Pith. sign in

REVIEW 2 major objections 6 minor 28 references

A chart review process aided by natural language processing and multi-wave adaptive sampling to expedite validation of code-based algorithms for large database studies

T0 review · 2 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that pairing NLP-assisted annotation with multi-wave adaptive sampling and Bayesian stopping rules can cut per-chart review time by about 40% and avoid reviewing most charts in a validation study, as demonstrated in a…

desk verdict Adaptive sampling demonstration is solid; the NLP time-savings claim needs reanalysis before it can be used. read the letter →

arxiv 2507.22943 v1 pith:BMCY7DZM submitted 2025-07-25 cs.CL stat.ME

Shirley V Wang , Georg Hahn , Sushama Kattinakere Sreedhara , Mufaddal Mahesri , Haritha S. Pillai , Rajendra Aldis , Joyce Lii , Sarah K. Dutcher
show 8 more authors
Rhoda Eniafe Jamal T. Jones Keewan Kim Jiwei He Hana Lee Sengwee Toh Rishi J Desai Jie Yang
This is my paper
classification cs.CLstat.ME
keywords chartreviewnaturallanguageprocessingmulti-waveadaptivesamplingBayesianstoppingrulepositivepredictivevaluevalidationstudyintentionalself-harmclaimsdatabase
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes and demonstrates an expedited chart validation process for code-based algorithms used in large healthcare databases. The process combines two efficiency mechanisms: natural language processing (NLP) that highlights relevant portions of electronic health record notes, reducing the time a human annotator spends per chart, and multi-wave adaptive sampling with a predefined Bayesian stopping rule, which halts review once the primary performance metric is estimated with enough precision. In the case study, validating a claims-based screening algorithm for intentional self-harm in patients with obesity, the authors report that NLP-assisted review took about 40% less time per chart, and that applying the stopping rule would have prevented review of 77% of the 530 sampled charts. The estimated positive predictive value at the stopping point was 0.60 with a 95% credible interval of (0.47, 0.72), essentially unchanged after full review, supporting the claim that precision loss is limited. If the process holds up, validation studies could become routine enough to support quantitative bias analyses in database research.

What carries the argument

The process rests on three linked components. First, study participants are stratified by whether they are claims-positive for the outcome, whether their EHR notes contain a relevant UMLS concept unique identifier (CUI) found by MetaMap, and whether state death records indicate suicide; patients negative on both claims and EHR are treated as true negatives, and death-record suicides are treated as true positives without annotation. Second, an open-source NLP-aided annotation tool called CORA loads linked claims and notes in chronological order and highlights text matching the self-harm CUI list, directing annotator attention and cutting review time. Third, sequential batches of ten charts, five claims-positive and five claims-negative, are annotated, and a $\mathrm{Beta}(1,1)$ prior is updated with the observed binomial counts to form a posterior distribution for the positive predictive value; the stopping rule stops for success when the lower bound of the 95% credible interval exceeds 0.75 and for futility when the upper bound falls below 0.75. Inverse sampling weights convert the stratified sample into estimates of negative predictive value, sensitivity, and specificity for the full cohort.

What would settle it

One could settle the central claim by linking out-of-network records or state death data for the unreviewed claims-positive patients with no health-system contact and computing their true-positive rate; if it differs materially from the 0.60 PPV seen in reviewed charts, the generalizable PPV and any bias adjustment built on it fail. A second check would annotate a random sample of the claims-negative, EHR-negative group to test the assumption that they are all true negatives.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that an NLP-assisted, adaptively sampled chart review can validate a claims-based outcome algorithm with substantially less human effort than full manual review. The central empirical claims are that annotators using the NLP-highlighting interface took a median 6.0 minutes per chart versus 11.4 minutes without it, a roughly 40 to 48 percent reduction; that cumulative Bayesian estimates of positive predictive value met a predefined futility stopping rule after 12 batches of 10 charts, only 23% of the full sample; and that the inverse-weighted sensitivity, specificity, and negative predictive value at that stopping point were consistent with the values after all 530 charts were reviewed, although sensitivity intervals remained wide. The paper also characterizes the reasons behind false-positive and false-negative annotations, showing that many false positives arose from mental-status assessments recording no suicidal thoughts and from distant past history, and it argues that these gains make routine validation of algorithms for database studies more feasible.

