Misclassification risks associated with interferences and assay quality

Concerns about biotin interference have revealed a need for clarity about the role and importance of individual factors that contribute to lab errors in the analytical phase and the potential risk for misclassification of patient results.

June 12, 2019 Dr. Dušanka Kasapić, Senior Medical Affairs Manager, Core Reagents, Roche Diagnostics

Errors caused by interferences are a small fraction of the overall lab test error rate and errors caused by biotin interference are an even smaller fraction of that. While risk of misclassification due to biotin interference is possible, findings from a risk assessment demonstrate it to be significantly lower than the risk from common sources of error laboratories manage with expertise every day. Published clinical studies also show the potential for result misclassification is associated with overall assay performance (e.g., clinical sensitivity and specificity, PPVs and NPVs). Laboratories should continue to have strict and appropriate protocols in place to reduce and prevent lab errors. There is also an opportunity for labs to take a leadership role in educating clinicians about analytical limitations, interferences as well as ways they can partner to further optimise laboratory results and support clinical decision making.

Role of interferences in overall laboratory error rate

As laboratory medicine is increasingly considered a cornerstone of predictive medical decision-making, understanding and removing or reducing lab errors is paramount.

Among the common causes of test error are inadequate sample collection and transport, patient misidentification, inappropriate test requests and delay in reporting results that fall within the critical limits. Interfering substances are also known sporadic causes of error. These can include heterophilic antibodies, HAMA, therapeutic drugs and high doses of certain supplements, including biotin (vitamin b7). However, the frequency with which interferences occurs and consequently erroneous results that may affect clinical management is difficult to quantify.

Errors arising from interference account for a very small proportion of lab test errors. The overall lab test error rate is 0.012-0.6 per cent; the proportion of these errors that occur in the analytical phase is 7-13 per cent. The percentage of errors in the analytical phase in relation to the overall lab test error rate is 0.078 per cent (Figure 1). Only a very small portion of these errors are due to interferences, and the potential of assay errors caused by biotin interferences is even lower.

Laboratory awareness and successful management of the common causes of errors have helped reduce the overall test error rate. Due to advances in automation and sample handling, fewer errors now occur in the analytical phase of analysis than in the pre-and post-analytical phases.

While all laboratory tests and all immunoassay manufacturers are affected by interferences, laboratories have an opportunity to educate clinicians and hospital administrators about steps undertaken to reduce lab errors from common sources including interferences, and measures in place to minimise risks.

Biotin interference risk in perspective

While the error rates caused by biotin interferences are a fraction of errors caused by overall assay interferences, biotin requires a special effort at raised awareness.

For the last few years, supplement companies have been marketing biotin supplements in 2.5 mg, 5 mg and 10 mg doses, 167 – 333 times the amount contained in a daily multivitamin and the adequate daily intake, for hair, skin and nail health, despite the lack of supportive clinical evidence. High-dose biotin is also currently being used in clinical trial settings as a potential treatment for patients with and secondary progressive multiple sclerosis.

There is no risk of assay interferences associated with the intake of biotin as part of a standard multivitamin (typically 0.03 -0.06 mg). Intake of high doses of biotin, however, has the potential to lead to interference with immunoassays.

U.S. sales data of biotin over a four-year period (2014-April 2018) indicate biotin sales have been trending slightly upwards. The steadiest growth has been in products containing 2.5 mg biotin and under doses, which have little to no risk of interference if taken according to the package insert. Sales in products containing 5 mg dose declined.

While an increase in biotin supplement sales does not necessarily correlate to an increased risk of biotin interference, laboratories and clinicians should be aware of biotin use among their patients and proactively instruct patients to stop taking biotin before a blood draw if very high biotin concentrations (e.g. 10 mg or more) are taken. If lab results seem discordant with clinical observation, clinicians should ask questions, consult with the laboratory, and repeat the test as necessary.

While all laboratory tests and all immunoassay manufacturers are affected by interferences, laboratories have an opportunity to educate clinicians and hospital administrators about steps undertaken to reduce lab errors

Assessment of risk misclassification due to biotin and other sources for variations

To get a clearer picture of the risk of misclassification due to interferences, Roche analysed and estimated the risk resulting from biotin interference for the Elecsys assays with a risk assessment model.

The estimation of risk was done by analysis of the probability of occurrence in the context of the intended use considering the clinical practice and severity of harm.

In general, this risk calculation includes “extrapolated” data for biotin prevalence (anticipating extremely values that have never been seen in prevalence studies so far), the biotin interference curve, and real-world data distribution of specific patient results derived from the intended use population.

Risk of misclassification for most Roche assays was determined typically in the range of 1 in 10,000,000 or lower. This is approximately 1,000-10,000 times lower than other risks of other errors quantified in the literature such as HAMA at 3-5 in 10,000, assay imprecision error at approximately 1 in 10,000 and biological variation error at approximately up to 1 in 1,000 (Figure 2).

Biotin interference in troponin testing

Biotin interference probability in troponin assays was also evaluated. If a dose of biotin was taken right before the onset of an Acute Coronary Syndrome (ACS), by the time the patient reaches the ER (mean 3.5-6 hours after onset of symptoms), biotin plasma concentrations will have decreased significantly. Studies show biotin peaks in the bloodstream between 1-2 hours post ingestion.

The misclassification risk for troponin is mitigated by fast wash-out of biotin and the utilisation of serial sampling in clinical practice, in accordance with the American Heart Association (AHA) and European Society of Cardiology (ESC) guidelines (Figure 3). If an AMI is suspected, AHA guidelines recommend baseline and serial troponin testing within 3-6 hours.

The risk of misclassifying a Troponin T test result due to biotin interference was assessed to be < 1 in 10 million in an acute coronary syndrome cohort.

Immunoassay overall quality performance and misclassification risk in cardiac troponin testing

Assay design contributes to its overall quality performance. Biotin has been part of the development of immunoassays for decades due to two primary reasons: a) streptavidin and biotin naturally form a strong, highly specific and stable bond; and b) the addition of biotin to antibodies or other proteins does not alter their biologic properties. This system results in a plethora of highly sensitive, specific and accurate immunoassays supporting informed clinical decision-making in general and emergency settings.

With Troponin assays, for example, the risk of missed AMI is linked to the sensitivity and negative predictive value (NPV) of the test. Clinical studies demonstrate greater risk of misclassification for hs cTnT assays based on overall assay performance (Figure 4).

The risk of misclassifying a patient due to biotin interference (<1 in 10,000,000) is a fraction of the risk due to overall assay performance (6-15 in 1,000).

Adverse events also impact potential for risk misclassification (Figure 5). Data shows the Elecsys cTnThs (Elecsys, Roche) had 5 adverse events, vs the Abbott hs-cTnl which had 90, and Beckman-Coulter, 806. The goal for future developments should be to further minimise all kinds of potential interferences.  

Opportunity for lab leadership

Protecting patient safety is global healthcare’s highest priority, and clinical laboratories can partner with clinicians to increase understanding of potential interferences and misclassification risk as well as strategies for mitigating risks and educate patients.

Clinical labs should use assays with overall high-quality performance and reliability, especially in critical settings. Immunoassays should be of the highest available standards in quality and innovation across criteria that matter most including longitudinal precision and accuracy, sensitivity, specificity, lot-to-lot consistency and adverse events.

With the clarity of facts and data, and clinician education, laboratories can confidently take a leadership role on the issue of interferences by educating clinicians about interferences, including biotin, and how labs and clinicians can work together to reduce risks and optimise lab results.  

References available on request.