Diagnostic Accuracy

Diagnostic accuracy studies estimate how well a test classifies disease—often via sensitivity, specificity, predictive values, or discrimination (AUC)—against a reference standard or outcome. Related work includes risk scores and prognostic biomarkers that forecast future events rather than confirm current disease.

A bedside assay with 95% sensitivity, a low-prevalence screening PPV, and a plasma protein that forecasts dementia years ahead are all “accuracy-ish,” but they answer different clinical questions and break in different ways.

Claim ledger

What the evidence shows

Drawn from 8 studies in this library. Mix labels say which citation roles are present; they are not a strength score. Supports means evidence for a finding; Challenges means evidence against a stated position; Qualifies marks scope.

• ○ Limited citations

  Phone pulse waveforms can detect atrial fibrillation accurately versus ECG in clinic samples.

  Smartphone photoplethysmography detected atrial fibrillation with sensitivity about 95.6% and specificity about 96.6% versus ECG in an analysed clinic sample (adequate signal required; pacing excluded).

  1 supporting · 2 qualifying

  Supports

  1. 1Smartphone PPG to detect atrial fibrillation

  Qualifies

  1. 1How well does CA125 find ovarian cancer in GP care?— different setting — low-prevalence GP screening PPV, not clinic AF vs ECG
  2. 2Blood GFAP and NfL predict dementia years ahead— different question — long-horizon prognosis, not point-of-care arrhythmia detection

  Study comparison

  Study	Role	Design	N	Population	Outcome
  Smartphone PPG to detect atrial fibrillation2019	Supports	Cross-sectionalDiagnostic accuracy vs 12-lead ECG in primary care	N=223 · 241 enrolled; 223 analysed after exclusions	Primary-care patients screened for atrial fibrillation with FibriCheck PPG	Sensitivity and specificity of mobile PPG for AF vs 12-lead ECG
  How well does CA125 find ovarian cancer in GP care?2020	Qualifiesdifferent setting — low-prevalence GP screening PPV, not clinic AF vs ECG	CohortPopulation-based primary-care EHR cohort	N=50780 · Women with CA125 tested in English primary care	Women having CA125 measured in English general practice	PPV, sensitivity, specificity, and AUC of CA125 ≥35 U/ml for ovarian cancer
  Blood GFAP and NfL predict dementia years ahead2024	Qualifiesdifferent question — long-horizon prognosis, not point-of-care arrhythmia detection	CohortUK Biobank prospective plasma GFAP/NfL with registry dementia follow-up	N=48542 · 1,312 incident all-cause dementia over ~13 years	UK Biobank adults with baseline Olink plasma GFAP and NfL	Incident all-cause dementia / ADRD HRs and prediction AUC for GFAP and NfL

• ○ Limited citations

  A multiplex assay can identify Staphylococcus from blood cultures with very high accuracy.

  A multiplex molecular assay on monomicrobial blood cultures identified Staphylococcus at genus level with sensitivity about 99.4% and specificity about 99.7% in a large evaluated set.

  1 supporting · 2 qualifying

  Supports

  1. 1Rapid gram-positive blood culture ID

  Qualifies

  1. 1Can a rapid test diagnose sickle cell at the bedside?— different target — sickle-cell bedside detection, not blood-culture ID
  2. 2Smartphone PPG to detect atrial fibrillation— different modality — PPG AF screening, not microbiology ID

  Study comparison

  Study	Role	Design	N	Population	Outcome
  Rapid gram-positive blood culture ID2013	Supports	OtherProspective diagnostic accuracy study of automated microarray panel vs reference culture	N=1157 · 1,157 prospectively collected monomicrobial positive blood cultures	Monomicrobial positive blood cultures	Sensitivity and specificity for gram-positive organisms and resistance markers
  Can a rapid test diagnose sickle cell at the bedside?2015	Qualifiesdifferent target — sickle-cell bedside detection, not blood-culture ID	OtherLateral-flow Sickle SCAN vs gold-standard diagnosis on duplicate patient samples	N=137 · Patient samples measured in duplicate	Clinical samples submitted for sickle-cell diagnosis	Agreement of a point-of-care sickle-cell test with the laboratory gold standard
  Smartphone PPG to detect atrial fibrillation2019	Qualifiesdifferent modality — PPG AF screening, not microbiology ID	Cross-sectionalDiagnostic accuracy vs 12-lead ECG in primary care	N=223 · 241 enrolled; 223 analysed after exclusions	Primary-care patients screened for atrial fibrillation with FibriCheck PPG	Sensitivity and specificity of mobile PPG for AF vs 12-lead ECG

• ○ Limited citations

  In GP care, CA125 looks accurate but a positive rarely means ovarian cancer when disease is rare.