Load-bearing premise

The load-bearing premise is that claims-positive patients with no contact at the linked health system around the outcome date have the same true-positive rate as those who were seen there, even though the former group cannot be reviewed; if their documentation differs, the reported PPV is biased.

Editorial extensions

If this is right

  • If the method is applied to other code-based algorithms, validation studies that currently require hundreds of hours of chart review could be completed after reviewing only a fraction of the planned charts.
  • The reported PPV of 0.60 for the intentional self-harm screening algorithm can feed quantitative bias analyses for database studies of self-harm in patients with obesity instead of relying on unvalidated code.
  • For algorithms that are clearly high-performing or clearly poor, the stopping rule yields the largest savings; gains shrink for algorithms whose performance hovers near the threshold.
  • The false-positive profile gives concrete code-refinement targets, such as excluding notes that only document absence of suicidal thoughts and distinguishing distant past history from a current event.
  • The framework is tool-agnostic, so newer NLP approaches, including large language models, can be substituted for the MetaMap-based highlighting without changing the sampling or stopping machinery.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The efficiency gain is likely to carry over to other outcomes only when the keyword or CUI list has high recall; an incomplete list would create silent false negatives in the unreviewed 'EHR-negative' stratum, so the process should be accompanied by a sensitivity analysis that annotates a sample of that stratum.
  • The NLP-aided versus unaided timing comparison used 40 charts reviewed by different annotators, so part of the measured time difference could reflect between-annotator speed; a within-annotator randomized crossover would isolate the NLP effect.
  • Because the algorithm being validated is fixed during the process, any future adaptive tuning of the code list between waves would need validation on a fresh sample to avoid overfitting to the reviewed charts.
  • The same machinery could be extended to estimate performance separately by exposure or comparator group, which would make the resulting measurement characteristics more directly usable in exposure-specific quantitative bias analyses.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 6 minor

Summary. The paper proposes a two-part process to expedite validation of code-based algorithms in large database studies: (1) NLP highlighting of relevant text in EHR notes to speed up manual chart review, and (2) multi-wave adaptive sampling with a Bayesian stopping rule to reduce the number of charts reviewed. The method is illustrated in a case study validating a claims-based ICD-10-CM screening algorithm for intentional self-harm among 62,129 patients with obesity in Mass General Brigham (MGB) linked claims-EHR data. The authors report that NLP assistance reduced per-chart review time by 40% (abstract) or 48% (discussion) and that a predefined futility stopping rule, met after 120 of 530 charts, would have prevented review of 77% of charts with limited compromise to the PPV estimate.

Significance. If the efficiency gains are real, the proposed process could substantially lower the cost of chart validation and make routine algorithm validation more feasible, which would strengthen quantitative bias analyses in database studies. The paper has notable strengths: the proposed workflow is clearly specified; the full-sample comparison of PPV at stopping versus after all 530 charts supports the soundness of the stopping-rule decision; the open-source CORA tool, protocol, analytic code, and de-identified annotation inputs are made available; and the authors explicitly acknowledge key limitations such as the untestable assumption about claims-positive patients without MGB contact. The main gap is that the NLP time-savings claim, which is one of the two pillars of the expedited-review argument, is not statistically established and is internally inconsistent between the abstract and the discussion.

