Quad Screen Risk Calculator

Conditional quad AFP/hCG/uE3/inhibin-A four-marker log-MoM screening Gaussian-class likelihood calculation with independently validated full correlated covariance matrices and matched prior. Log-MoMadjustment, model applicability, target definition and clinical validation external; illustrative parameters are not a real screening model. No naive marker independence, diagnosis, individual outcome certainty or testing recommendation.

Description

Quad screening combines independently adjusted marker information with a calibrated probability model.

Quad screening combines independently adjusted marker information with a calibrated probability model.^[1]

A posterior probability depends on the supplied four-dimensional distributions and prior.

When to use Quad Screen Risk Calculator

Use this tool to calculate or review the supplied observations under the stated rules. Establish the observation date, units, measurement context and independently applicable clinical reference or calibrated model.

How to use Quad Screen Risk Calculator and understand its results

Establish the observation date, units, measurement context and independently applicable clinical reference or calibrated model.
Enter callerVerifiedSource, callerVerifiedContext, callerVerifiedAsOfDate, independentlyAdjustedLogMomValues, callerValidatedAffectedMean, callerValidatedUnaffectedMean, callerValidatedAffectedCovariance, callerValidatedUnaffectedCovariance, callerValidatedPriorProbability, using the field meanings and rules below, then run the tool.
Read result.conditionalPosteriorModelProbability, result.conditionalLogLikelihoodRatio. Use the units and sign conventions below.
Correct any reported input issue and run again. Before using a successful result, check the stated units, signs, reference scope and observation coverage.

Input field	Meaning	Value type	Requirement
callerVerifiedSource	Caller verified source; supplied assertion, not independently verified; output retention depends on the rules below	string	Required
callerVerifiedContext	Caller verified context; supplied assertion, not independently verified; output retention depends on the rules below	string	Required
callerVerifiedAsOfDate	Caller verified as of date; real Gregorian YYYY-MM-DD date for the supplied record	string	Required
independentlyAdjustedLogMomValues	Independently adjusted log mom values	list of number	Required
callerValidatedAffectedMean	Caller validated affected mean	list of number	Required
callerValidatedUnaffectedMean	Caller validated unaffected mean	list of number	Required
callerValidatedAffectedCovariance	Caller validated affected covariance	numeric matrix	Required
callerValidatedUnaffectedCovariance	Caller validated unaffected covariance	numeric matrix	Required
callerValidatedPriorProbability	Caller validated prior probability	number	Required

Supply nonblank caller source/context and a real Gregorian as-of date, then four independently adjusted log-MoM coordinates with matched affected/unaffected mean vectors and square covariance matrices. Require exact matrix symmetry and positive definiteness by Cholesky factorization, and a prior strictly between zero and one. Evaluate each Gaussian log density without the shared normalizing constant, subtract unaffected from affected, combine with prior log odds and apply a numerically stable logistic conversion. Return posterior probability and log likelihood ratio; provenance fields are validated but not returned. The tool does not derive MoM adjustments, train a model or validate its clinical calibration.

logDensity=−0.5×||L^−1(x−mean)||²−Σln(L_jj); logLR=logDensity_affected−logDensity_unaffected; posterior=logistic(log(prior/(1−prior))+logLR)

Use the matching field units and the exact boundary, ordering and reference rules above.

Output field	Meaning	Value type
result	Structured result; its nested fields below contain the actual calculation or selected reference	object
result.conditionalPosteriorModelProbability	Conditional affected-model posterior as a probability from zero to one under the supplied distributions and prior	number
result.conditionalLogLikelihoodRatio	Affected minus unaffected Gaussian log density; natural logarithm of the model likelihood ratio	number

json

{
  "input": {
    "callerVerifiedSource": "independently-validated-model/edition",
    "callerVerifiedContext": "quad AFP/hCG/uE3/inhibin-A four-marker log-MoM screening; target condition/cohort/assay/dating matched",
    "callerVerifiedAsOfDate": "2026-01-01",
    "independentlyAdjustedLogMomValues": [
      0,
      0,
      0,
      0
    ],
    "callerValidatedAffectedMean": [
      0,
      0,
      0,
      0
    ],
    "callerValidatedUnaffectedMean": [
      0,
      0,
      0,
      0
    ],
    "callerValidatedAffectedCovariance": [
      [
        1,
        0,
        0,
        0
      ],
      [
        0,
        1,
        0,
        0
      ],
      [
        0,
        0,
        1,
        0
      ],
      [
        0,
        0,
        0,
        1
      ]
    ],
    "callerValidatedUnaffectedCovariance": [
      [
        1,
        0,
        0,
        0
      ],
      [
        0,
        1,
        0,
        0
      ],
      [
        0,
        0,
        1,
        0
      ],
      [
        0,
        0,
        0,
        1
      ]
    ],
    "callerValidatedPriorProbability": 0.01
  },
  "output": {
    "result": {
      "conditionalPosteriorModelProbability": 0.01,
      "conditionalLogLikelihoodRatio": 0
    }
  }
}

Limitations and assumptions

Dating evidence, acquisition conditions and applicable clinical references remain independently established. The output describes supplied observations or a conditional model without diagnosing an outcome, choosing a dose or verifying clinical applicability.

Missing required fields, unknown fields, wrong types and violated rules return input issues. Numeric values and outputs must be finite. Lists and matrices must be dense, without missing cells. Numeric lists allow at most 10000 entries; ordinary string lists allow at most 128 strings of 128 UTF-16 units each; numeric matrices allow at most 128 rows and 128 columns, subject to narrower requirements above. Extremely large composite outputs are rejected.

Alternative or Complementary methods

For a related step, use Combined Screening Risk Calculator: Combined screening uses adjusted marker coordinates with a calibrated probability model. For a complementary comparison, use PAPP-A MoM Calculator: PAPP-A multiples of the median express an observed value relative to an applicable matched median.

References

Quad Screen Risk Calculator — tool.io

Similar or alternative tools

Combined Screening Risk Calculator
Combined screening uses adjusted marker coordinates with a calibrated probability model.
PAPP-A MoM Calculator
PAPP-A multiples of the median express an observed value relative to an applicable matched median.