Search Results

ShinetellTM Digital Pregnancy Test is a pregnancy test. It is used for the qualitative detection of hCG in human urine as an aid in early detection of pregnancy. For in vitro diagnostic use, for over the counter use.

Shinetell™ Digital Pregnancy Test is used for in vitro qualitative detection of Human Chorionic Gonadotropin (HCG) in human urine, and is designed to be tested in dip or midstream mode. The test device consists of a single test strip assembled in a plastic housing, with an absorbent tip. The device is in a ready-to-use format. Shinetell™ Digital Pregnancy Test uses lateral flow immunoassay and light reflection for the detection of the HCG in urine specimens. The test would detect the light intensity by using the LED as the light source. After that, the result can be displayed on the display screen.

The provided document describes the Shinetell™ Digital Pregnancy Test, a device for the qualitative detection of hCG in human urine as an aid in early detection of pregnancy, for over-the-counter use. The document focuses on demonstrating the device's performance characteristics and its substantial equivalence to a legally marketed predicate device.

Here's a breakdown of the acceptance criteria and the study that proves the device meets these criteria:

1. Acceptance Criteria and Reported Device Performance

The document doesn't explicitly list "acceptance criteria" in a separate table, but the performance characteristics evaluated serve as the de facto criteria for demonstrating substantial equivalence. The key performance metrics and their reported results are:

2. Sample Sizes Used for the Test Set and Data Provenance

• Analytical Performance (Precision/Sensitivity):

  • Sample Size: For each hCG concentration (0, 12.5, 15, 18.75, 22.5, 25, 50, 100, 200 mIU/mL), 10 replicates were tested per day for 5 days for each of 3 device lots. With 3 operators, this resulted in 50 tests per operator per lot per concentration (10 replicates/day * 5 days), totaling 150 tests per concentration per lot.
  • Overall Sensitivity Sample Size: For the sensitivity conclusions, the most critical concentrations (0, 12.5, 25, 50, 100, 200 mIU/mL) each had 300 tests (3 lots * 50 replicates * 2 methods, or summarized directly as 300 total results).
  • Data Provenance: The document does not specify the country of origin for these spiked urine samples. It implies a laboratory setting for the spiking and testing, rather than patient samples. The study is retrospective in the sense that controlled samples were created for testing.

• Specificity and Cross-Reactivity:

  • Non-pregnant females: 300 samples (100 from each age group: pre-menopausal, peri-menopausal, post-menopausal).
  • Cross-reactive substances/hCG ß-core fragment: Number of samples not explicitly stated per substance/concentration, but it involved "negative and positive female urine samples" spiked with the substances.
  • Data Provenance: Not specified, but implied to be collected samples.

• Interfering Substances:

  • Number of samples not explicitly stated per substance/concentration, but involved "urine samples containing 0, 5 and 25 mIU/mL hCG" spiked with the substances.
  • Data Provenance: Not specified, implied to be collected urine samples spiked in a lab setting.

• Method Comparison Study:

  • Sample Size: 200 women presenting to test for pregnancy, suspected to be pregnant (early stage, less than 5 weeks).
  • Data Provenance: Retrospective, collected from patients at three Point-of-Care (POC) sites. Country of origin not specified.

• Lay Person Study:

  • First Study (self-testing): 200 women (individual pregnancy status self-tested).
  • Second Study (blinded spiked samples): 100 laypersons for 5 mIU/mL hCG aliquots and 100 laypersons for 25 mIU/mL hCG aliquots.
  • Data Provenance: "Individuals with varying educational and occupational backgrounds from three sites." Country of origin not specified. The first study used patient samples (their own urine); the second used controlled, spiked samples. Both are prospective in terms of the layperson testing event.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

• Analytical Performance (Precision/Sensitivity): Ground truth was established by the known concentration of hCG spiked into the negative urine samples, traceable to the 5th WHO International Standard. No human experts were used to interpret these results beyond confirming the digital readout.
• Specificity, Cross-reactivity, Interfering Substances: Ground truth was established by the known negative status of the samples or the known spiked concentrations.
• Method Comparison Study: The predicate device served as the "ground truth" for comparison. The study states "3 different professionals using the candidate device and 1 professional using the predicate device at each site." Their qualifications are not specified beyond "professionals."
• Lay Person Study:
  • First Study (self-testing): The "professional results" served as ground truth for comparison with layperson results. The qualifications of these "professionals" are not specified.
  • Second Study (blinded spiked samples): The known spiked concentration (5 mIU/mL hCG meant to be negative, 25 mIU/mL hCG meant to be positive) acted as the ground truth. A "study administrator" was present to observe; their role was not to establish ground truth but to monitor the study.

4. Adjudication Method for the Test Set

The document does not describe a formal "adjudication method" involving multiple readers/experts to resolve discrepancies for the test sets.

• For analytical performance, ground truth was by spiking, so no adjudication needed.
• For method comparison and lay person studies, comparison was made against the predicate device results or professional results (in the case of actual patient samples). In the second layperson study, ground truth was by known concentration. Discrepancies (if any arose) or how they were handled (e.g., re-testing, exclusion) are not explicitly detailed.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done

No. This is a digital pregnancy test with a clear "Pregnant" or "Not Pregnant" readout, not an imaging device requiring human interpretation for complex diagnostic decisions. Therefore, a multi-reader multi-case (MRMC) comparative effectiveness study to assess how human readers improve with AI vs. without AI assistance is not applicable to this device. The "human readers" (laypersons) are using the device to get a direct digital output, not making an interpretation with or without AI assistance.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

Yes, in essence, the "Analytical Performance" section (Precision/Sensitivity, Hook Effect, Specificity, Interfering Substances) represents the standalone performance of the device, as it evaluates the device's ability to detect hCG at various concentrations and in the presence of challenging substances, independent of human interpretation variability. The digital output is the algorithm's direct result.

7. The Type of Ground Truth Used

The types of ground truth used varied depending on the performance characteristic being evaluated:

• Known Spiked Concentrations: For sensitivity, hook effect, cross-reactivity (with defined interferers), and the second layperson study. This is a highly controlled, artificial ground truth.
• Clinical (Patient) Samples with Known Status (or compared to predicate/professional): For method comparison study and the first layperson study. Here, the "ground truth" was either the result from the legally marketed predicate device or the result from a "professional" using the candidate device, implying a clinical assessment of the patient's pregnancy status (e.g., confirmatory lab tests, clinical history).

8. The Sample Size for the Training Set

The document does not provide information on the training set size because this is not an AI/ML device that requires a separate "training set" in the conventional sense. This is an immunoassay device with a digital readout mechanism based on light intensity detection. Its "design" is based on biochemical and optical principles, not on learned parameters from a large dataset. Therefore, there's no "training set" as would be seen for, say, an image-recognition AI algorithm.

9. How the Ground Truth for the Training Set Was Established

As noted in point 8, there is no "training set" for this type of device in the context of an AI/ML algorithm. The device's operational parameters (e.g., thresholds for light intensity to display "Pregnant" or "Not Pregnant") are likely established during the device's development and internal validation processes using controlled, known concentrations of hCG, similar to the analytical performance testing described.

Ask a specific question about this device