major comments (2)
  1. [Chart Review Process, step 5; Results] The NLP time-reduction claim is not supported by a valid statistical comparison. The design is described as follows: each annotator reviewed 20 charts without NLP highlights that had previously been reviewed by the other annotator with NLP highlights, giving 40 charts with recorded timing in both conditions. Because each chart is reviewed by a different annotator under the two conditions, annotator identity is perfectly confounded with NLP condition within each chart. No paired test, confidence interval, or mixed model with annotator fixed effects is reported, and the statement that the two annotators had 'similar timing' during training is not a formal test of exchangeability. Furthermore, the abstract's 40% reduction appears to be obtained by comparing the overall NLP median of 7.0 minutes with the no-NLP subset median of 11.4 minutes, whereas the Results report within-subset medians of 6.0 minutes (NLP) versus 11.4 minutes (no NLP), which give a 48% reduction, the number used in the Discussion. This is an apples-to-oranges comparison and an internal inconsistency. The authors should present a paired analysis that separates annotator effects from NLP effects, report an effect estimate with a confidence interval or a posterior interval, and correct the abstract's percentage so that it matches the analysis actually performed.
  2. [Figure 2; Discussion, Limitations] The PPV estimate used for the case study is based on claims-positive patients with MGB healthcare contact near the outcome date (Figure 2, boxes f/g), and the authors assume that the proportion of true positives is the same among claims-positive patients without MGB contact (boxes d/e). This assumption is explicitly acknowledged as necessary, but it is load-bearing for the generalizability of the case-study PPV and for the stopping-rule decision, since the reviewed sample excludes the d/e subgroup entirely. The paper provides no sensitivity analysis or external benchmark to bound the possible bias. The authors should either present a sensitivity analysis under plausible scenarios (e.g., differential PPV in d/e) or clearly restrict the PPV claim to the subpopulation with linked EHR contact, and temper the conclusion that the algorithm's PPV is approximately 0.63.
minor comments (6)
  1. [Results] Please clarify which median (7.0 vs 6.0 minutes) corresponds to which set of charts, and explicitly derive the percentage reduction reported in the abstract (40%) from the results; as written, the abstract number cannot be reproduced from the within-subset comparison.
  2. [Discussion] The statement 'reduced the amount of time to annotate charts by 48%' should be reconciled with the abstract's 40%; if the correct within-subset estimate is 48%, the abstract should be corrected, and if the intended comparison uses the overall median, the denominator and sample should be stated explicitly.
  3. [Chart Review Process, step 5] The claim that the two annotators had 'similar timing in terms of review' during training should be supported by quantitative summaries (medians, ranges, or a formal test), since this statement is used to argue that annotator speed does not confound the NLP comparison.
  4. [Results] The report of Cohen's kappa equal to 100% after 30 charts would be more informative with the raw number of agreements and a confidence interval; a perfect kappa in a small sample is not strong evidence of interchangeability.
  5. [Table 2] The sensitivity estimate changes from 0.93 at the stopping point to 0.82 after the full sample, with both confidence intervals very wide; please state explicitly whether this difference is within the expected sampling variability, given that the stopping rule was based only on PPV.
  6. [References] Reference 10 (Wright et al., JAMIA 2011) does not appear to describe multi-wave adaptive sampling and may be a citation error; please verify that this reference supports the statement about multi-wave sampling methods.

Circularity Check

1 steps flagged · score 2.0 of 10

No substantive circularity: the case-study estimates are empirical and self-contained, with one minor self-citation used to justify the sampling design.

  1. other [Methods, Chart Review Process, Step 4 (selection of random sample batch; page 9 of manuscript)]
    "In parallel work, we have demonstrated that this approach balances simplicity and efficiency compared to more complex sampling strategies in a multi-wave setting, across a variety of scenarios.17"

    This is the stated justification for the chosen multi-wave sampling design (batches of k=10 with five claims+ and five claims- per batch). Reference 17 is the authors' own parallel work (Hahn, Schneeweiss, Wang), so the design choice is imported from a self-citation rather than from an independent external derivation. However, this citation is not load-bearing for the paper's central empirical results: the PPV stopping outcome (0.60; 0.47-0.72), the 77% of charts not reviewed under the stopping rule, and the NLP timing comparison are all measured or computed from the collected annotations and prespecified thresholds, not derived from reference 17. The circularity is therefore minor.

full rationale

The paper's central claims are empirical and do not reduce to their inputs by construction. The futility stopping result follows directly from the prespecified Beta-Binomial rule (upper credible bound below 0.75) applied to the accumulated annotations: after 12 batches the 95% credible interval was (0.47, 0.72), so stopping is a straightforward application of the stated threshold. The '77% of charts prevented' figure is arithmetic from the completed full review (1 - 120/530), not a fitted target. The NLP time comparison is an empirical measurement reported as median 6.0 minutes with NLP assistance versus 11.4 minutes without; the discrepancy between 40% (abstract) and 48% (discussion) is an internal consistency or reporting issue for the correctness pass, not a circularity issue. The one element that leans on prior work is the choice of the multi-wave sampling design, which is justified by the authors' own parallel simulation paper (reference 17); this is a minor self-citation but does not feed into the estimated performance metrics. The explicit assumption that PPV among claims-positive patients without MGB contact equals that among patients with MGB contact is a stated, testable limitation rather than a circular derivation. Overall, no step of the derivation is equivalent to its inputs by definition.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

No new entities are postulated. The CORA tool is open-source prior work, and the process is a workflow rather than a new theoretical object.

