# Oral single-ingredient berberine hydrochloride × Reduction of HbA1c in adults with type 2 diabetes Verdict 3067 · Evidence Check Stored-material review date: 2026-10-05 Original evidence review date: 2026-09-08 Review-criteria version: 1.2 Public page/material preparation date: 2026-10-27 Clinical papers were not read directly; existing review summaries were checked. No new clinical-literature search was performed. Unverified does not mean that research is absent or that there is no effect. ## Assessment scope - Population: Adults with type 2 diabetes. Detailed diagnosis, age, baseline HbA1c, eligibility/exclusion and losses during pretreatment remain unverified in the current summary. - Ingredient, formulation and route: Oral industrially synthesized single-ingredient berberine hydrochloride tablets. Botanical extracts, other derivatives and combinations are not interchangeable; actual batch, purity, added ingredients and standardization remain unverified. - Dose: The nominal dose is 600 mg 2 times daily (1.2 g/day). Actual intake, adherence, dose reduction, discontinuation and rescue treatment remain unverified. - Comparator: Under identical gentamicin pretreatment: the 98-person berberine-only active-ingredient arm versus the 103-person berberine/probiotic double-placebo arm. These are allocation counts, not analysis counts. - Measured outcome: The between-arm difference in absolute HbA1c change at week 13. Retained LS mean changes are −0.99 %p versus −0.59 %p, difference −0.40 %p (95% CI −0.67 to −0.13; adjusted P=.0006). Actual model, selected-arm analysis n and adjustment definitions remain unverified. - Duration: All arms received 7 days of gentamicin pretreatment, followed by 12 weeks of trial agent and assessment at week 13. This is not expanded to effects without antibiotics, pretreatment effects or persistence after stopping. - Questions outside this scope: Original C, 54 points, Warning and B/S/R1/I1/E+/B0/CX are retained. Table statuses do not establish a new score, effect or clinical certification.; 409 randomized, 391 with primary data and 18 (4.4%) missing refer to all 4 arms. They are not transferred to actual analysis, completion or missing-data denominators for selected-arm allocation of 98/103 or newly added together.; The 2008 pilot’s 36 randomized, 15/16 analyzed and 9.5±0.5%→7.5±0.4%/within-arm P<.01 are active-comparator context. Exact between-arm differences, CI and type of ± remain unverified; this is not independent placebo-comparison replication or metformin equivalence.; No new claim is made of diabetes cure, complication prevention, metformin replacement, identical effects across all berberine products, optimal dose, long-term persistence or subgroup confirmation.; Original warnings concerning gastrointestinal adverse effects, hypoglycemia with glucose-lowering drugs, checking CYP2D6, 2C9 and 3A4 substrate medicines, and avoiding use in pregnancy/lactation/infants are retained. Actual event/person counts, harm denominators, follow-up and interaction magnitudes remain unverified; personal safety is not certified. ## Checklist items ### N01 · Is the week 13 HbA1c question fixed for the single-active-ingredient and double-placebo arms? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: This concerns the between-arm difference in absolute HbA1c change at week 13 with industrially synthesized single-ingredient berberine hydrochloride. Combination arms, metformin comparisons and effects without pretreatment are excluded from this question. - Why unverified: The current claim conditions and text specify the selected comparison, fixing this limited scope. - Reason for missing information: The question's boundary is verified. This is neither new verification of the full original paper nor confirmation of identical effects for other berberine products. - Required materials and conditions: The formulation, population, timing, endpoint and comparison conditions, and their change history. - Materials to obtain first: The formulation, population, timing, endpoint and comparison conditions, and their change history. - Further research: If the boundary changes, retain the existing conditions and separately assess whether the new comparison asks the same question. - Conditions for verification: Link common pretreatment, the single active ingredient, double placebo and the between-arm difference in absolute HbA1c change at week 13; exclude other conditions. - Why this condition is needed: Prevent mixing results with different interventions, comparators or times. - Application exceptions: The original subclaim list is empty; English wording is not counted as another independent claim. - Remaining uncertainty: The question's boundary is verified. This is neither new verification of the full original paper nor confirmation of identical effects for other berberine products. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01, S02 ### N02 · Are the stated ingredient, source and formulation and actual product specifications verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The claim conditions specify industrially synthesized berberine hydrochloride as a single-ingredient tablet, not a botanical extract. Actual batch, purity, added ingredients and standardization remain unverified in the retained summary. - Why unverified: The stated molecule and formulation are known, but product specifications for identity matching remain unresolved. - Reason for missing information: Manufacturing, batch, purity, added-ingredient and supplied-product comparison information for the tablet is unavailable. - Required materials and conditions: Trial-product descriptions, ingredient lists, batch and purity/added-ingredient specifications. - Materials to obtain first: Trial-product descriptions, ingredient lists, batch and purity/added-ingredient specifications. - Further research: Compare product-analysis and manufacturing records to verify agreement between stated amounts and the actual molecule/formulation. - Conditions for verification: Retain the stated identity as synthetic single-ingredient berberine hydrochloride tablets; verify actual trial-product manufacturing, batch, purity, added ingredients and standardization against that identity. Recording the distinction between stated identity and unverified specifications does not replace actual product-identity verification. - Why this condition is needed: Actual product identity is needed to avoid transferring results to other extracts, derivatives or combinations. - Application exceptions: Dihydroberberine, botanical extracts and topical products are not identical to the single-ingredient tablet. - Remaining uncertainty: Manufacturing, batch, purity, added-ingredient and supplied-product comparison information for the tablet is unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N03 · Are the nominal dose and actual intake/adherence verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The nominal dose is 600 mg 2 times daily (1.2 g/day). Arm-level intake verification, adherence rates, dose reductions, discontinuation and rescue treatment are absent from the current summary. - Why unverified: The nominal intake is stated, but actual exposure and adherence effects remain unverified. - Reason for missing information: Adherence and medication-change information needed to establish the stated dose as actual intake is unavailable. - Required materials and conditions: Arm-level intake records, adherence criteria and dose-reduction/discontinuation/rescue-treatment information. - Materials to obtain first: Arm-level intake records, adherence criteria and dose-reduction/discontinuation/rescue-treatment information. - Further research: Distinguish assignment effects from actual adherence effects when comparing intake records and planned analyses. - Conditions for verification: Retain the stated 600 mg 2 times daily (1.2 g/day) while verifying actual intake, adherence and changes. - Why this condition is needed: A nominal dose does not automatically establish actual exposure. - Application exceptions: Unverified adherence does not mean that every participant took none of the treatment. - Remaining uncertainty: Adherence and medication-change information needed to establish the stated dose as actual intake is unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N04 · Are common gentamicin pretreatment and the selected intervention's contribution distinguished? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: All arms first received gentamicin for 7 days, including the berberine-only active-ingredient and double-placebo arms. This comparison does not estimate effects without antibiotics or the pretreatment's own effect. - Why unverified: The text and claim conditions specify common pretreatment, fixing it as a condition of the current contrast. - Reason for missing information: The common pretreatment's identity and duration are verified. Effects without pretreatment, pretreatment adherence and selection effects remain separately unverified. - Required materials and conditions: Pretreatment plans, actual intake and separate material comparing conditions without pretreatment. - Materials to obtain first: Pretreatment plans, actual intake and separate material comparing conditions without pretreatment. - Further research: If a question without pretreatment is needed, examine a separate comparison with pretreatment presence/absence designed in advance. - Conditions for verification: Retain 7 days of gentamicin pretreatment in every arm; do not transfer results to berberine effects without that condition. - Why this condition is needed: Distinguish the common antibiotic condition from applicability to ordinary supplements. - Application exceptions: Do not interpret the 7-day pretreatment as tolerability run-in selection, washout or an effect without pretreatment. - Remaining uncertainty: The common pretreatment's identity and duration are verified. Effects without pretreatment, pretreatment adherence and selection effects remain separately unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N05 · Is the sequence of pretreatment, trial-agent duration and week 13 assessment verified? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The retained conditions describe 7 days of gentamicin pretreatment followed by 12 weeks of trial agent and HbA1c assessment at week 13. - Why unverified: The timing conditions and the week 13 endpoint in the evidence agree. - Reason for missing information: Nominal duration and assessment time are verified. Visit windows, actual arm-level exposure durations and sustained long-term effects were not verified. - Required materials and conditions: Visit schedules, assessment windows, administration/discontinuation dates and actual arm-level exposure durations. - Materials to obtain first: Visit schedules, assessment windows, administration/discontinuation dates and actual arm-level exposure durations. - Further research: Compare planned visits with actual measurement dates to examine timing deviations and handling of missing data. - Conditions for verification: Separate 7 days of pretreatment, 12 weeks of trial agent and assessment at week 13, linking them as the conditions of a single assessment. - Why this condition is needed: Exposure duration and outcome-assessment time are separate information. - Application exceptions: Do not change 12 weeks of trial agent into assessment at week 12 or calculate a new calendar duration. - Remaining uncertainty: Nominal duration and assessment time are verified. Visit windows, actual arm-level exposure durations and sustained long-term effects were not verified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N06 · Are the type 2 diabetes population, baseline status and selection process verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The population is adults with type 2 diabetes. The current summary lacks detailed diagnostic and age information, baseline HbA1c distributions, eligibility/exclusion and arm-level losses during pretreatment. - Why unverified: Disease and adulthood are known, but information needed to assess detailed population matching and representativeness remains unresolved. - Reason for missing information: Detailed selection, baseline characteristics and losses during pretreatment remain unverified. - Required materials and conditions: Protocol eligibility/exclusion criteria, baseline tables and pretreatment/allocation flow diagrams. - Materials to obtain first: Protocol eligibility/exclusion criteria, baseline tables and pretreatment/allocation flow diagrams. - Further research: Compare baseline tables and flow diagrams to assess whether the actual analysis population matches the selected population. - Conditions for verification: Verify the adult type 2 diabetes population, diagnosis, age, baseline HbA1c, co-medication, eligibility/exclusion and allocation after pretreatment. - Why this condition is needed: A population label alone does not establish applicability to every adult with diabetes. - Application exceptions: Pretreatment's presence does not establish that tolerability selection occurred. - Remaining uncertainty: Detailed selection, baseline characteristics and losses during pretreatment remain unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N07 · Are the berberine-only active ingredient and double-placebo compositions verified? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The selected contrast is the 98-person berberine-only active-ingredient arm versus the 103-person double-placebo arm. The probiotic-combination arm is excluded; the comparator received the same gentamicin pretreatment. - Why unverified: The retained arm identities and double-placebo composition make the selected contrast clear. - Reason for missing information: The comparison composition is verified within the retained summary. Placebo appearance/ingredients and actual arm-level co-use/adherence require separate verification. - Required materials and conditions: The 4-arm administration table, placebo ingredients/appearance and arm-level co-use/adherence information. - Materials to obtain first: The 4-arm administration table, placebo ingredients/appearance and arm-level co-use/adherence information. - Further research: Link the administration table to analysis contrasts to check for mixing results from other active arms. - Conditions for verification: Link the berberine-only active-ingredient arm and berberine/probiotic double-placebo arm as a contrast under identical pretreatment. - Why this condition is needed: Do not transfer contributions from combination arms to a single active ingredient's effect. - Application exceptions: The single-active-ingredient arm does not mean exposure without common gentamicin. - Remaining uncertainty: The comparison composition is verified within the retained summary. Placebo appearance/ingredients and actual arm-level co-use/adherence require separate verification. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N08 · Are the 4-arm randomization/blinding, concealment and baseline balance verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: PREMOTE is retained as a prospectively registered 4-arm randomized double-blind placebo trial. Actual sequence generation, concealment, maintenance of assessment blinding and baseline balance remain unverified in the current summary. - Why unverified: The design name is known, but procedural information for assessing bias remains unresolved. - Reason for missing information: RCT and double-blind labels alone cannot verify actual allocation and measurement procedures. - Required materials and conditions: Randomization-generation/concealment procedures, baseline tables and placebo/assessor-blinding procedures. - Materials to obtain first: Randomization-generation/concealment procedures, baseline tables and placebo/assessor-blinding procedures. - Further research: Compare methods, protocols and baseline tables by outcome to examine possible allocation and measurement differences. - Conditions for verification: Retain the prospectively registered 4-arm randomized double-blind design label; verify actual allocation generation/concealment, maintained assessment blinding and baseline balance from methods and baseline data. Recording the design label separately from unverified procedures does not replace procedure verification. - Why this condition is needed: Distinguish a design label from actual conduct. - Application exceptions: A blinding label certifies neither absence of bias nor successful blinding of every assessment. - Remaining uncertainty: RCT and double-blind labels alone cannot verify actual allocation and measurement procedures. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N09 · Are HbA1c change units and actual testing methods verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The selected endpoint is the between-arm difference in absolute HbA1c change at week 13. A decrease in %p is neither a relative % decrease in HbA1c nor final HbA1c itself. Testing methods and laboratory standardization remain unverified. - Why unverified: The absolute-change unit and direction are stated, but actual measurement conditions remain unresolved. - Reason for missing information: Testing methods, laboratories and consistency between baseline/final measurements are unavailable. - Required materials and conditions: HbA1c measurement methods, standardization and baseline/week 13 measurement timing and repetition rules. - Materials to obtain first: HbA1c measurement methods, standardization and baseline/week 13 measurement timing and repetition rules. - Further research: Compare measurement plans and results for use of identical methods and handling of missing measurements. - Conditions for verification: Verify the %p unit and direction of absolute HbA1c change at week 13, methods, laboratories and time-specific standards. - Why this condition is needed: Distinguish absolute change from relative change and final values. - Application exceptions: Do not substitute clinical complications, cure or mortality outcomes for HbA1c. - Remaining uncertainty: Testing methods, laboratories and consistency between baseline/final measurements are unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N10 · Do retained LS mean changes, the between-arm difference, CI and P describe the same contrast? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: PREMOTE's LS mean changes are −0.99 %p versus −0.59 %p, with a between-arm difference of −0.40 %p (95% CI −0.67 to −0.13; adjusted P=.0006). These are retained original reported values; no new difference or interval was calculated. - Why unverified: The evidence results and reader answer state the numbers, direction and interval for the same contrast. - Reason for missing information: Attribution of displayed values to the original text is verified. The original paper's model, analysis n and calculation definitions for intervals/adjustments were not newly verified. - Required materials and conditions: Original result tables, analysis plans and definitions of adjusted P and CI. - Materials to obtain first: Original result tables, analysis plans and definitions of adjusted P and CI. - Further research: Compare the original table's contrast, units and model against planned analyses to verify attribution of displayed values. - Conditions for verification: Link the original −0.99 versus −0.59 %p, difference −0.40 %p, 95% CI −0.67 to −0.13 and adjusted P=.0006 unchanged. - Why this condition is needed: Retain original values and interpretation conditions accurately even for favorable results. - Application exceptions: Do not present the original difference as new arithmetic or a new clinical verdict, or convert it into a raw mean. - Remaining uncertainty: Attribution of displayed values to the original text is verified. The original paper's model, analysis n and calculation definitions for intervals/adjustments were not newly verified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N11 · Are the LS mean label, actual model, analysis population and adjustments verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The LS mean change label is known, but the retained summary does not verify model equations, covariates, ITT/PP, missing-data handling or actual analysis n in the two arms. - Why unverified: The estimate type is known, but the population and adjustments underlying it remain unresolved. - Reason for missing information: The actual LS mean model, analysis population, covariates and selected-arm analysis n remain unverified. - Required materials and conditions: Statistical analysis plans, model equations, analysis-population definitions, arm-level analysis n and missing-data handling. - Materials to obtain first: Statistical analysis plans, model equations, analysis-population definitions, arm-level analysis n and missing-data handling. - Further research: Compare planned models with result-table populations and adjustment variables to distinguish the estimated effects. - Conditions for verification: Verify the least-squares mean label together with the actual model, ITT/PP, missing-data handling, adjustment variables and analysis n for the selected contrast. - Why this condition is needed: Do not interpret an LS mean as a raw mean or an effect in everyone randomized. - Application exceptions: Do not delete original effect values because analysis methods are unavailable or fill the gap with ITT. - Remaining uncertainty: The actual LS mean model, analysis population, covariates and selected-arm analysis n remain unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N12 · Are the adjusted P/95% CI labels and testing/multiplicity definitions verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: P=.0006 is labelled adjusted P, and the original 95% CI is retained. The test/multiplicity adjustment method and whether the CI uses the same adjustment scope remain unverified in the current summary. - Why unverified: P and interval strings are known, but detailed testing and adjustment definitions remain unresolved. - Reason for missing information: Multiplicity methods, test type, significance level and adjustment scopes for CI and P are unavailable. - Required materials and conditions: Original statistical methods, analysis plans and contrast-specific adjustment tables. - Materials to obtain first: Original statistical methods, analysis plans and contrast-specific adjustment tables. - Further research: Compare testing/adjustment definitions with result tables; do not replace original rounded values with new calculations. - Conditions for verification: Verify the testing target, adjustment method, significance level and scope of multiple contrasts in the 4-arm trial for displayed adjusted P=.0006 and 95% CI. - Why this condition is needed: Do not confuse adjusted values in a multi-arm trial with raw P. - Application exceptions: Do not calculate new P/CI or automatically treat printing differences between values as a conflict. - Remaining uncertainty: Multiplicity methods, test type, significance level and adjustment scopes for CI and P are unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N13 · Are the overall 409 participants distinguished from allocation counts for the two selected arms? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: 409 is the overall allocation count across 4 arms; allocation for the selected contrast is 98 versus 103 people. These numbers are not arm-level HbA1c analysis n or completion counts. - Why unverified: The original evidence and crux specify overall and selected-arm allocation roles, allowing their distinction. - Reason for missing information: Allocation-count attribution is verified. Actual analysis, completion and missing-data n for the selected contrast remain unverified. - Required materials and conditions: The 4-arm flow diagram and arm-level allocation/analysis/completion/missing-data tables. - Materials to obtain first: The 4-arm flow diagram and arm-level allocation/analysis/completion/missing-data tables. - Further research: Link identical arm identities across allocation, completion, measurement and analysis stages to check for cross-use of denominators. - Conditions for verification: Record 409 people across all 4 arms and selected-arm allocation of 98/103 unchanged; do not repurpose them as analysis/completion n. - Why this condition is needed: Do not mix overall trial denominators with those of the selected contrast. - Application exceptions: Do not calculate 98+103 as a new sample size or use all 409 as analysis n for the selected contrast. - Remaining uncertainty: Allocation-count attribution is verified. Actual analysis, completion and missing-data n for the selected contrast remain unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N14 · Are the overall primary-data count and selected-arm analysis n/missing-data handling verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The retained overall primary-data count is 391/409, with 18 people (4.4%) missing. Analysis, completion and missing-data denominators and reasons/handling for the selected 98/103 arms are not separately verified. - Why unverified: Overall loss is known, but material verifying actual selected-contrast denominators and missing-data effects remains unresolved. - Reason for missing information: Overall missingness alone cannot establish the selected arms' analysis n, attrition distribution or handling. - Required materials and conditions: HbA1c analysis n by the 4 arms, attrition reasons, imputation/model handling and sensitivity analyses. - Materials to obtain first: HbA1c analysis n by the 4 arms, attrition reasons, imputation/model handling and sensitivity analyses. - Further research: Link arm-level flow to model handling of missing data to distinguish complete-case and assignment effects. - Conditions for verification: Retain 391/409 and 18 people (4.4%) as information for all 4 arms while verifying actual analysis n, attrition reasons and handling by selected arm. - Why this condition is needed: Do not transfer an assessment of low overall missingness to actual selected-arm denominators. - Application exceptions: Do not recalculate 391/409 or distribute the same proportion to selected arms. - Remaining uncertainty: Overall missingness alone cannot establish the selected arms' analysis n, attrition distribution or handling. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N15 · Are arm-level variance and important-difference criteria verified separately from the original CI? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The between-arm 95% CI −0.67 to −0.13 %p is retained. Arm-level SD/SE, covariance and minimum-important-difference criteria/support remain unverified in the current summary. - Why unverified: The interval exists, but direct material on variance and clinical-importance criteria remains unresolved. - Reason for missing information: Arm-level variance/covariance and justified important-difference criteria for this population/time are unavailable. - Required materials and conditions: Arm-level variance/SE, model covariance and evidence for prespecified important-difference criteria. - Materials to obtain first: Arm-level variance/SE, model covariance and evidence for prespecified important-difference criteria. - Further research: Compare the range allowed by the CI against verified important-difference criteria without creating a new sufficiency score. - Conditions for verification: Retain the between-arm CI while verifying arm-level SD/SE, estimation variance and justified HbA1c important-difference criteria. - Why this condition is needed: Statistical significance and clinical importance are separate conditions. - Application exceptions: Do not inherit historical CX or C54 as a new precision/importance judgment. - Remaining uncertainty: Arm-level variance/covariance and justified important-difference criteria for this population/time are unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N16 · Are the NCT clue and detailed registration comparison for the prespecified primary endpoint verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The text describes HbA1c as a prespecified primary endpoint and registration before recruitment; the citation includes NCT02861261. The original registration field says 'Unconfirmed'. Registration, protocol and plan originals remain unread in this work. - Why unverified: The registration identifier and historical account exist; the unconfirmed structured field and current original-registration comparison gap remain retained. - Reason for missing information: Detailed comparison of registration versions, recruitment start, primary endpoints and analysis plans is currently unavailable. - Required materials and conditions: Initial/amended NCT02861261 registrations, protocols, analysis plans and paper tables. - Materials to obtain first: Initial/amended NCT02861261 registrations, protocols, analysis plans and paper tables. - Further research: Compare initial plans with published endpoints, timing and analyses, recording reasons for changes. - Conditions for verification: Retain the historical prespecified-primary/prior-registration account, citation NCT02861261 and unconfirmed original registration field together; verify the scope of actual registration/plan-version comparisons. - Why this condition is needed: Distinguish identifiers and historical comparisons from current direct reading. - Application exceptions: Do not turn an unconfirmed registration field into nonregistration or a citation identifier into successful current registration verification. - Remaining uncertainty: Detailed comparison of registration versions, recruitment start, primary endpoints and analysis plans is currently unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N17 · Is the 2008 pilot's within-arm reduction kept separate from replication of the placebo comparison? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The 2008 pilot randomized 36 people, with 15 berberine and 16 metformin participants analyzed. The retained account reports berberine 9.5±0.5%→7.5±0.4%, within-arm P<.01, and no reported exact between-arm difference/CI. It is not counted as independent confirmation of PREMOTE's placebo comparison. - Why unverified: Comparator and analysis-direction differences are stated; within-arm reduction is not transferred as replication under identical conditions. - Reason for missing information: The active-comparator/within-arm boundary is verified. The pilot's exact between-arm difference, CI and type of ± remain unverified. - Required materials and conditions: Original pilot tables, metformin administration, each arm's baseline/final values and between-arm analyses. - Materials to obtain first: Original pilot tables, metformin administration, each arm's baseline/final values and between-arm analyses. - Further research: Compare actual active-comparator between-arm results while retaining separate conditions from the placebo question. - Conditions for verification: Separate metformin active-comparator/within-arm results from the current between-arm placebo comparison; retain every original number and unverified item. - Why this condition is needed: Favorable adjacent results do not establish independent confirmation when their analysis direction and comparator differ. - Application exceptions: Do not use within-arm P<.01 as between-arm P or an effect equivalent to metformin. - Remaining uncertainty: The active-comparator/within-arm boundary is verified. The pilot's exact between-arm difference, CI and type of ± remain unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S02 ### N18 · Are the pilot's allocation/analysis counts and type of ± verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The pilot randomized 36 people; 15 berberine and 16 metformin participants were analyzed. The type of ± in 9.5±0.5%→7.5±0.4%, endpoint-specific n, exact between-arm difference and CI remain unverified in the current summary. - Why unverified: Some denominator roles are known, but variance definitions and endpoint-specific analysis/precision remain unresolved. - Reason for missing information: The type of ±, attrition reasons, missing-data handling, endpoint-specific n and between-arm precision information are unavailable. - Required materials and conditions: The 2008 flow diagram, table footnotes, ± definitions, analysis populations and between-arm result tables. - Materials to obtain first: The 2008 flow diagram, table footnotes, ± definitions, analysis populations and between-arm result tables. - Further research: Link allocation and analysis populations and verify table-footnote variance definitions and between-arm contrasts. - Conditions for verification: Distinguish 36 randomized from 15/16 analyzed; verify SD/SE definitions for ±, endpoint-specific analysis n, missing data and between-arm intervals. - Why this condition is needed: Do not repurpose allocation counts from an adjacent study as analysis n for the current trial. - Application exceptions: Do not convert ± into CI or state that all 36 were analyzed. - Remaining uncertainty: The type of ±, attrition reasons, missing-data handling, endpoint-specific n and between-arm precision information are unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S02 ### N19 · Are reports, registration and different arms distinguished from unique trials? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: PREMOTE's citation PMID, PMCID, DOI, NCT and 4 arms are gateways/arms of the same trial report. The 2008 pilot is a different report with an active comparator. No new total direct-replication or unique-participant count is calculated. - Why unverified: Retained identities and comparison differences are linked; multiple gateways/arms of the same trial are not counted as independent replications. - Reason for missing information: The relationship between the two current reports is verified within a limited scope. Unique direct replications and exhaustive overlap checks across the current literature were not newly established. - Required materials and conditions: Trial-specific registration/paper linkage, recruitment/sites/periods and related-report lists. - Materials to obtain first: Trial-specific registration/paper linkage, recruitment/sites/periods and related-report lists. - Further research: Compare identifiers and recruitment conditions to organize report-trial families and exclude duplicate arms/recitations. - Conditions for verification: Link the PREMOTE paper/registration/4 arms and the separate 2008 pilot as trial families, leaving unique direct-replication counts unverified. - Why this condition is needed: Do not automatically add papers, registrations, arms or participants. - Application exceptions: A separate 2008 report does not establish independent replication of this placebo comparison. - Remaining uncertainty: The relationship between the two current reports is verified within a limited scope. Unique direct replications and exhaustive overlap checks across the current literature were not newly established. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01, S02 ### N20 · Are funding, product support and the supplier's actual role verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: PREMOTE reports public funding and BBR/placebo supply by Northeast Pharmaceutical, with no design/data role reported. Specific support agreements, author employment/patents and complete COI remain unverified in the current summary. - Why unverified: Public funding, product provision and declared roles are known; detailed independence information remains unresolved. - Reason for missing information: Details of product-support agreements, author employment/patents, declared COI and institutional roles are unavailable. - Required materials and conditions: Paper funding/COI, author affiliations, product-support agreements and role statements. - Materials to obtain first: Paper funding/COI, author affiliations, product-support agreements and role statements. - Further research: Compare funding, product provision and author identities separately; do not automatically discard results solely because of product support. - Conditions for verification: Retain public funding and Northeast Pharmaceutical BBR/placebo provision/no-role reports; verify authors, institutions, COI and support agreements. - Why this condition is needed: Public funding and product support can coexist. - Application exceptions: A report of no supplier role certifies neither current independent verification nor absence of bias. - Remaining uncertainty: Details of product-support agreements, author employment/patents, declared COI and institutional roles are unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N21 · Are the 2008 pilot's support and product/author interests verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The 2008 pilot is retained as supported by NIH/NCCIH, NIDDK and non-US government research funding. Product provision, author employment, patents and declared COI details remain unverified. - Why unverified: Funding institutions are known, but product/author independence for this adjacent report is