  In GP care, CA125 ≥35 U/ml for ovarian cancer showed sensitivity about 77%, specificity about 93.8%, and AUC about 0.92, but PPV was only about 10.1% because incidence was about 0.9%—illustrating prevalence’s grip on predictive value.

  1 supporting · 2 qualifying

  Supports

  1. 1How well does CA125 find ovarian cancer in GP care?

  Qualifies

  1. 1Smartphone PPG to detect atrial fibrillation— higher-prevalence clinic AF sample — PPV/NPV not transportable as raw sensitivities
  2. 2Diabetes prevalence and risk-score accuracy— related lesson — risk-score AUCs with number-needed-to-screen trade-offs

  Study comparison

  Study	Role	Design	N	Population	Outcome
  How well does CA125 find ovarian cancer in GP care?2020	Supports	CohortPopulation-based primary-care EHR cohort	N=50780 · Women with CA125 tested in English primary care	Women having CA125 measured in English general practice	PPV, sensitivity, specificity, and AUC of CA125 ≥35 U/ml for ovarian cancer
  Smartphone PPG to detect atrial fibrillation2019	Qualifieshigher-prevalence clinic AF sample — PPV/NPV not transportable as raw sensitivities	Cross-sectionalDiagnostic accuracy vs 12-lead ECG in primary care	N=223 · 241 enrolled; 223 analysed after exclusions	Primary-care patients screened for atrial fibrillation with FibriCheck PPG	Sensitivity and specificity of mobile PPG for AF vs 12-lead ECG
  Diabetes prevalence and risk-score accuracy2008	Qualifiesrelated lesson — risk-score AUCs with number-needed-to-screen trade-offs	Cross-sectionalSUNSET Amsterdam population sample; ethnicity-stratified screening criteria	N=1434 · 339 Hindustani Surinamese, 605 African Surinamese, 490 Dutch	Amsterdam adults aged 35–60 (Hindustani Surinamese, African Surinamese, Dutch)	Diabetes prevalence and risk-score AUC / numbers-needed-to-screen by ethnicity

• ○ Limited citations

  Diabetes risk scores trade accuracy against how many people you must screen locally.

  Diabetes risk scores in an Amsterdam sample showed AUCs about 0.74–0.80 with numbers-needed-to-screen of 3–7 depending on the outcome definition—accuracy metrics tied to local prevalence and cut-offs.

  1 supporting · 2 qualifying

  Supports

  1. 1Diabetes prevalence and risk-score accuracy

  Qualifies

  1. 1How well does CA125 find ovarian cancer in GP care?— different disease and prevalence regime — ovarian cancer screening PPV
  2. 2Blood GFAP and NfL predict dementia years ahead— different task — prognostic plasma proteins, not cross-sectional diabetes scoring

  Study comparison

  Study	Role	Design	N	Population	Outcome
  Diabetes prevalence and risk-score accuracy2008	Supports	Cross-sectionalSUNSET Amsterdam population sample; ethnicity-stratified screening criteria	N=1434 · 339 Hindustani Surinamese, 605 African Surinamese, 490 Dutch	Amsterdam adults aged 35–60 (Hindustani Surinamese, African Surinamese, Dutch)	Diabetes prevalence and risk-score AUC / numbers-needed-to-screen by ethnicity
  How well does CA125 find ovarian cancer in GP care?2020	Qualifiesdifferent disease and prevalence regime — ovarian cancer screening PPV	CohortPopulation-based primary-care EHR cohort	N=50780 · Women with CA125 tested in English primary care	Women having CA125 measured in English general practice	PPV, sensitivity, specificity, and AUC of CA125 ≥35 U/ml for ovarian cancer
  Blood GFAP and NfL predict dementia years ahead2024	Qualifiesdifferent task — prognostic plasma proteins, not cross-sectional diabetes scoring	CohortUK Biobank prospective plasma GFAP/NfL with registry dementia follow-up	N=48542 · 1,312 incident all-cause dementia over ~13 years	UK Biobank adults with baseline Olink plasma GFAP and NfL	Incident all-cause dementia / ADRD HRs and prediction AUC for GFAP and NfL