free parameters (3)
  • PPV stopping threshold = 0.75
    The success/futility boundary for the Bayesian credible interval; chosen by the authors, not derived from data. It directly determines when the study stops and therefore the 77% chart savings.
  • Batch size k = 10
    Each wave samples 10 charts, split 5 claims+ and 5 claims-; a design choice that affects precision trajectory and stopping time.
  • Review window for notes = +/-60 days around the outcome date
    Notes within this window are used for annotation; a window choice that could miss events if documentation falls outside it.
assumptions (4)
  • domain assumption Patients in group 1 (claims negative and EHR negative) are truly negative for intentional self-harm.
    Assumed in Step 3 and Figure 2 without chart review; if MetaMap misses concepts, some true cases would be mislabeled as negative, biasing sensitivity and NPV. Acknowledged as a limitation in the Discussion.
  • domain assumption Patients with cause of death coded as suicide (group 3) are truly positive.
    Assumed in Step 3 and Figure 2; death records are treated as a perfect reference standard for this group.
  • domain assumption The proportion of true positives among claims+ patients is equal for those with and without MGB healthcare contact around the outcome date.
    Stated in the Discussion as a necessary assumption; used to generalize the PPV from reviewed charts to all claims+ patients. This is the weakest load-bearing premise.
  • domain assumption The reference standard labels assigned by annotators are correct.
    Inter-annotator agreement (kappa=100% after 30 charts) is not validation against an external gold standard; the EHR notes are the only source for the reference standard.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A chart review process aided by natural language processing and multi-wave adaptive sampling to expedite validation of code-based algorithms for large database studies." pith.science (2026). https://pith.science/paper/BMCY7DZM

@misc{pith2026250722943,
  author       = {Pith},
  title        = {Pith review of: A chart review process aided by natural language processing and multi-wave adaptive sampling to expedite validation of code-based algorithms for large database studies},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BMCY7DZM}},
  note         = {Machine review of arXiv:2507.22943}
}
read the original abstract

Background: One of the ways to enhance analyses conducted with large claims databases is by validating the measurement characteristics of code-based algorithms used to identify health outcomes or other key study parameters of interest. These metrics can be used in quantitative bias analyses to assess the robustness of results for an inferential study given potential bias from outcome misclassification. However, extensive time and resource allocation are typically re-quired to create reference-standard labels through manual chart review of free-text notes from linked electronic health records. Methods: We describe an expedited process that introduces efficiency in a validation study us-ing two distinct mechanisms: 1) use of natural language processing (NLP) to reduce time spent by human reviewers to review each chart, and 2) a multi-wave adaptive sampling approach with pre-defined criteria to stop the validation study once performance characteristics are identified with sufficient precision. We illustrate this process in a case study that validates the performance of a claims-based outcome algorithm for intentional self-harm in patients with obesity. Results: We empirically demonstrate that the NLP-assisted annotation process reduced the time spent on review per chart by 40% and use of the pre-defined stopping rule with multi-wave samples would have prevented review of 77% of patient charts with limited compromise to precision in derived measurement characteristics. Conclusion: This approach could facilitate more routine validation of code-based algorithms used to define key study parameters, ultimately enhancing understanding of the reliability of find-ings derived from database studies.

Figures

Figures reproduced from arXiv: 2507.22943 by the authors.

Figure 1
Figure 1. Process diagram for natural language processing assisted chart review with adaptive, multi-wave sampling [PITH_FULL_IMAGE:figures/full_fig_p018_1.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

28 extracted references · 24 canonical work pages

  1. [1]

    Food and Drug Administration: Framework for FDA's Real-World Evidence Pro- gram

    U.S. Food and Drug Administration: Framework for FDA's Real-World Evidence Pro- gram. https://www.fda.gov/media/120060/download

    work page

  2. [2]

    General Principles on Plan, Design and Analysis of Pharmacoepidemiological Stud- ies That Utilize Real-World Data for Safety Assessment of Medicines

    ICH. General Principles on Plan, Design and Analysis of Pharmacoepidemiological Stud- ies That Utilize Real-World Data for Safety Assessment of Medicines. https://data- base.ich.org/sites/default/files/ICH_M14_Step3_DraftGuideline_2024_0521.pdf

    work page

  3. [3]

    The FDA's sentinel initiative--A comprehen- sive approach to medical product surveillance

