Stewardship Prompts to Improve Antibiotic Selection for Pneumonia: The INSPIRE Randomized Clinical Trial.
Gohil SK, Septimus E, Kleinman K, Varma N, Avery TR, Heim L, Rahm R, Cooper WS, Cooper M, McLean LE, Nickolay NG, Weinstein RA, Burgess LH, Coady MH, Rosen E, Sljivo S, Sands KE, Moody J, Vigeant J, Rashid S, Gilbert RF, Smith KN, Carver B, Poland RE, Hickok J, Sturdevant SG, Calderwood MS, Weiland A, Kubiak DW, Reddy S, Neuhauser MM, Srinivasan A, Jernigan JA, Hayden MK, Gowda A, Eibensteiner K, Wolf R, Perlin JB, Platt R, Huang SS
- DOI
- 10.1001/jama.2024.6248
- Record issued
- 2026-08-16
- Engine
- 7.39.0
- Exported
- 2026-09-22
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/fc42ca9e-a7c9-4270-a58e-9dc4989bdbe9 is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×3−1.5★
- ClaimsEfficacy rests on an unvalidated surrogate endpoint−0.5★
- ClaimsTreatment effect not shown to be clinically meaningful−0.5★
- ReportingData & code availability partially met−0.25★
- The numeric-impossibility checks (GRIM/GRIMMER/DEBIT/SPRITE) did not run: 1 reported mean was read, and its group size is not stated where the value is printed (this source has no machine-readable table structure). These checks need the count the mean was averaged over, so none was performed.
- No data or code availability links were detected to verify.
- 01Efficacy rests on an unvalidated surrogate endpoint
The primary outcome is extended-spectrum antibiotic days of therapy, a process measure (antibiotic use) rather than a hard clinical outcome. The paper does not provide evidence linking reduced antibiotic use to improved clinical outcomes; safety outcomes (ICU transfer, length of stay) were unchanged, and no validated surrogate-to-clinical outcome link is established.
“The primary outcome was empiric (first 3 days of hospitalization) extended-spectrum antibiotic days of therapy.”
- 02Treatment effect not shown to be clinically meaningful
The reported 28.4% reduction in extended-spectrum antibiotic days of therapy is a relative reduction in a process measure, not anchored to a clinically meaningful threshold or patient benefit. The absolute reduction from 613.9 to 428.5 days per 1000 empiric days is not framed against a minimal clinically important difference, and safety outcomes were unchanged, so the clinical meaningfulness is unclear.
“a 28.4% (95% CI, 22.2%-34.1%; P <.001) significantly lower rate of empiric extended-spectrum days of therapy in the CPOE bundle group compared with routine stewardship.”
- 03Other integrity concern
Trial NCT03697070 was first submitted to ClinicalTrials.gov on 2018-10-03, after the registered study start date of 2018-10-01. Retrospective registration means the protocol and outcomes were not on the public record before the study ran, which is what prospective registration exists to establish.
NCT03697070
reviewer’s wording
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
This is a well-conducted and well-reported cluster-randomized trial with a clear premise, rigorous design, comprehensive demographic reporting, and appropriate statistical methods. The main weaknesses are a vague data availability statement and a minor lack of explicit blinding rationale. Several minor copyedit inconsistencies in reported numbers should be corrected.
Both reviewers independently scored all eight dimensions and agreed on every status; no divergence required reconciliation. The study is interventional (cluster-randomized trial). Non-applicable sub-criteria (e.g., species/strain, housing, IACUC) were excluded. The statistics verification covered only 6 tests/effect estimates; the rest are unverified. The citation check found no retracted or unresolved references.
Numerical inconsistencies
1 finding · worst lowValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
- Internal contradictions in the reported numbersAssessed
Recomputed 6 tests: 6 consistent, 0 inconsistent; 6 via agent-written checks.
- CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Primary outcome rate ratio p-value from CI
“The overall rate ratio (RR) when clustering by hospital and period was 0.72 (95% CI, 0.66- 0.78; P < .001) for the primary outcome”
Taken as given: The rate ratio is 0.72 with 95% CI 0.66-0.78.; The CI is two-sided at 95%.; The p-value is two-tailed.Method: Recomputed p-value from the reported rate ratio and 95% CI using the normal approximation for the log rate ratio.How we recomputed it: pCI(0.72, 0.66, 0.78, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Secondary outcome vancomycin rate ratio p-value from CI
“Vancomycin days - of - therapy ... 0.77 (0.71-0.83) <.001”
Taken as given: The rate ratio is 0.77 with 95% CI 0.71-0.83.; The CI is two-sided at 95%.; The p-value is two-tailed.Method: Recomputed p-value from the reported rate ratio and 95% CI using the normal approximation for the log rate ratio.How we recomputed it: pCI(0.77, 0.71, 0.83, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Secondary outcome antipseudomonal rate ratio p-value from CI
“Antipseudomonal days - of - therapy ... 0.68 (0.61-0.75) <.001”
Taken as given: The rate ratio is 0.68 with 95% CI 0.61-0.75.; The CI is two-sided at 95%.; The p-value is two-tailed.Method: Recomputed p-value from the reported rate ratio and 95% CI using the normal approximation for the log rate ratio.How we recomputed it: pCI(0.68, 0.61, 0.75, 1) - CONSISTENTreported p = .130 · recomputed p = .125Reviewers 1, 2Safety outcome length of stay hazard ratio p-value from CI
“Length-of-Stay ... 0.96 (0.91-1.01) 0.13”
Taken as given: The hazard ratio is 0.96 with 95% CI 0.91-1.01.; The CI is two-sided at 95%.; The p-value is two-tailed.Method: Recomputed p-value from the reported hazard ratio and 95% CI using the normal approximation for the log hazard ratio.How we recomputed it: pCI(0.96, 0.91, 1.01, 1) - CONSISTENTreported p = .620 · recomputed p = .617Reviewers 1, 2Safety outcome days to ICU transfer hazard ratio p-value from CI
“Days-to-ICU Transfers ... 1.04 (0.89-1.21) 0.62”
Taken as given: The hazard ratio is 1.04 with 95% CI 0.89-1.21.; The CI is two-sided at 95%.; The p-value is two-tailed.Method: Recomputed p-value from the reported hazard ratio and 95% CI using the normal approximation for the log hazard ratio.How we recomputed it: pCI(1.04, 0.89, 1.21, 1) - CONSISTENTreported p = .020 · recomputed p = .022Reviewers 1, 2Safety outcome days to antibiotic escalation hazard ratio p-value from CI
“Days-to-Antibiotic Escalations ... 0.82 (0.69-0.97) 0.02”
Taken as given: The hazard ratio is 0.82 with 95% CI 0.69-0.97.; The CI is two-sided at 95%.; The p-value is two-tailed.Method: Recomputed p-value from the reported hazard ratio and 95% CI using the normal approximation for the log hazard ratio.How we recomputed it: pCI(0.82, 0.69, 0.97, 1)
- lowinternal contradictionAbstract reports 48.1% men, while Table 1 shows 48.4% for baseline routine group; likely due to rounding or different denominator.
Abstract: '48.1% were men'; Table 1: 'Male 12,852 (48.4)'
Table 1reviewer’s wording - lowinternal contradictionDenominator for routine stewardship baseline group is 25,031 in text but 25,030 in one percentage calculation.
Receipt of any empiric extended-spectrum antibiotic was 51.5% (12,901/25,030) during the baseline period ... for the routine stewardship group
Resultsreviewer’s wording
Overstated conclusions
3 findings · worst highConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
- Efficacy rests on an unvalidated surrogate endpointAssessed
- Treatment effect not shown to be clinically meaningfulAssessed
- Conclusions only partially backed by the presented evidenceAssessed
8 major claims checked against the paper's own evidence: 1 only partially supported (evidence backs part of the claim; gaps or caveats remain); the rest adequately supported.
- partialReviewer 1The intervention was safe, with no increase in antibiotic escalation.Time to antibiotic escalation was longer in the CPOE group (HR 0.82, P=0.02), which could be interpreted as a safety concern, though the paper frames it as a positive finding.Evidence: Days-to-antibiotic escalation HR 0.82 (95% CI, 0.69-0.97; P=0.02).
“Time to antibiotic escalation (from standard- to extended-spectrum antibiotics) was 18.1% longer in the CPOE bundle group compared with the routine stewardship group (overall hazard ratio, 0.82 [95% CI, 0.69-0.97]; P = .02).”
Results - supportedReviewer 1CPOE prompts significantly reduced empiric extended-spectrum antibiotic use in hospitalized patients with pneumonia.The primary outcome shows a significant reduction with a rate ratio of 0.72 (95% CI, 0.66-0.78; P < .001), directly supporting the claim.Evidence: Primary outcome rate ratio 0.72 (95% CI, 0.66-0.78; P < .001) in Table 2.
