Interference and Point-of-Care Testing Devices Nam K. Tran, PhD, - - PowerPoint PPT Presentation
Interference and Point-of-Care Testing Devices Nam K. Tran, PhD, HCLD, (ABB), FACB Associate Clinical Professor Director of Clinical Chemistry, Special Chemistry/Toxicology and POCT 1 Learning Objectives Identify common interferences
1
Nam K. Tran, PhD, HCLD, (ABB), FACB Associate Clinical Professor Director of Clinical Chemistry, Special Chemistry/Toxicology and POCT
2
testing
affected patient care.
interfering substances on POC testing.
3
Definition: POCT is defined as testing at or near the site of patient care
4
Examples:
5
Examples:
Being FDA approved as a POCT device does not mean it is not susceptible to interfering substances!!!
6
Pre-Analytical Total Testing Process: Lab testing occurs over three critical phases:
7
Pre-Analytical Analytical
Total Testing Process: Lab testing occurs over three critical phases:
8
Pre-Analytical Analytical Post-Analytical
Total Testing Process: Lab testing occurs over three critical phases:
9
Pre-Analytical Analytical Post-Analytical TREATMENT
Total Testing Process: Lab testing occurs over three critical phases:
10
Pre-Analytical Analytical Post-Analytical TREATMENT Errors in the Pre-Analytical Phase: Most frequent source of errors (up to 70%). Incorrect Patient preparation Sample collection Transportation Accessioning Processing
Components
11
Pre-Analytical Analytical Post-Analytical TREATMENT Errors in the Pre-Analytical Phase: Most frequent source of errors (up to 70%). Incorrect
Patient preparation Sample collection Transportation Accessioning Processing Incorrect patient ID Mislabeling of specimens Hemolysis Wrong specimen type Improper specimen collection Interfering substances Components Sources of Error
12
Pre-Analytical Analytical Post-Analytical TREATMENT Testing
Patient preparation Sample collection Transportation Accessioning Processing Incorrect patient ID Mislabeling of specimens Hemolysis Wrong specimen type Improper specimen collection Interfering substances Errors in the Analytical Phase: Infrequent in laboratory tests, however may be higher in POCT due to non-lab trained personnel operating devices. Components Sources of Error
13
Pre-Analytical Analytical Post-Analytical TREATMENT Testing
Patient preparation Sample collection Transportation Accessioning Processing Incorrect patient ID Mislabeling of specimens Hemolysis Wrong specimen type Improper specimen collection Interfering substances QC/calibration Operator error Bad reagents Errors in the Analytical Phase: Infrequent in laboratory tests, however may be higher in POCT due to non-lab trained personnel operating devices. Components Sources of Error
14
Pre-Analytical Analytical Post-Analytical TREATMENT Results interpretation Entry to LIS/EMR Contacting providers Sample archiving
Testing Patient preparation Sample collection Transportation Accessioning Processing Incorrect patient ID Mislabeling of specimens Hemolysis Wrong specimen type Improper specimen collection Interfering substances QC/calibration Operator error Bad reagents Errors in the Post-Analytical Phase: Second most common among laboratory-based results. Components Sources of Error
15
Pre-Analytical Analytical Post-Analytical TREATMENT Results interpretation Entry to LIS/EMR Contacting providers Sample archiving
Testing Patient preparation Sample collection Transportation Accessioning Processing Incorrect patient ID Mislabeling of specimens Hemolysis Wrong specimen type Improper specimen collection Interfering substances QC/calibration Operator error Bad reagents Misinterpretation of results IT problems Errors in the Post-Analytical Phase: Second most common among laboratory-based results. Components Sources of Error
16
Pre-Analytical Analytical Post-Analytical TREATMENT Results interpretation Entry to LIS/EMR Contacting providers Sample archiving
Testing Patient preparation Sample collection Transportation Accessioning Processing Incorrect patient ID Mislabeling of specimens Hemolysis Wrong specimen type Improper specimen collection Interfering substances QC/calibration Operator error Bad reagents Misinterpretation of results IT problems Errors in the Post-Analytical Phase: Second most common among laboratory-based results. Components Sources of Error
17
18
each year – highest number of adverse events for any in vitro diagnostic device.
