from an undiagnosed patient. We contacted Roche Diagnostics to comment about this unusual case. The susceptibility of specimens with protein concentration above 16 g/dL (160 g/L) to cause gel formation during the vitamin B12 assay incubation was observed during method development, and this information was provided in the method sheet: “Please be aware that the occurrence of gel formation in natural patient samples is dependent of the individual sample composition. The description of 160 g/L in the method sheet of the Vitamin B12 II assay can only serve as a hint for our customers because the value of 160 g/L was derived from an artificial experiment. This experiment comprised the spiking of purified human IgG and human albumin into samples. In summary, this means that natural patient samples with a protein concentration below 160 g/L can also show gel formation.” The alarms and error codes were developed both to ensure proper instrument function and to prevent reporting of inaccurate results. As a manufacturer, they were pleased to hear the combination of error codes and information on the package insert supported “diagnostic serendipity” (G. Turcotte and K. Küper, personal communication, 18 January 2017).

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.