Replies to post #7635 on NanoLogix Inc (NNLX)

[Abig]

Table 2B
Prevalence, sensitivity, specificity, positive, and negative predictive values after the
addition of bacitracin to all washing steps
Percent (95% CI)
Prevalence 31.3 (24.2–39.5)
Sensitivity 100 (89.1–100)
Specificity 93.6 (86.9–97.2)
Positive predictive value 85.1 (71.1–93.3)
Negative predictive value 100 (95.5–100)

We have found that by adding bacitracin to each washing step, one
may greatly reduce interference with Enterococcus. While this allowed
for the increased specificity found when samples were retested, it
should be noted that there were still 7 samples that were positive by
the rapid test but negative by traditional culture. While it is possible
that these samples may have additional contaminants causing the
results to be persistently falsely positive, it may be possible that a GBS
non–beta-hemolytic variant would be missed by traditional culture
but detected by the antibody-based rapid test. Indeed, one of these
samples was confirmed to contain GBS by PCR.

[parentsgratae]

PCR? That is the best you have? How does PCR determine antibiotic sensitivity [AST];? the best treatment that will hit the infection, whether the GBS is penicillin resistant. You must have a culture based test that determines the AST. Which is one of many reasons Dr. Faro [and the TMC] now has studies/publications on our tests; PCR is not adequate.
Parents

[Longpicker]

Fulcrum,
What would be the reasoning behind this action by NNLX if what you are writing about was a concern to management?

"April 05, 2013:
NanoLogix plans to begin marketing this summer of their new N-Assay Multiwell Diagnostic test. The N-Assay is a machine-readable test that will be available in an assortment of size configurations for detection and identification of a variety of bacteria. Tests to date have provided consistent results in as little as 30 minutes with both high sensitivity and specificity."

[parentsgratae]

fulcrumgavel...did you read more than the abstract because the test you cite is not that accurate in sensitivity or specificity as it seems?
The study you cite is a very nice paper published in 2009, from a french group. This was a large study in which over 900 patients were enrolled. The test did show a very high degree of specificity and sensitivity, for a certain portion of the tests. If you read more than the abstract, you would have seen that the real data is actually very different. Out of 968 patients enrolled, more than 10% were "not available." this was due to things like mucous being present (which is present in a lot of these samples), or user error. So what the authors did, was to omit these results from analysis.

UTHSC Houston chose to include samples which did not give exactly what they wanted, for the purpose of being candid and accurate. (They were under the impression that users would presumably want to know how the test truly performs, ie, it's accuracy, as well as its sensitivity and specificity.)

I believe the accuracy/credibility of the study you cite is in question.

Parents

[Petro_Drlg_Engineer]

Fulcrumgavel; seems if one ignores over 10% of the patients/results, then you can get a better result?

Copied from your referenced paper:
"RESULTS
We prospectively enrolled 968 pregnant women who presented for intrapartum care from April 2007 through March 2008. They all underwent both intrapartum GBS culture and molecular testing. Characteristics of the study population are presented in Table 1.
The intrapartum GBS colonization rate according to culture was 14.8% (95% CI, 12.7–17.1). Molecular GBS test results were not available for 105 (10.8%) women. Of these 42 (40%) were because of PCR inhibition, 39 (37%) because of a significant presence of mucus, and 24 (23%) because of manipulation errors in loading the cartridge at the beginning of the study. Consequently, the overall molecular GBS test yield was 89.2% (95% CI, 87.0–91.0), and the positive test yield was 95.8% (95% CI, 91.1–98.1). Among the 863 women with available results, the molecular GBS test had a sensitivity of 98.5%, a specificity of 99.6%, a positive predictive value of 97.8%, and a negative predictive value of 99.7% (Table 2)."

It works, except for when it doesn't... Is that it?

Only in my opinion from your referenced paper.
PDE