• ○ Limited citations

  Higher blood GFAP and NfL associate with later dementia risk — prognosis, not same-day diagnosis.

  In UK Biobank, higher baseline plasma GFAP and NfL were associated with roughly doubled hazards of incident dementia over ~13 years, with incremental discrimination when added to demographic risk scores—prognostic biomarker performance, not a bedside diagnostic rule-in test.

  1 supporting · 2 qualifying

  Supports

  1. 1Blood GFAP and NfL predict dementia years ahead

  Qualifies

  1. 1Smartphone PPG to detect atrial fibrillation— different task — detecting current AF vs ECG, not forecasting dementia
  2. 2Rising frailty flags one-year death risk— related prognostic framing — frailty trajectory vs death, not plasma NfL

  Study comparison

  Study	Role	Design	N	Population	Outcome
  Blood GFAP and NfL predict dementia years ahead2024	Supports	CohortUK Biobank prospective plasma GFAP/NfL with registry dementia follow-up	N=48542 · 1,312 incident all-cause dementia over ~13 years	UK Biobank adults with baseline Olink plasma GFAP and NfL	Incident all-cause dementia / ADRD HRs and prediction AUC for GFAP and NfL
  Smartphone PPG to detect atrial fibrillation2019	Qualifiesdifferent task — detecting current AF vs ECG, not forecasting dementia	Cross-sectionalDiagnostic accuracy vs 12-lead ECG in primary care	N=223 · 241 enrolled; 223 analysed after exclusions	Primary-care patients screened for atrial fibrillation with FibriCheck PPG	Sensitivity and specificity of mobile PPG for AF vs 12-lead ECG
  Rising frailty flags one-year death risk2018	Qualifiesrelated prognostic framing — frailty trajectory vs death, not plasma NfL	CohortEnglish primary-care EHR; monthly eFI trajectories before death vs matched controls	N=26298 · 13,149 decedents and 13,149 matched controls	English primary-care patients with electronic frailty index trajectories	12-month mortality risk by rapidly rising vs stable frailty class

• ● Supporting studies only

  Other tools (sickle cell, pain subtype, frailty trajectories) answer their own accuracy questions.

  Other validated tools in this set include a point-of-care sickle-cell device detecting HbS/HbC at low percentages suitable for neonates, StEP pinprick signs highly sensitive/specific for neuropathic vs non-neuropathic pain in specialist clinics, and rapidly rising electronic frailty index trajectories associated with doubled 12-month mortality versus stable frailty.

  3 supporting · 2 qualifying

  Supports

  1. 1Can a rapid test diagnose sickle cell at the bedside?
  2. 2StEP pain-subtype assessment
  3. 3Rising frailty flags one-year death risk

  Qualifies

  1. 1How well does CA125 find ovarian cancer in GP care?— different prevalence and pathway — GP ovarian-cancer screening
  2. 2Blood GFAP and NfL predict dementia years ahead— different prognostic biomarker — plasma proteins for dementia

