Search Results

The Evidence Pregnancy Test is a qualitative, one-step assay for the detection of human chorionic gonadotropin (hCG) in urine. The Evidence Pregnancy Test is intended to be used by the lay consumer.

Evidence Pregnancy Test is a pregnancy test to be used for detecting human Chorionic Gonadotropin (hCG) in urine. The presence of hCG usually appears about the eighth or ninth day after fertilization. The Evidence Pregnancy Test will detect hCG in urine as early as the first day after a missed menses. The Evidence Pregnancy Test will be made available to the lay consumer.

The Evidence Pregnancy Test employs a unique combination of monoclonal-dye conjugate and polyclonal-solid phase antibodies to selectively identify human Chorionic Gonadotropin (hCG) in urine. As the urine flows through the absorbent portion of the device, the antibody-dye conjugate binds to the hCG forming an antibody-antigen complex binds to the anti-hCG antibody in the positive reaction zone and produces a pink-rose color band if hCG concentration is greater than 25 mIU/ml. In the absence of hCG, there is no line in the positive reaction zone. Unbound conjugate binds to the reagents in the control zone, producing a pink-rose color band, demonstrating that the reagents are functioning correctly.

The provided text describes the 510k submission for the "Evidence Pregnancy Test" by Universal Diagnostics L.L.C. and includes information about its performance evaluation.

Here's a breakdown of the requested information:

Acceptance Criteria and Reported Device Performance

Acceptance Criteria are not explicitly stated in the provided text. However, the product's performance is compared to two substantially equivalent devices: the Hoffman-La Roche Pregnosis Pregnancy Test and the Pacific Biotech Cards + Q.S. Pregnancy Test. The goal of the study was to demonstrate "substantial equivalence."

The performance metrics reported are:

Note regarding Specificity (vs. Pregnosis): The reported specificity of 60% for the Evidence Pregnancy Test when compared to Pregnosis (professional user) is notably lower than the other comparisons. The text clarifies that two "false positive" results from the Evidence test were actually confirmed positive by a more sensitive test (Pacific Biotech Cards + Q.S.), suggesting the Evidence test is more sensitive than Pregnosis, and thus, the "false positives" might be true positives that Pregnosis missed.

Study Details:

1. Sample size used for the test set and the data provenance:

   • Pregnosis vs. Evidence (Professional User): 100 samples (95 positive, 5 negative by Pregnosis standard).
   • Cards + Q.S. vs. Evidence (Professional User): 2 samples (both positive by Cards + Q.S. standard).
   • Medical Technologist vs. Study Participant (Lay Consumer): 100 study participants.
   • Data Provenance: Not explicitly stated, but given the manufacturer is in France, it's possible the clinical trial was conducted there or in the US. It appears to be a prospective clinical trial as it's described as "A clinical trial was done to confirm the performance."

2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g., radiologist with 10 years of experience):

   • For the "Pregnosis vs. Evidence" and "Cards + Q.S. vs. Evidence" comparisons, the "ground truth" was established by the comparative devices (Pregnosis and Cards + Q.S.) as read by "Professional Users." The number and specific qualifications of these professional users are not detailed.
   • For the "Medical Technologist vs. Study Participant" comparison, the "ground truth" was defined by the results obtained by a Medical Technologist. The number of medical technologists is not specified.

3. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

   • No explicit adjudication method is described. The comparison seems to be direct against a reference test or a professional's reading.
   • For the two "false positive" results in the Pregnosis comparison, an external device (Pacific Biotech Cards + Q.S.) was used for confirmation, which could be considered an informal adjudication to clarify discrepancies.

4. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

   • This is not an AI-assisted device. The study does not describe an MRMC comparative effectiveness study involving AI. It focuses on the performance of a standalone rapid diagnostic test for pregnancy.

5. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:

   • Yes, the performance data presented effectively represents standalone performance of the Evidence Pregnancy Test. The device itself (the "algorithm") produces a visual result (a pink-rose color band) which is then interpreted by a human user (either a professional or a lay consumer). The "Medical Technologist vs. Study Participant" comparison also assesses the standalone performance of the device when used by the intended lay consumer.

6. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

   • The ground truth was established by:
     • Reference device results (Pregnosis and Cards + Q.S.) as read by "Professional Users."
     • Medical Technologist's interpretation of the Evidence Pregnancy Test results.
   • Ultimately, the underlying ground truth for a pregnancy test is the presence or absence of hCG, which is a biochemical marker. The reference devices and professional readings are considered proxies for this biochemical ground truth.

7. The sample size for the training set:

   • The document does not specify a separate training set for the device. The "Performance Data" section describes clinical trials to confirm performance, which implies testing the final product. As this is a rapid diagnostic test based on immunoassay, it typically does not involve a "training set" in the machine learning sense. The device's design and parameters are likely established through laboratory development and optimization, not machine learning training.

8. How the ground truth for the training set was established:

   • Not applicable, as no training set (in the machine learning context) is mentioned for this immunoassay device. The device's design is based on the known biochemical interaction of antibodies and hCG.

Ask a specific question about this device