    Ball R, Robb M, Anderson SA, Dal Pan G. The FDA's sentinel initiative--A comprehen- sive approach to medical product surveillance. Clin Pharmacol Ther. Mar 2016;99(3):265-8. doi:10.1002/cpt.320

    work page doi:10.1002/cpt.320 2016

  4. [4]

    Proposed Guidance Document: Study Protocols That Use Real-world Data

    CMS. Proposed Guidance Document: Study Protocols That Use Real-world Data. https://www.cms.gov/medicare-coverage-database/view/medicare-coverage-docu- ment.aspx?mcdid=39

    work page

  5. [5]

    Quantitative Bias Analysis in Regula- tory Settings - PubMed

    Lash TL, Fox MP, Cooney D, Lu Y, Forshee RA. Quantitative Bias Analysis in Regula- tory Settings - PubMed. American journal of public health. 2016 Jul;106(7)doi:10.2105/AJPH.2016.303199

    work page arXiv 2016

  6. [6]

    Desai RJ, Wang SV, Sreedhara SK, et al. Process guide for inferential studies using healthcare data from routine clinical practice to evaluate causal effects of drugs (PRINCIPLED): considerations from the FDA Sentinel Innovation Center. BMJ. 2024-02-12 2024:e076460. doi:10.1136/bmj-2023-076460

    work page doi:10.1136/bmj-2023-076460 2024

  7. [7]

    Natural language processing - PubMed

    Joshi A. Natural language processing - PubMed. Science (New York, NY). 09/13/1991;253(5025)doi:10.1126/science.253.5025.1242

    work page arXiv 1991

  8. [8]

    Using natural language processing to improve ef- ficiency of manual chart abstraction in research: the case of breast cancer recurrence - Pub- Med

    Carrell DS, Halgrim S, Tran D-T, et al. Using natural language processing to improve ef- ficiency of manual chart abstraction in research: the case of breast cancer recurrence - Pub- Med. American journal of epidemiology. 03/15/2014;179(6)doi:10.1093/aje/kwt441

    work page doi:10.1093/aje/kwt441 2014

Show all 28 references
  1. [9]

    Adaptive Validation Design: A Bayesian Ap- proach to Validation Substudy Design With Prospective Data Collection - PubMed

    Collin LJ, MacLehose RF, Ahern TP, et al. Adaptive Validation Design: A Bayesian Ap- proach to Validation Substudy Design With Prospective Data Collection - PubMed. Epidemiol- ogy (Cambridge, Mass). 2020 Jul;31(4)doi:10.1097/EDE.0000000000001209

    2020 doi

  2. [10]

    A method and knowledge base for automated in- ference of patient problems from structured data in an electronic medical record - PubMed

    Wright A, Pang J, Feblowitz JC, et al. A method and knowledge base for automated in- ference of patient problems from structured data in an electronic medical record - PubMed. Journal of the American Medical Informatics Association : JAMIA. 2011 Nov- Dec;18(6)doi:10.1136/amiaj...

    2011 doi

  3. [11]

    A systematic re- view of validated suicide outcome classification in observational studies - PubMed

    Swain RS, Taylor LG, Braver ER, Liu W, Pinheiro SP, Mosholder AD. A systematic re- view of validated suicide outcome classification in observational studies - PubMed. International journal of epidemiology. 10/01/2019;48(5)doi:10.1093/ije/dyz038

    2019 doi

  4. [12]

    Extreme obesity is associated with suicidal behavior and suicide attempts in adults: results of a population-based representative sample - PubMed

    Wagner B, Klinitzke G, Brähler E, Kersting A. Extreme obesity is associated with suicidal behavior and suicide attempts in adults: results of a population-based representative sample - PubMed. Depression and anxiety. 2013 Oct;30(10)doi:10.1002/da.22105

    2013 doi

  5. [13]

    Associations of obesity with psychiatric disor- ders and suicidal behaviors in a nationally representative sample - PubMed

    Mather AA, Cox BJ, Enns MW, Sareen J. Associations of obesity with psychiatric disor- ders and suicidal behaviors in a nationally representative sample - PubMed. Journal of psycho- somatic research. 2009 Apr;66(4)doi:10.1016/j.jpsychores.2008.09.008

    2009 doi

  6. [14]

    The FDA Sentinel Real World Evidence Data Enterprise (RWE‐DE)

    Desai Rishi J, Marsolo K, Smith J, et al. The FDA Sentinel Real World Evidence Data Enterprise (RWE‐DE). Pharmacoepidemiology and Drug Safety. 2024/10/01;33(10)doi:10.1002/pds.70028