“CPOE prompts promoting standard-spectrum antibiotics for patients at low risk of infection with MDROs significantly reduced empiric extended-spectrum antibiotic use in hospitalized patients with pneumonia by 28.4%”
Abstract - supportedReviewer 1The reduction in antibiotic use was achieved without increasing ICU transfers or length of stay.Safety outcomes showed no significant differences in ICU transfer or length of stay, supporting the claim.Evidence: Safety outcomes: length of stay HR 0.96 (95% CI, 0.91-1.01; P=0.13); ICU transfer HR 1.04 (95% CI, 0.89-1.21; P=0.62).
“without increasing intensive care unit transfers or length of stay”
Table 2 - supportedReviewer 1The reduction in antibiotic use was sustained during the COVID-19 pandemic.The paper reports that reductions continued during the last 4 months of the trial, which overlapped with the pandemic.Evidence: Figure 2A-B and text: 'These reductions continued during the COVID-19 pandemic in the last 4 months of the trial.'
“These reductions continued during the COVID-19 pandemic in the last 4 months of the trial”
Results - supportedReviewer 1The CPOE algorithm classified more than 96% of patients as low risk, and less than 2% of these had MDRO-positive cultures.The paper reports these percentages directly, supporting the claim.Evidence: Text: 'The INSPIRE algorithm classified more than 96% of patients with pneumonia in both stewardship groups as low risk; less than 2% of these patients subsequently had an MDRO-positive culture.'
The INSPIRE algorithm classified more than 96% of patients with pneumonia in both stewardship groups as low risk; less than 2% of these patients subsequently had an MDRO-positive culture
Resultsreviewer’s wording - supportedReviewer 2CPOE prompts significantly reduced empiric extended-spectrum antibiotic use by 28.4%.The primary outcome analysis shows a rate ratio of 0.72 (95% CI 0.66-0.78, P<.001), directly supporting the claim.Evidence: Primary outcome rate ratio 0.72 (95% CI 0.66-0.78; P<.001) indicating 28.4% reduction.
the group using CPOE prompts had a 28.4% reduction in empiric extended-spectrum days of therapy (rate ratio, 0.72 [95% CI, 0.66-0.78]; P < .001)
Abstractreviewer’s wording - supportedReviewer 2The reduction was achieved without increasing ICU transfers or length of stay.Safety outcomes showed no significant differences in ICU transfer or length of stay, supporting the claim.Evidence: Safety outcomes: ICU transfer HR 1.04 (95% CI 0.89-1.21, P=0.62); length of stay HR 0.96 (95% CI 0.91-1.01, P=0.13).
“Safety outcomes of mean days to ICU transfer (6.5 vs 7.1 days) and hospital length of stay (6.8 vs 7.1 days) did not differ significantly between the routine and CPOE intervention groups.”
Abstract - supportedReviewer 2The intervention is safe and effective for non-critically ill adults with pneumonia.The trial's design and results support this claim for the studied population.Evidence: Primary and safety outcomes as reported.
“Empiric extended-spectrum antibiotic use was significantly lower among adults admitted with pneumonia to non-ICU settings in hospitals using education, feedback, and real-time CPOE prompts recommending standard-spectrum antibiotics for patients at low risk of MDRO infection, compared with routine stewardship practices.”
Conclusion
Premise concern: surrogate not validated for clinical benefit; effect size not shown to be clinically meaningful.
- INADEQUATESurrogate endpointThe primary outcome is extended-spectrum antibiotic days of therapy, a process measure (antibiotic use) rather than a hard clinical outcome. The paper does not provide evidence linking reduced antibiotic use to improved clinical outcomes; safety outcomes (ICU transfer, length of stay) were unchanged, and no validated surrogate-to-clinical outcome link is established.
“The primary outcome was empiric (first 3 days of hospitalization) extended-spectrum antibiotic days of therapy.”
- INADEQUATEEffect sizeThe reported 28.4% reduction in extended-spectrum antibiotic days of therapy is a relative reduction in a process measure, not anchored to a clinically meaningful threshold or patient benefit. The absolute reduction from 613.9 to 428.5 days per 1000 empiric days is not framed against a minimal clinically important difference, and safety outcomes were unchanged, so the clinical meaningfulness is unclear.
“a 28.4% (95% CI, 22.2%-34.1%; P <.001) significantly lower rate of empiric extended-spectrum days of therapy in the CPOE bundle group compared with routine stewardship.”
Data authenticity concerns
1 finding · worst mediumAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
- Other integrity concernAssessed
3 integrity concerns flagged (0 high).