  Study comparison

  Study	Role	Design	N	Population	Outcome
  Can a rapid test diagnose sickle cell at the bedside?2015	Supports	OtherLateral-flow Sickle SCAN vs gold-standard diagnosis on duplicate patient samples	N=137 · Patient samples measured in duplicate	Clinical samples submitted for sickle-cell diagnosis	Agreement of a point-of-care sickle-cell test with the laboratory gold standard
  StEP pain-subtype assessment2009	Supports	Cross-sectionalTwo-part tool development then independent clinic validation of StEP	N=137 · Part 2 validation after exclusions; Part 1 development n=187	Specialist clinic patients with chronic low back pain	Discrimination of radicular vs axial / neuropathic vs non-neuropathic LBP (StEP signs)
  Rising frailty flags one-year death risk2018	Supports	CohortEnglish primary-care EHR; monthly eFI trajectories before death vs matched controls	N=26298 · 13,149 decedents and 13,149 matched controls	English primary-care patients with electronic frailty index trajectories	12-month mortality risk by rapidly rising vs stable frailty class
  How well does CA125 find ovarian cancer in GP care?2020	Qualifiesdifferent prevalence and pathway — GP ovarian-cancer screening	CohortPopulation-based primary-care EHR cohort	N=50780 · Women with CA125 tested in English primary care	Women having CA125 measured in English general practice	PPV, sensitivity, specificity, and AUC of CA125 ≥35 U/ml for ovarian cancer
  Blood GFAP and NfL predict dementia years ahead2024	Qualifiesdifferent prognostic biomarker — plasma proteins for dementia	CohortUK Biobank prospective plasma GFAP/NfL with registry dementia follow-up	N=48542 · 1,312 incident all-cause dementia over ~13 years	UK Biobank adults with baseline Olink plasma GFAP and NfL	Incident all-cause dementia / ADRD HRs and prediction AUC for GFAP and NfL

Debates

Tensions and limits

Some items are genuine disagreements on the same question. Others mark different assays, populations, or outcomes.

Timeline

How understanding moved

Study years are when the paper was published. Evidence edits are dated changes to this page's claims. Explanations are when PaperFren added a Discovery — not a claim that the science happened that day.

Papers

8 studies in this library bear on Diagnostic Accuracy, ordered by citations.

• StEP pain-subtype assessment

  A structured interview-plus-exam tool (StEP) was developed and validated to separate neuropathic/radicular from non-neuropathic axial low-back pain.

  PLoS medicine · 2009 · 190 citations

• How well does CA125 find ovarian cancer in GP care?

  In UK primary care, CA125 ≥35 U/ml had high NPV but only about 10% PPV for ovarian cancer, and elevated values also flagged other cancers.

  PLoS medicine · 2020 · 102 citations

• Blood GFAP and NfL predict dementia years ahead

  In 48,542 UK Biobank participants followed ~13 years, elevated plasma GFAP and NfL preceded dementia up to 15 years and improved prediction beyond CAIDE/DRS risk scores (AUC up to ~0.89).

  BMC medicine · 2024 · 98 citations

• Smartphone PPG to detect atrial fibrillation

  FibriCheck PPG showed ~96% sensitivity and specificity versus cardiologist ECG on analysable traces.

  JMIR mHealth and uHealth · 2019 · 93 citations

• Rising frailty flags one-year death risk

  Rapid monthly rises in the electronic frailty index marked older adults twice as likely to die within a year, with high specificity.

  BMC medicine · 2018 · 89 citations

• Diabetes prevalence and risk-score accuracy

  Hindustani Surinamese had ~26% diabetes prevalence; a clinical risk score showed moderate-to-good AUCs (~0.74–0.80) across ethnic groups.

  BMC public health · 2008 · 87 citations

• Rapid gram-positive blood culture ID

  A microarray panel identified Staphylococcus from positive blood cultures with ~99% sensitivity and specificity versus culture.

  PLoS medicine · 2013 · 86 citations

• Can a rapid test diagnose sickle cell at the bedside?

  Sickle SCAN, a point-of-care immunoassay, was validated against laboratory gold standards for detecting hemoglobin S and related genotypes.

  BMC medicine · 2015 · 81 citations

Questions

What is still open

1. Point-of-care classification accuracy (AF, blood culture ID, sickle cell, pain subtype) is not the same claim as low-prevalence screening PPV or long-horizon prognostic discrimination. Mixing sensitivities with hazards or PPVs invents a false head-to-head.

   Watch this

2. High sensitivity/specificity in a selected clinic sample does not guarantee useful predictive values in a low-prevalence screening population—the CA125 GP analysis is the cautionary case in this library.

   Watch this

Ask PaperFren about Diagnostic Accuracy