19
each year – highest number of adverse events for any in vitro diagnostic device.
20
each year – highest number of adverse events for any in vitro diagnostic device.
Most of these reported errors are due to erroneous results from interfering substances and operator error.
21
each year – highest number of adverse events for any in vitro diagnostic device.
Most of these reported errors are due to erroneous results from interfering substances and operator error.
22
each year – highest number of adverse events for any in vitro diagnostic device.
Most of these reported errors are due to erroneous results from interfering substances and operator error.
23
Anemia and polycythemia causes falsely high or falsely low results respectively.
24
20 40 60 10 20 30 40 50 60 70
BIAS (mg/dL) Hematocrit (%) y = -0.9465x + 30.951 R² = 0.4171
Note: Bias = BGMS – Plasma Glucose Sample Size: 60 Hematocrit Range: 19 - 60% Glucose Range: 90 - 296 mg/dL
25
26
Oxidizing and reducing substances interfere with electrochemical sensors causing falsely high or low results.
Tang Z, et al. Am J Clin Pathol 2000;113:75-86
28
50 100 150 200 250 300 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 Time Following Admission (Hours) Blood Glucose (mg/dL) Started Ascorbic Acid (66 mg/kg/hr) Stopped Ascorbic Acid
BGMS B***
History: Patient is a 21 y/o woman with 90% TBSA burns from MVA.
50 100 150 200 250 300 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 Time Following Admission (Hours) Blood Glucose (mg/dL) Stopped Ascorbic Acid
BGMS B PLASMA GLUCOSE (Lab)
Started Ascorbic Acid (66 mg/kg/hr)
50 100 150 200 250 300 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 Time Following Admission (Hours) Blood Glucose (mg/dL) Stopped Ascorbic Acid
BGMS A BGMS B*** PLASMA GLUCOSE (Lab)
Started Ascorbic Acid (66 mg/kg/hr)
32
33
Specimen temp alters biosensor enzyme kinetics. Hypotension/shock affect capillary specimens.
34
Some glucose meters cannot differentiate between certain non- glucose sugars (e.g., maltose, galactose)
35
37
FDA MAUDE Database website: http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfmaude/ search.cfm, Accessed on August 20, 2014
BGMS A BGMS B BGMS C Timeframe 1997-14 2013-14 2007-11 Adverse Events (Deaths) 28 (13) 5 (0) 0 (0) Erroneous Results 557 168 15 Non-Clinical Event 387 59 21 TOTAL 1094 232 36
38
to similar interferences (will vary based on manufacturer).
measurements and not plasma or whole blood.
interferences.
most devices compared against
such as the YSI.
39
40
41
42
Background: Anesthesia reports “impossible venous blood gas values” in one patient where end tidal CO2 was greater than the venous blood gas (VBG).
43
Background: Anesthesia reports “impossible venous blood gas values” in one patient where end tidal CO2 was greater than the venous blood gas (VBG).
44
Background: Anesthesia reports “impossible venous blood gas values” in one patient where end tidal CO2 was greater than the venous blood gas (VBG).
45
Background: Anesthesia reports “impossible venous blood gas values” in one patient where end tidal CO2 was greater than the venous blood gas (VBG).
46
Background: Anesthesia reports “impossible venous blood gas values” in one patient where end tidal CO2 was greater than the venous blood gas (VBG).
equilibrate atmospheric air (1 atm = 760 mmHg = 0.21 x 760 = 150 mmHg for pO2!!!)
47
48
49
50
51
52
53
54
55
56
the absorption spectra
different absorption in the red to IR wavelengths.
(A = elc).
with non-cyanide-based reagents.
Absorbance
HHb O2Hb COHb MetHb
Prism White light
500 550 600 650 700 nm
57
Electrode VS.
High Resistance Low Resistance
58
Electrode VS. Hematocrit (%) Resistance (Ω)
59
Electrode VS.