- mediumotherTrial NCT03697070 was first submitted to ClinicalTrials.gov on 2018-10-03, after the registered study start date of 2018-10-01. Retrospective registration means the protocol and outcomes were not on the public record before the study ran, which is what prospective registration exists to establish.
NCT03697070
reviewer’s wording
Reporting gaps
1 finding · worst mediumRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
- Data/code availability incompleteAssessed
The introduction cites multiple prior studies on antibiotic overuse and stewardship strategies, acknowledges limitations of prior approaches (e.g., focusing on duration rather than initial selection), and provides a logical rationale linking the premise to the objective of reducing empiric extended-spectrum antibiotic use. Limitations of prior research are addressed by the novel real-time risk-based prompt approach.
“Real-time identification of patients with low risk for MDRO pneumonia might therefore reduce empiric extended-spectrum antibiotic exposure.”
“Real-time identification of patients with low risk for MDRO pneumonia might therefore reduce empiric extended-spectrum antibiotic exposure.”
Randomization was performed within matched pairs using Mahalanobis distance, with a 1:1 ratio. The unit of randomization was the hospital (cluster). Blinding is not explicitly described, but the nature of the intervention (CPOE prompts) makes blinding of clinicians infeasible; the paper does not state this rationale explicitly. Power analysis is reported (97% power to detect 12.5% difference). Inclusion/exclusion criteria are clearly defined. Outlier handling is addressed through sensitivity analyses and pre-specified analysis populations. Controls are inherent in the routine stewardship group. Independent replication is not applicable for a single pivotal trial.
Sex, age, race/ethnicity, comorbidities, and health status are reported in Table 1. Since both sexes are enrolled, sex justification is not applicable. Species/strain and housing conditions are not applicable for a human trial.
“Sex a Male 12,852 (48.4) 11,869 (47.6) 10,966 (48.2) 10,545 (48.1) Female 13,710 (51.5) 13,085 (52.3) 11,768 (51.7) 11,400 (51.8)”
“the mean (SD) age of patients was 68.1 (17.0) years, 48.1% were men”
The paper states that the Harvard Pilgrim Health Care Institute IRB provided centralized oversight, and individual informed consent was waived. This satisfies both IRB approval and consent handling. Regulatory compliance is implied through IRB oversight and CONSORT reporting.
“The Harvard Pilgrim Health Care Institute Institutional Review Board provided centralized oversight, with reliance agreements and operational committee approvals from participating hospitals. Individual informed consent was waived.”
“The Harvard Pilgrim Health Care Institute Institutional Review Board provided centralized oversight, with reliance agreements and operational committee approvals from participating hospitals.”
“Individual informed consent was waived.”
The intervention is a CPOE prompt system, not a drug or device. The paper describes the algorithm and the EHR system (MEDITECH). Statistical software (SAS 9.4, R 4.0.0) is identified. No antibodies, cell lines, or organisms are used, so those criteria are not applicable.
“All analyses were performed using SAS version 9.4 (SAS Institute) or R version 4.0.0 (R Foundation).”
“Hospitals were eligible to participate if they used the MEDITECH electronic health record (EHR) system”
“All analyses were performed using SAS version 9.4 (SAS Institute) or R version 4.0.0 (R Foundation).”
Tests are named (generalized linear mixed-effects models, proportional hazards models). Assumptions are handled by design (clustering accounted for). Exact p-values are reported (e.g., P < .001). Effect sizes with 95% CIs are provided. Software is identified. Data presentation includes per-group n and rates. Mathematical plausibility checks were not possible for all numbers due to complex models, but no obvious errors were found.
“The overall rate ratio (RR) when clustering by hospital and period was 0.72 (95% CI, 0.66- 0.78; P < .001) for the primary outcome”
The paper states 'Data Sharing Statement: See Supplement 3.' This is vague and does not specify a concrete access route. No repository deposit or accession numbers are provided. No custom code is shared.
“Data Sharing Statement: See Supplement 3.”
“Data Sharing Statement: See Supplement 3.”
Trial registration number is provided (NCT03697070). CONSORT guidelines are mentioned. All outcomes are reported, including safety outcomes. Limitations are explicitly discussed. Conclusions are proportional to the evidence. Funding and COI disclosures are provided.
“There are several limitations in this study. First, the trial was performed in private community hospitals and applicability to other settings is unknown.”
“Trial Registration ClinicalTrials.gov Identifier: NCT03697070”
“There are several limitations in this study. First, the trial was performed in private community hospitals and applicability to other settings is unknown.”
