Technology

External Validation Summary

IntelligentMDx External Validation Study Data:

Click here for pdf version (237kb).



Study Site

Assay

Sample Type

Reference Method

Sample #

Sensitivity

Specificity

Confirmation Method Sensitivity & Specificity 5

UAB

INF-A

Swab

Culture/DFA

38

100%

100% 1

 

UVA

INF-A

Swab

Culture/EIA

98

95.9%

97.9% 2

100%

UVA

INF-A

Swab

Cepheid Flu A/B TaqMan

98

100%

100%

 

BWH

HSV NT

Swab

Culture/DFA (typing)

105

98.1%

96.2% 3

100%

BWH

HSV typing multiplex

Swab

Culture/DFA (typing)

105

HSV1 92.9% 4

HSV2 100%

HSV1 98.7%

HSV2 98.8%

HSV1 100%

HSV2 100%

NYS

HSV NT

CSF

CLEP-validated TaqMan

100

100%

100%

 

NYS

HSV typing multiplex

CSF

CLEP-validated TaqMan

100

HSV1 100%

HSV2 100%

HSV1 100%

HSV2 100%

 

UAB

HSV NT

Swab or CSF

In-house validated PCR

72

100%

100%

 

WUSTL

qBKV

Plasma or urine

In-house validated LightCycler assay

100

100%

100%

 

Additional studies ongoing for other assays, at these and other validation study sites

UAB: Department of Pediatrics at the University of Alabama, Birmingham
UVA: University of Virginia Health System, Charlottesville
BWH: Brigham and Women's Hospital with Harvard Medical School, Boston
NYS: New York State Department of Health at the Wadsworth Center, Albany
WUSTL: Washington University in St. Louis School of Medicine
CC: Cleveland Clinic
INF-A: Influenza A Virus
HSV: Herpes Simplex Virus
NT: Non-typing for HSV type 1 (HSV1) and HSV type 2 (HSV2)
qBKV: Quantitative BK Virus
DFA: Direct Immunofluorescence Assay
EIA: Enzyme Immunoassay

Sensitivity and Specificity were calculated based on comparison to the standard reference method used at each study location.

  1. One sample had potential low level contamination. All further testing of this sample at UAB and IntelligentMDx showed that it was negative leading to 100% concordance.
  2. Three discordant IntelligentMDx results were reproducible with additional TaqMan testing at UVA and IntelligentMDx, and were confirmed with Cepheid Flu A/B TaqMan results.
  3. Three discordant IntelligentMDx results were reproducible with additional TaqMan and culture testing at BWH and IntelligentMDx, and were confirmed by further Cepheid HSV typing ASR results at IntelligentMDx.
  4. Four discordant IntelligentMDx results. Three were the same samples described above in 3. The fourth was the result of a previously undescribed HSV1 strain, and was no longer discordant when the annealing temperature was optimized to a more permissive temperature. The sensitivity is raised to 96.4% if the more permissive temperature is used.
  5. Cepheid’s nucleic acid-based tests were used to confirm test results with discrepant samples. The sensitivity and specificity for IntelligentMDx’s test were recalculated based on the confirmation method results.