Ohm’s Law (V = IR)
60
Electrode VS.
Ohm’s Law (V = IR)
hemoglobin based on a conversion factor (estimated hemoglobin = hematocrit / 3.4)*
61
Electrode VS.
62
63
64
65
66
67
Handheld Results Hct = 68% Hb = 21.9 g/dL
68
Handheld Results Hct = 68% Hb = 21.9 g/dL CBC Results Hct = 41% Hb = 13.2 g/dL
69
Hct = 43% Hb = 13.8 g/dL Handheld Results Hct = 68% Hb = 21.9 g/dL RE-MIXING! CBC Results Hct = 41% Hb = 13.2 g/dL
70
Hct = 43% Hb = 13.8 g/dL Inadequate mixing may result in artificial changes in total hemoglobin measurements. Handheld Results Hct = 68% Hb = 21.9 g/dL RE-MIXING! CBC Results Hct = 41% Hb = 13.2 g/dL
71
Electrode
= Plasma Protein High Resistance
72
Electrode
= Plasma Protein
hemodilution, hematocrit may be falsely lower and causing an underestimation of total hemoglobin.
Low Resistance from low plasma protein concentration!
73
Electrode
= Plasma Protein
hemodilution, hematocrit may be falsely lower and causing an underestimation of total hemoglobin.
measurement of hemoglobin versus a benchtop blood gas analyzer using a spectrophotometric- based method for hemoglobinometry.
74
were evaluated.
handheld POC analyzer and spectrophotometric methods through the core laboratory.
P = 0.011.
patients would have received unnecessary transfusions.
75
were evaluated.
handheld POC analyzer and spectrophotometric methods through the core laboratory.
P = 0.011.
patients would have received unnecessary transfusions.
76
were evaluated.
handheld POC analyzer and spectrophotometric methods through the core laboratory.
P = 0.011.
patients would have received unnecessary transfusions.
Toner RW, et al. Appl Health Econ Health Policy 2011;9:29-37
77
78
79
80
81
82
y = 0.5092x + 4.0176 R² = 0.5253 4 5 6 7 8 9 10 11 4 5 6 7 8 9 10 11 12
83
y = 0.5249x + 3.9443 R² = 0.5407 4 5 6 7 8 9 10 11 4 5 6 7 8 9 10 11 12
84
y = 0.9345x + 0.4057 R² = 0.9205 4 5 6 7 8 9 10 11 4 5 6 7 8 9 10 11
85
Notes: Reference Method = Beckman LH hematology analyzer
86
Notes: Reference Method = Beckman LH hematology analyzer Median (IQR) Bias: 0.78 (0.78) g/dL P < 0.001 N = 50
87
Notes: Reference Method = Beckman LH hematology analyzer Median (IQR) Bias: 0.73 (0.60) g/dL P < 0.001 N = 50
88
Notes: Reference Method = Beckman LH hematology analyzer Median (IQR) Bias: 0.22 (0.20) g/dL P = 0.510 N = 50
89
Notes: *** P<0.001, Central Lab = Spectrophotometric Method, n = 20 patients 6 6.5 7 7.5 8 8.5 9 9.5 10 1 2 3 4 5 Total Hemoglobin (g/dL) Time Point Serial Testing Performance at 7 and 8 g/dL
bias between spectrophotometry vs. conductance-based measurements. *** Spectrophotometric-based Methods
90
Notes: *** P<0.001, Central Lab = Spectrophotometric Method, n = 20 patients 6 6.5 7 7.5 8 8.5 9 9.5 10 1 2 3 4 5 Total Hemoglobin (g/dL) Time Point Serial Testing Performance at 7 and 8 g/dL
bias between spectrophotometry vs. conductance-based measurements. *** Spectrophotometric-based Methods Conductance-based Methods
91
6 6.5 7 7.5 8 8.5 9 9.5 10 1 2 3 4 5 Total Hemoglobin (g/dL) Time Point Serial Testing Performance at 7 and 8 g/dL
bias between spectrophotometry vs. conductance-based measurements.