Registered (1 ID: ClinicalTrials.gov). Reporting guideline cited: CONSORT.
Broken references and links
None found · partly checkedReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
Nothing surfaced — but not everything feeding this category ran (missing: data/code link verification), so read this as a partial clean bill.
Checked 31 references by DOI: 28 verified — 3 no DOI (shown, not verified).
- NO DOIAntibiotic Resistance Threats in the United States, 2019No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOICore Elements of Hospital Antibiotic Stewardship ProgramsNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIOn the generalised distance in statisticsNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
Copyediting
5 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 5 minor suggestions below.
5 copyedit issues flagged: mostly consistency, typo.
- MINORconsistencyAbstract, Results“48.1% were men”→ Ensure consistency with Table 1 sex percentages (e.g., 48.4% male in baseline).Abstract reports 48.1% men, while Table 1 shows 48.4% for baseline routine group; likely due to rounding or different denominator.
- MINORtypoResults, Patient Characteristics“23 078/26 ,640”→ Remove spaces: '23,078/26,640'.Spacing inconsistency in numbers.
- MINORconsistencyResults, Antibiotic Prescribing“51.5% (12,901/25,030)”→ Check denominator: baseline routine group N is 25,031 elsewhere; ensure consistency.Denominator 25,030 vs 25,031 in Table 1; likely a typo.
- MINORtypoTable 1, Sex row“Female 13,710 (51.5) 13,085 (52.3) 11,768 (51.7) 11,400 (51.8)”→ Check that percentages sum to 100 with male percentages; likely rounding.Percentages for male and female may not sum exactly to 100 due to rounding.
- MINORconsistencyResults, Antibiotic Prescribing and MDRO Risk Estimation“Receipt of any empiric extended-spectrum antibiotic was 51.5% (12,901/25,030) during the baseline period”→ Verify denominator: 25,030 vs 25,031 reported elsewhere for routine baseline.Minor discrepancy in denominator (25,030 vs 25,031) may be a typo.
The published work is robust and trustworthy, with no major rigor gaps. An informed reader should weigh the vague data availability statement and the retrospective registration as minor transparency limitations. The copyedit inconsistencies in reported numbers are minor and do not affect conclusions, but a correction or erratum could clarify them.
- 1.HIGHdata codeReplace the vague 'Data Sharing Statement: See Supplement 3.' with a concrete description of data access conditions, including the managed-access platform or data-access committee, contact, and timeframe.The current statement is inadequate for a data-driven clinical trial and fails to meet reproducibility expectations.
- 2.HIGHreportingAdd a statement in the Methods (Study Design and Intervention) explaining that blinding of clinicians was not feasible due to the nature of the CPOE intervention, and describe any blinding of outcome assessors or data analysts.The absence of any blinding description is a reporting gap that reviewers flagged.
- 3.MEDIUMdata codeDeposit de-identified aggregate data or a data dictionary in a public repository (e.g., Dryad, Zenodo) with a DOI, if permissible under privacy constraints.A concrete repository deposit would strengthen reproducibility and transparency.
- 4.MEDIUMdata codeShare custom analysis code (e.g., for the MDRO risk models) in a public repository (e.g., GitHub) with a version and DOI.Code sharing is not mentioned and would enhance the reproducibility of the analyses.
- 5.MEDIUMcopyeditReconcile the abstract's '48.1% were men' with Table 1's baseline male percentage (48.4%) and ensure consistent rounding or denominator.The internal contradiction between the abstract and Table 1 could confuse readers.
- 6.MEDIUMcopyeditFix the spacing in '23 078/26 ,640' to '23,078/26,640' in the Results section.Spacing inconsistency in numbers is a typographical error that should be corrected.
- 7.MEDIUMcopyeditVerify the denominator for the routine stewardship baseline group: 25,031 in text vs 25,030 in one percentage calculation, and correct the inconsistency.A denominator discrepancy could be a typo but should be resolved to avoid confusion.
- 8.LOWreportingAdd a note in the Methods about how missing data (e.g., unknown sex) were handled in the analysis.Clarifying missing data handling improves methodological transparency.
- 9.LOWreportingConsider describing the CPOE prompt software version and any custom code used for the algorithm to improve software identification.More specific software identification would strengthen the key_resources reporting.
- 10.LOWreportingAdd a statement on whether any independent replication or validation cohort is planned or available.Addressing replication potential would strengthen the study's generalizability discussion.
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.