    2024 doi

  7. [15]

    Accessed December 17, 2024, Browse Sentinel Documentation / Sentinel Routine Querying Tool Documentation - Sentinel Version Control System

    Sentinel Routine Querying System Overview. Accessed December 17, 2024, Browse Sentinel Documentation / Sentinel Routine Querying Tool Documentation - Sentinel Version Control System

    2024

  8. [16]

    An overview of MetaMap: historical perspective and recent ad- vances - PubMed

    Aronson AR, Lang F-M. An overview of MetaMap: historical perspective and recent ad- vances - PubMed. Journal of the American Medical Informatics Association : JAMIA. 2010 May- Jun;17(3)doi:10.1136/jamia.2009.002733 32

    2010 arXiv

  9. [17]

    Adaptive multi-wave sampling for efficient chart vali- dation

    Hahn G, Schneeweiss S, Wang S. Adaptive multi-wave sampling for efficient chart vali- dation. 2025/03/08;doi:10.48550/arXiv.2503.06308

    2025 doi arXiv

  10. [18]

    PyQt5 - Comprehensive Python Bindings for Qt v5

    Limited RC. PyQt5 - Comprehensive Python Bindings for Qt v5. https://pypi.org/pro- ject/PyQt5/

  11. [19]

    A Coefficient of Agreement for Nominal Scales

    Cohen J. A Coefficient of Agreement for Nominal Scales. Educational and Psychological Measurement. 1960-04;20(1)doi:10.1177/001316446002000104

    1960 doi

  12. [20]

    Weighted Chi Square: an Extension of the Kappa Method

    Cohen J. Weighted Chi Square: an Extension of the Kappa Method. Educational and Psychological Measurement. 1972-04;32(1)doi:10.1177/001316447203200106

    1972 doi

  13. [21]

    Weighted kappa: Nominal scale agreement provision for scaled disagreement or partial credit

    Cohen J. Weighted kappa: Nominal scale agreement provision for scaled disagreement or partial credit. Psychological Bulletin. 1968;70(4)doi:10.1037/h0026256

    1968 doi

  14. [22]

    Accessed April 14, 2025, 2025

    Basic HHS Policy for Protection of Human Research Subjects, 45 CFR §46.102(l)(2). Accessed April 14, 2025, 2025. https://www.ecfr.gov/current/title-45/subtitle-A/subchapter- A/part-46/subpart-A#46.102

    2025

  15. [23]

    A primer on quantita- tive bias analysis with positive predictive values in research using electronic health data

    Newcomer SR, Xu S, Kulldorff M, Daley MF, Fireman B, Glanz JM. A primer on quantita- tive bias analysis with positive predictive values in research using electronic health data. Journal of the American Medical Informatics Association : JAMIA. 2019 Jul 31;26(12)doi:10.1093/ja- ...

    2019 doi

  16. [24]

    Challenges to Classifying Suicidal Ideations, Communications, and Be- haviours

    Silverman MM. Challenges to Classifying Suicidal Ideations, Communications, and Be- haviours. International Handbook of Suicide Prevention. 2011;doi:10.1002/9781119998556.ch1

    2011 doi

  17. [25]

    Validation to correct for outcome misclassification bias

    Lanes S, Beachler DC. Validation to correct for outcome misclassification bias. Phar- macoepidemiology and Drug Safety. 2023/06/01;32(6)doi:10.1002/pds.5601

    2023 doi

  18. [26]

    doi:10.1007/978-3-030- 82673-4

    Applying Quantitative Bias Analysis to Epidemiologic Data. doi:10.1007/978-3-030- 82673-4

    doi

  19. [27]

    Common misconceptions about validation studies - Pub- Med

    Fox MP, Lash TL, Bodnar LM. Common misconceptions about validation studies - Pub- Med. International journal of epidemiology. 08/01/2020;49(4)doi:10.1093/ije/dyaa090

    2020 doi

  20. [28]

    Use of the positive predictive value to correct for disease misclas- sification in epidemiologic studies - PubMed

    Brenner H, Gefeller O. Use of the positive predictive value to correct for disease misclas- sification in epidemiologic studies - PubMed. American journal of epidemiology. 12/01/1993;138(11)doi:10.1093/oxfordjournals.aje.a116805

    1993 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.