prompted unnecessary transfusions at time point #5 for patients using the 7 g/dL cutoff. Unnecessary Transfusion Risk *** Notes: *** P<0.001, Central Lab = Spectrophotometric Method, n = 20 patients
92
6 6.5 7 7.5 8 8.5 9 9.5 10 1 2 3 4 5 Total Hemoglobin (g/dL) Time Point Serial Testing Performance at 7 and 8 g/dL
bias between spectrophotometry vs. conductance-based measurements.
prompted unnecessary transfusions at time point #5 for patients using the 7 g/dL cutoff.
at risk for potential transfusions if the 8 g/dL cutoff was used. Unnecessary Transfusion Risk *** Notes: *** P<0.001, Central Lab = Spectrophotometric Method, n = 20 patients
93
6 6.5 7 7.5 8 8.5 9 9.5 10 1 2 3 4 5
Central Laboratory Method
Total Hemoglobin (g/dL) Time Point Serial Testing Performance at 7 and 8 g/dL
bias between spectrophotometry vs. conductance-based measurements.
prompted unnecessary transfusions at time point #5 for patients using the 7 g/dL cutoff.
at risk for potential transfusions if the 8 g/dL cutoff was used. Unnecessary Transfusion Risk *** Notes: *** P<0.001, Central Lab = Spectrophotometric Method, n = 20 patients
94
Device 1 Device 2 Device 3 Timeframe 2011-2016 2011-2016 2014-2016* Erroneous Results 8 Improper Transfusions 5 https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfmaude/results.cfm, Accessed on July 19, 2016
95
96
97
98
No current FDA approved integrated solutions for detecting hemolysis at the point-of-care
99
101
102
Mumma B, et al. AACC Poster Presentation 2018
n = 797; 3-hour, n=646) from 850 patients.
103
Mumma B, et al. AACC Poster Presentation 2018
n = 797; 3-hour, n=646) from 850 patients.
hour samples.
104
n = 797; 3-hour, n=646) from 850 patients.
hour samples.
had biotin >20 ng/mL (0.13% [95% CI: 0-0.7%]).
Mumma B, et al. AACC Poster Presentation 2018
105
Mumma B, et al. AACC Poster Presentation 2018
106
107
Average Plasma Biotin: 1.15 (0.97) ng/mL
Specimens collected as part of clinical validation
108
Average Plasma Biotin: 1.15 (0.97) ng/mL
20 40 60 80 100 120 140 160 Number of Patients Biotin Concentration (ng/mL)
Gen 5 TnT Biotin Interference Threshold is 20 ng/mL Biotin quantified by GC-TOF-MS
109
Average Plasma Biotin: 1.15 (0.97) ng/mL
20 40 60 80 100 120 140 160 Number of Patients Biotin Concentration (ng/mL)
Gen 5 TnT Biotin Interference Threshold is 20 ng/mL Biotin quantified by GC-TOF-MS BIOTIN IS LESS LIKELY TO BE A PROBLEM IN CARDIAC TROPONIN TESTING AND IS POPULATION SPECIFIC!
110
111
care urine pregnancy tests were affected by biotin.
pregnancy test exhibited interference as low as 6 microgram/mL of urine biotin! Williams G, et al. Clin Biochem 2018;53:168-170
112
113
test interferences. Go to grand rounds, build partnerships, and provide multi-modality means to disseminate knowledge.
114
test interferences. Go to grand rounds, build partnerships, and provide multi-modality means to disseminate knowledge.
115
test interferences. Go to grand rounds, build partnerships, and provide multi-modality means to disseminate knowledge.
be at risk for certain interferences (e.g., biotin, vitamin C, etc). MAUDE database is also helpful!
116
disseminate knowledge.
be at risk for certain interferences (e.g., biotin, vitamin C, etc). MAUDE database is also helpful!
tests could flag both on the provider and laboratory side certain substances are identified.
117
impact POC testing as much as traditional lab testing!
as glucose meters have resulted in injury and death.
resulted in inappropriate treatment decisions.
POC immunoassays such as urine pregnancy tests.
minimizing errors associated with interfering substances.
118