Search Results

The Biodata Testosterone Maia Kit is a radioimmunoassay for the quantitative determination of testosterone in human serum or plasma. The device is intended to be used for the diagnosis of elevated or depressed levels of testosterone. It is FOR IN VITRO DIAGNOSTIC USE ONLY.

Neoplasms of the testes and adrenal cortex and hyperthyroidism may result in elevated levels of testosterone in the male. Elevated testosterone levels are observed in females with endometrial carcinoma, hirsutism, neoplasms of the ovaries, and Cushing syndrome. Patients with hypogonadism of pituitary origin, anorchia, gonadel dysgenesis, Klinefelter syndrome, and pseudohermaphroditism have decreased testosterone levels.

The Biodata Testosterone Maia kit is based on the competitive binding principles of radioimmunoassay. Testosterone (unlabeled antigen) in samples, standards, or controls competes with testosterone labeled with radioactive 1251 ("testosterone - 1251") (labeled antigen) for a limited number of testosterone antibody sites. The amount of testosterone - 1251 bound by the antibody is inversely proportional to the amount of testosterone present in the sample, standard, or control. Testosterone Maia Separation Reagent (an antibody covalently bound to a magnetic particle capable of binding the testosterone antibody) is added to the reaction mixture to facilitate the separation of the bound and free fractions of labeled testosterone. The bound fraction of each standard, sample and control is sedimented in a magnetic separator. Each bound fraction is then counted in a gamma counter calibrated to detect 125].

The concentration of testosterone in the samples can be determined by comparing the relative percent binding of the standards with known concentrates of testosterone on the standard curve. The testosterone concentration in the sample is then compared to the reference range of testosterone concentrations established by each laboratory or the expected volumes listed in the package insert.

The provided text describes several studies related to the performance of the Biodata Testosterone Maia Kit, but it does not explicitly state pre-defined acceptance criteria for each study. Instead, it presents the results and concludes that the performance is "satisfactory," "high degree of precision," "acceptable," "no significant cross-reaction," or "does not significantly affect."

Therefore, I will create a table with the study type, the reported device performance, and then explain the details of each study as they relate to the requested information.

1. Table of Acceptance Criteria and Reported Device Performance

As no explicit "acceptance criteria" are stated in the document, the table will reflect the reported findings/conclusions for each performance study, indicating whether the device met the implied performance standards.

Details of the Studies:

2. Sample sizes used for the test set and the data provenance

• Intra-Assay Precision:
  • Test Set Sample Size: 3 controls (Serotest S-18, 3 Lyphocheck controls, 3 patient pools) with 20 replicates each. Total: 7 samples * 20 replicates = 140 measurements.
  • Data Provenance: Not explicitly stated, but derived from "commercially-available Lyphocheck controls" and "patient pool controls" which suggests laboratory-generated data, likely in a controlled setting.
• Inter-Assay Precision:
  • Test Set Sample Size: 7 controls (Serotest S-18, 3 Lyphocheck controls, 3 patient pools) tested across 10 separate assays.
  • Data Provenance: Not explicitly stated, but results from "different operators" conducting assays, indicating laboratory-generated data.
• Recovery Studies:
  • Test Set Sample Size: 3 patient plasma pools, each spiked with 50 ul of bovine serum or 50 ul of testosterone solution at 4 different concentrations. The number of replicates for each spike value is n=3.
  • Data Provenance: "randomly selected patient samples" and laboratory-prepared solutions.
• Dilution Studies:
  • Test Set Sample Size: 3 patient samples, serially diluted up to a factor of 32.
  • Data Provenance: "patient samples" and a "low (0.70 ng/mL) testosterone standard," suggesting laboratory-generated data.
• Sensitivity Study:
  • Test Set Sample Size: 20 replicates of the zero standard and 3 replicates of the first standard (0.25 ng/mL).
  • Data Provenance: Laboratory-generated data ("BATCH 1 REAGENTS").
• Specificity Study:
  • Test Set Sample Size: 8 potentially cross-reactive analytes.
  • Data Provenance: Laboratory-generated data (measuring apparent response to specific analytes).
• Sex Hormone Binding Globulin Interference Study:
  • Test Set Sample Size: 5 serum pools, measured at 6 time points (0, 2, 4, 8, 16, and 32 minutes) after heating. (n=5 for calculating mean +/- 1 S.D.).
  • Data Provenance: "Five serum pools," likely from patient samples, and laboratory-induced conditions (heating).
• Variation of Recommended Assay Procedures Study:
  • Test Set Sample Size: 3 controls and 3 pools of sera.
  • Data Provenance: Controls and "pools of sera" (likely patient samples), tested under varied laboratory conditions.
• Gas Chromatography Mass Spectroscopy Tested Samples Comparison:
  • Test Set Sample Size: 13 control sera samples (some with A/B identifiers).
  • Data Provenance: "Several control sera previously checked by Gas Chromatography Mass Spectroscopy in Germany." This indicates retrospective use of samples with a highly reliable reference method.
• Correlation Studies:
  • Test Set Sample Size:
    • 120 samples from males.
    • 63 samples from females.
    • 60 samples from pre-puberal children.
  • Data Provenance: "samples from males," "samples from females," "samples from pre-puberal children." Specific country of origin is not stated, but compared against DPC Testosterone Kit and an "old Testosterone Maia Kit," which are commercial products. This suggests retrospective or prospective testing of patient samples.
• Expected Values Study (Reference Range):
  • Test Set Sample Size:
    • 659 serum samples from males (aged 18 to 71 years).
    • 300 serum samples from females.
  • Data Provenance: "serum samples from males," "serum samples from females." Country of origin not stated, but these are patient samples used to establish population-based reference ranges.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

