Interferences SES

Interferences - are some methods better than others?: 

Interferences - are some methods better than others? Graham Jones Department of Chemical Pathology St Vincent’s Hospital, Sydney

Contents: 

Contents Background Choosing your instrument Using your instrument

Introduction: 

Introduction Our aim: to produce timely, accurate results to allow optimal patient care Interferences - substances present in a sample, or events affecting a sample, which lead to the production of inaccurate results Accuracy: results which reflect the result which would have been obtained if the interference had not been present

Interference Importance: 

Interference Importance May lead to a clinical error Wrong management with bad outcome Interference-related clinical errors quite rare Most clinical errors require several mishaps concurrently Many 'near misses' BUT: can cost time, additional testing, reduced doctor confidence

Error Importance: 

Error Importance Erroneous and Non-believable eg potassium of 10.0 due to haemolysis or EDTA contamination Result: ignore or recollect specimen Erroneous and Believable eg potassium of 5.5 due to haemolysis or EDTA contamination result: unnecessarily cease potassium supplements

Common Interferences: 

Common Interferences In-vitro haemolysis Bilirubin Lipaemia Drugs Immunoglobulins Events (eg delayed separation) Other (artificial blood)

Common Interferences: 

Common Interferences In-vitro haemolysis Bilirubin Lipaemia Drugs Immunoglobulins Events (eg delayed separation) Other (artificial blood) The visible interferences

Given factors: 

Given factors We wish to have accurate results We wish to avoid errors due to interferences We aim to give out results when they are accurate We aim to withhold results which are inaccurate This implies different cutoff levels for different analytes

Choosing your instrument: 

Choosing your instrument

Assesment of Interferents: 

Assesment of Interferents Melvin Glick Clin Chem (1987) 33: 1453-1458 Add known amounts of RBC lysate; Intralipid; bilirubin to normal serum Standard procedures Plot percent change in result vs interferent concentration 'Interferographs'

Interferographs: Glick: 

Interferographs: Glick 100 200 0 Final/original result x 100 (%) 1000 500 0 Haemolysate added (as haemoglobin. mg/dL) 110% 90% * Bilirubin C.Bilirubin Glucose * Urea * Chloride * Creatinine GGT

Glick: 

Glick Most work performed in 1980s Work performed using his own blood (reliable supply, but limited quantity) Limited comprehensive third party data available for current instruments Data from our own studies Haemolysis Interference in Modern Instruments Clin Biochem Revs 2000;21:124 Icterus Interference in Modern Instruments Clin Biochem Revs 2000;21:124

Interferogram: 

Interferogram Roche Modular andlt;Pandgt; Haemolysis Haemolysate added to patient samples and concentrations measured

Instruments: 

Instruments

Comparing Interference Performance: Amylase and Haemolysis: 

Comparing Interference Performance: Amylase and Haemolysis Haemoglobin (mg/dL) Using RCPA-AACB Allowable Limits of Performance 160 990

Slide16: 

Instrument Comparison: 

Instrument Comparison Some Instruments are better than others but All are affected by interferences Data is NOT transferable between instruments There is room for improvement by manufacturers

Effect of Haemolysis - methods: 

Effect of Haemolysis - methods Examples of tests where different instruments show wide variations in response to haemolysis (Data from 2000).

Method Comparison: 

Method Comparison Some methods better than others Suggest choosing methods which are less prone to interference May require third party supplier

Using your instrument: 

Using your instrument

Using Your Instrument: 

Using Your Instrument Once the instrument is chosen the fun begins A protocol must be set which allows appropriate response to samples with interferences requires detailed knowledge of your method / instrument Sources: Manufacturer Literature Own studies

Slide22: 

Slide23: 

Olympus Cholesterol Reagent and Modular Cholesterol Reagent Olympus Results Modular Results

Slide24: 

Slide25: 

Olympus Results: 10% at 500 2.5, 4.1 and 6.0 mmol/L Modular Results 10% at 700 3.5 mmol/L Response best expressed as absolute (not not percentage)

Data Sources: 

Data Sources Best data is from your own instrument No factors Full data set Perform experiment as needed. Manufacturer information best when all results available Beware of 'No Interference' limits (eg 10%) Format of limits may not be useful

How accurate do we need to be?: 

How accurate do we need to be? RCPA-AACB Quality Assurance limits Change greater than 2 SD of analytical precision Change related to biological variation 10% Other fixed percentage or absolute values A difference that may lead to a change in clinical management - subjective*

Error Budget: 

Error Budget Int. error Other errors Total error

The Accuracy - Utility Balance: 

The Accuracy - Utility Balance More accuracy More rejections More recollections More delays Unhappier ptns and Drs Fewer clinical errors Less accuracy Fewer rejections Fewer recollections Shorter TAT More clinical errors

Interference Limits: 

Interference Limits No easy solution Take all factors into account Likely clinical effects is the main parameter (personal opinion) Include pathologist / clinician in decision making

Slide31: 

SydPath Haemolysis Protocol

Other quality factors: 

Other quality factors Sample type: Serum, heparin plasma, EDTA plasma, fluoride oxalate, Citrate, gel separators. Sample stability As whole blood, as serum/plasma At RT, 4 degrees, -20 degrees

Slide33: 

What can I do that will make a difference to your business? A supplier’s question…

Suppliers…..: 

Suppliers….. Quality data on interferences, sample types and analyte stability can: Reduce recollections Reduce unnecessary recollections Reduce repetition in multiple laboratories Head office, literature watch, local data

Conclusions: 

Conclusions Interferences and our response to them are part of providing a quality laboratory service Choose methods and instruments with low interference Choose methods where data is available about interferences or generate local data Implement a policy for responding to interferences