The document describes in vitro diagnostic device performance studies, not clinical decision-making based on image interpretation or complex diagnoses where expert consensus is typically used for ground truth. The "ground truth" here is the actual concentration of testosterone or the known properties of the samples/analytes.

• For precision, recovery, dilution, sensitivity, specificity, and assay variation studies, the ground truth is established by the analytical methods themselves, known concentrations of standards, or the chemical properties of the substances. No human experts are involved in establishing this ground truth beyond setting up the experiment and analyzing the results according to established protocols.
• For the Gas Chromatography Mass Spectroscopy (GCMS) Tested Samples Comparison, GCMS itself serves as the "gold standard" or highly reliable ground truth method. No specific number or qualification of human experts involved in the GCMS analysis is provided, but GCMS is an instrumental analytical technique.
• For the Correlation Studies, the ground truth is essentially the results obtained by the predicate device (DPC Testosterone Kit) and the previous version of the device (old Testosterone Maia Kit), which are established diagnostic assays. There's no mention of human experts defining the "truth" for these comparative analyses.

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

Since this is an in vitro diagnostic device for quantitative measurement of testosterone, the concepts of "adjudication" by multiple human experts (e.g., 2+1, 3+1) typically used in medical image analysis or complex clinical diagnosis studies do not apply. The performance is assessed by comparing quantitative results against reference methods, known concentrations, or established statistical metrics (e.g., CV%, recovery %, correlation coefficients).

5. 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

No, a multi-reader multi-case (MRMC) comparative effectiveness study was not conducted. This type of study is relevant for evaluating diagnostic aids where human readers interpret medical images or data, and an AI system might assist them. The Biodata Testosterone Maia Kit is an in vitro diagnostic assay that produces a quantitative measurement, not an AI-powered interpretive tool for human readers.

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

Yes, the studies presented are all standalone performance evaluations of the Biodata Testosterone Maia Kit. The device (a radioimmunoassay kit) operates as an "algorithm only" in the sense that it performs a chemical reaction and measurement process to yield a numerical result. Human involvement is in operating the laboratory equipment and analyzing the raw count data, but the performance metrics (precision, accuracy, sensitivity, specificity, correlation) are inherent to the assay's chemical and measurement characteristics, not to human interpretation of its output in a "human-in-the-loop" clinical decision-making context.

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

The ground truth used varied by study type:

• Known concentrations/values: For precision, sensitivity, recovery, dilution, and specificity studies, the ground truth was based on analytically prepared standards with known concentrations of testosterone or other analytes.
• Gold Standard Analytical Method: For the Gas Chromatography Mass Spectroscopy Tested Samples Comparison, the results from GCMS were considered the ground truth.
• Predicate Device/Reference Method: For the Correlation Studies, the results from the DPC Testosterone Kit (the predicate device) and the old Testosterone Maia Kit served as the reference for comparison.
• Biological Samples: For studies establishing reference ranges (Expected Values Study) and those using "patient pools" or "patient samples," the ground truth was the actual presence and concentration of testosterone as measured by the device itself, validated through other performance metrics and comparison to a predicate.

8. The sample size for the training set

The document does not explicitly mention a "training set" in the context of machine learning or AI. This is a traditional in vitro diagnostic device. Therefore, there is no AI training set as such. The "training" of the assay refers to its development and optimization based on chemical principles and experimental results. The samples used for various performance evaluations (precision, recovery, etc.) could be considered as part of the overall development and validation, but not a distinct "training set" like in AI/ML.

9. How the ground truth for the training set was established

As there is no "training set" in the AI/ML sense, this question is not directly applicable. For a traditional diagnostic kit, the "ground truth" for its development is based on established biochemical principles, known concentrations of calibrators, and reference methods to ensure the assay accurately measures the target analyte (testosterone).

Ask a specific question about this device