unverified. - Reason for missing information: Pilot product provision, author affiliations/patents and declared COI information are unavailable. - Required materials and conditions: Original pilot funding/COI, product-provision and author-affiliation material. - Materials to obtain first: Original pilot funding/COI, product-provision and author-affiliation material. - Further research: Compare funding declarations separately from actual product/author relationships. - Conditions for verification: Retain the original NIH/NCCIH, NIDDK and non-US government support account; verify actual pilot product provision, author employment, patents and declared COI. Distinguishing government support from unverified details does not replace verification of product/author independence. - Why this condition is needed: Do not inherit adjacent material's funding as main-trial independence. - Application exceptions: Government support does not establish absence of interests. - Remaining uncertainty: Pilot product provision, author affiliations/patents and declared COI information are unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S02 ### N22 · Is independent replication with matching pretreatment, product, placebo and timing verified? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: Current evidence comprises direct PREMOTE and the adjacent metformin pilot. Other independent direct results for identical gentamicin pretreatment, single-ingredient tablets, double placebo and week 13 HbA1c are unverified in retained material. - Why unverified: The pilot uses different conditions; material verifying direct independent replication within the current scope is unavailable. - Reason for missing information: Other unique trials and their result distribution under identical conditions remain unverified. This does not establish absence of all research. - Required materials and conditions: Trial identities, registration, products, endpoint/analysis results and overlap tables under identical conditions. - Materials to obtain first: Trial identities, registration, products, endpoint/analysis results and overlap tables under identical conditions. - Further research: Predefine unique trials under identical conditions and compare replication results and condition differences individually. - Conditions for verification: Verify results and differences among unique independent trials matching this comparison; do not count a different active-comparator trial as replication. - Why this condition is needed: Do not expand a single direct trial's results into broad replication. - Application exceptions: With 0 new searches, do not establish the worldwide direct-study count as 1 or 0. - Remaining uncertainty: Other unique trials and their result distribution under identical conditions remain unverified. This does not establish absence of all research. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01, S02 ### N23 · Are optimal dose, persistence after stopping and subgroup confirmation required for this question? Verification status: Not applicable Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: This question concerns 1.2 g/day, common pretreatment and week 13 HbA1c. It makes no new claim of optimal dose, persistence after stopping or genotype/special-subgroup effects. - Why unverified: These extensions are outside the selected question; extension confirmation is not imposed as mandatory. - Reason for missing information: Limits of the narrow comparison remain; separate optimal-dose, long-term persistence and subgroup results are not added as new mandatory gaps. - Required materials and conditions: Direct dose comparisons, discontinuation/long-term follow-up and prespecified interaction data if a separate extension is asked. - Materials to obtain first: Direct dose comparisons, discontinuation/long-term follow-up and prespecified interaction data if a separate extension is asked. - Further research: When a separate extension question arises, review prespecified dose/follow-up/interaction comparisons under independent conditions. - Conditions for verification: Distinguish the selected 1.2 g/day/week 13 average HbA1c comparison from separate dose, persistence and subgroup claims. - Why this condition is needed: Existence of a narrow result differs from broad extensions. - Application exceptions: Not applicable means neither current unread original papers nor an extension effect of 0. - Remaining uncertainty: Limits of the narrow comparison remain; separate optimal-dose, long-term persistence and subgroup results are not added as new mandatory gaps. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N24 · Does an HbA1c difference also answer complication-prevention and cure questions? Verification status: Not applicable Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: This concerns week 13 HbA1c only; it is not converted into evidence supporting diabetes cure, complication prevention or metformin replacement. - Why unverified: Clinical events and cure are not outcomes of the selected numerical-change question; event trials are not imposed as mandatory for this question. - Reason for missing information: Only the HbA1c interpretation boundary is verified. Direct event-prevention/cure information was not verified by this table. - Required materials and conditions: Clinical-event definitions, follow-up, comparators, denominators and effect material for a separate event question. - Materials to obtain first: Clinical-event definitions, follow-up, comparators, denominators and effect material for a separate event question. - Further research: If an event question arises, review studies with adequate follow-up and event analyses separately from HbA1c trials. - Conditions for verification: Distinguish the HbA1c surrogate endpoint from separate complication, mortality and cure outcomes. - Why this condition is needed: Do not promote a favorable surrogate result into clinical-event prevention. - Application exceptions: Do not delete retained HbA1c results because event results are unavailable. - Remaining uncertainty: Only the HbA1c interpretation boundary is verified. Direct event-prevention/cure information was not verified by this table. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N25 · Are gastrointestinal-harm collection, event/person counts, denominators and discontinuation verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The text describes gastrointestinal adverse effects as common. Selected-arm event/person counts, severity, discontinuation, collection, harm denominators and temporal ordering remain unverified in the current summary. - Why unverified: The observed harm category and warning are known, but tallies needed to assess frequency, severity and causality remain unresolved. - Reason for missing information: Arm-level harm collection, tallies, discontinuation and actual exposure denominators are unavailable; risk values are not calculated. - Required materials and conditions: Arm-level AE definitions, person/event counts, severe events/discontinuation, collection/follow-up and exposure denominators. - Materials to obtain first: Arm-level AE definitions, person/event counts, severe events/discontinuation, collection/follow-up and exposure denominators. - Further research: Distinguish efficacy and safety populations when comparing event collection, timing and discontinuation. - Conditions for verification: Retain the original gastrointestinal warning; verify actual event definitions/collection, arm-level person/event counts, severity, discontinuation, exposure duration and harm denominators. Recording the distinction between warnings and tallies does not replace verification; efficacy denominators are not inherited as harm denominators. - Why this condition is needed: Efficacy significance establishes neither guaranteed safety nor harm causality. - Application exceptions: Do not turn the original warning into 0 events or use overall 409/391 as harm-analysis denominators. - Remaining uncertainty: Arm-level harm collection, tallies, discontinuation and actual exposure denominators are unavailable; risk values are not calculated. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N26 · Are original co-medication/vulnerable-group warnings and actual safety-information scope verified? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: Warnings concern hypoglycemia risk with glucose-lowering drugs, checking CYP2D6, 2C9 and 3A4 substrate medicines, and avoiding use in pregnancy, lactation and infants. Direct denominators, follow-up and interaction magnitudes for these populations remain unverified. - Why unverified: The original warning's populations and co-use scope are known, but direct safety-evidence details remain unresolved. - Reason for missing information: Exposure, events, follow-up and alternative-cause/adjustment information by vulnerable group/co-use are unavailable. - Required materials and conditions: Exposure, events, follow-up, interactions and original warning sources by co-use/vulnerable group. - Materials to obtain first: Exposure, events, follow-up, interactions and original warning sources by co-use/vulnerable group. - Further research: Compare population/drug/exposure-specific safety data separately against support for original warnings. - Conditions for verification: Retain original hypoglycemia co-use, CYP-drug and pregnancy/lactation/infant warnings; verify actual group-specific exposure, events, follow-up, interactions, alternative causes/adjustment and warning sources needed to assess the safety claim. Distinguishing original warnings from unverified direct data verifies neither safety, causality nor personal safety. - Why this condition is needed: Do not automatically transfer general efficacy-population results to vulnerable-group safety. - Application exceptions: No clinical dosing instructions, new risk values, causality-tool execution or personal-safety certification are provided. - Remaining uncertainty: Exposure, events, follow-up and alternative-cause/adjustment information by vulnerable group/co-use are unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01, S02 ### N27 · Are the retained summary read in this work and unread external materials distinguished? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The current input's Korean/English text, 2 evidence entries, 2 citations, structure and history were read. External PubMed, DOI, original full papers and NCT registration were neither accessed nor read in this work. - Why unverified: Actual reading and unattempted external access are distinguished, and summary-reading links point to exact local locations. - Reason for missing information: The current summary-reading scope is verified. External abstracts, full texts, registration and URL responses/document types remain unverified. - Required materials and conditions: Actual responses, versions, access/reading records and original locations when external material is reviewed. - Materials to obtain first: Actual responses, versions, access/reading records and original locations when external material is reviewed. - Further research: In authorized later review, record actual reading level and location by material and compare against the current summary. - Conditions for verification: Display actual reading of the retained local summary separately from current unattempted, unacquired and unread PubMed/DOI/paper/registration material. - Why this condition is needed: A link or historical account does not establish current full-text reading. - Application exceptions: There were 0 external requests in this work; unattempted access is not labelled failed access. - Remaining uncertainty: The current summary-reading scope is verified. External abstracts, full texts, registration and URL responses/document types remain unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01, S02 ### N28 · Are retained reference, correction and access dates distinguished? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The retained reference date is 2026-09-08; the generic English-wording correction is 2026-09-14. Actual material-specific access dates/levels are absent and remain null; current review time is recorded separately. - Why unverified: Event attribution and absent date fields are retained without inferring report-specific access dates from a global date. - Reason for missing information: Historical access dates/levels by material remain unverified. Only the distinction between current local-reading and original reference dates is verified. - Required materials and conditions: Original material-specific access/version records and complete preserved correction events. - Materials to obtain first: Original material-specific access/version records and complete preserved correction events. - Further research: Separate historical access and current-reading events when comparing dates, versions and material identities. - Conditions for verification: Retain original verdict-review date 2026-09-08 and correction 2026-09-14 historically; distinguish current actual review from null individual access dates. - Why this condition is needed: Review dates and external access dates are different facts. - Application exceptions: Do not inherit a report-level paper/registration comparison account as historical full-text access through every DOI gateway. - Remaining uncertainty: Historical access dates/levels by material remain unverified. Only the distinction between current local-reading and original reference dates is verified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01, S02 ### N29 · Is the scope of checking current literature, corrections and unpublished results available? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The retained research account describes comparison of PREMOTE, NCT02861261 and the 2008 pilot, with stated exclusions. No current complete-literature, unpublished/correction/retraction check or new clinical search was performed. - Why unverified: Converting the current summary alone cannot establish completeness of current evidence. - Reason for missing information: Search logs and exhaustive unpublished/correction/retraction material beyond the historical comparison scope are unavailable. - Required materials and conditions: Contemporaneous queries, search dates, eligibility/exclusion and unpublished/correction/retraction logs. - Materials to obtain first: Contemporaneous queries, search dates, eligibility/exclusion and unpublished/correction/retraction logs. - Further research: In a separately authorized evidence update, reproduce the original search scope and compare missing/changed material by identity. - Conditions for verification: Verify contemporaneous queries, eligibility/exclusion, unpublished/correction records and current complete-literature coverage for the same comparison from actual material. State the absence of a new search separately. - Why this condition is needed: Reading the complete retained summary differs from searching the complete current literature. - Application exceptions: Do not turn absence of new searching into 0 studies, no retractions or complete literature verification. - Remaining uncertainty: Search logs and exhaustive unpublished/correction/retraction material beyond the historical comparison scope are unavailable. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01, S02 ### N30 · Is the scope of specific English text and generic placeholder wording retained? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The English text specifically describes the original comparison. Generic verification wording remains in evidence sample-size, result and funding fields; it is not used as evidence of additional denominators, CI or verified funding. - Why unverified: The actual retained English text and generic wording remain unchanged, with the generic wording's factual scope limited. - Reason for missing information: Generic wording provides no additional analysis n, variance or COI information. No new correction to the original English text was performed in this work. - Required materials and conditions: Complete before/after versions of the English correction and corresponding actual Korean fields. - Materials to obtain first: Complete before/after versions of the English correction and corresponding actual Korean fields. - Further research: In a separate correction, preserve the original location and reason for changes when comparing the scope of accurate translation. - Conditions for verification: Distinguish actual Korean/English text from generic sample-size/result wording; do not create new numbers or acquisition facts from generic wording. - Why this condition is needed: A formal verification sentence does not establish actual clinical verification. - Application exceptions: Do not promote English placeholders or the original independent-cross-check wording into current full-text certification. - Remaining uncertainty: Generic wording provides no additional analysis n, variance or COI information. No new correction to the original English text was performed in this work. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01, S02 ### N31 · Are original C 54, axes and Warning distinguished from new table statuses? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: Original C, 54 points, B/S/R1/I1/E+/B0/CX, Warning, all original values and the empty subclaim list are retained. No new score, grade, effect, CI or NNT is created. - Why unverified: The complete original summary is retained with its original types; condition statuses are separated from the historical verdict. - Reason for missing information: The original verdict's preservation scope is verified. Table preparation or technical checks are not new clinical certification or a grade update. - Required materials and conditions: The complete original summary, verdict/correction events, input identity and in-house application history. - Materials to obtain first: The complete original summary, verdict/correction events, input identity and in-house application history. - Further research: Compare complete original/case identity and change events, separating content reassessment from technical preservation. - Conditions for verification: Retain original grade/score/axes/safety, null/empty values and numerical values unchanged; distinguish new clinical verdicts from condition statuses. - Why this condition is needed: Table conversion creates neither a new original verdict nor an efficacy grade. - Application exceptions: Do not reuse the original historical Codex independent cross-check as current other-author QA or clinical certification. - Remaining uncertainty: The original verdict's preservation scope is verified. Table preparation or technical checks are not new clinical certification or a grade update. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S01, S02 ## Evidence and sources ### S01 Gut microbiome-related effects of berberine and probiotics on type 2 diabetes (the PREMOTE study) [PubMed](https://pubmed.ncbi.nlm.nih.gov/33024120/) Materials actually read: Retained Korean/English summaries, evidence and citations were read. External PubMed, DOI, full text and registration remain unattempted, unacquired and unread. Findings recorded in the summary: LS mean changes −0.99 %p versus −0.59 %p; difference −0.40 %p, 95% CI −0.67 to −0.13, adjusted P=.0006. Distinguish overall 409 randomized/391 with primary data from selected-arm allocation of 98/103. Selected-arm analysis n, actual model and adjustment definitions remain unverified. Record level: These are retained original reported values. External full text/registration were not reread and no new calculation was performed. Independent replication: PREMOTE’s paper, registration and 4 arms are reports/gateways for the same trial. The 2008 pilot is a different active-comparator report. Unique direct replications under identical pretreatment, single-ingredient, double-placebo and week 13 HbA1c conditions, and total unique participants remain unverified. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | External PubMed bibliographic/abstract gateway | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | This is the retained PubMed link. Its current response, abstract provision/content and public access were not checked. A report linked to retained summaries and citations. External gateways, full texts, registration, protocols and analysis plans remain unread, with access not attempted in this conversion. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | External DOI/publisher gateway | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | This is the retained DOI link. Its current response, publisher provision, original-text version and public access were not checked. A report linked to retained summaries and citations. External gateways, full texts, registration, protocols and analysis plans remain unread, with access not attempted in this conversion. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | Original paper full text; direct location/version unverified | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | The original paper’s direct physical location, version, provision, public availability, acquisition and actual reading remain unverified. No direct URL is invented. A report linked to retained summaries and citations. External gateways, full texts, registration, protocols and analysis plans remain unread, with access not attempted in this conversion. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | External clinical-trial registration NCT02861261 | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | This is the retained citation’s NCT02861261 link. The original evidence-table registration field remains unconfirmed; current responses, registration versions and endpoint/analysis plans were not newly verified. A report linked to retained summaries and citations. External gateways, full texts, registration, protocols and analysis plans remain unread, with access not attempted in this conversion. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | Retained verdict summary | Retained verdict summary read | Retained copy acquired; summary read | Stored existing review summary read | Relevant retained summary, evidence and citation content was read. This is not the original clinical paper’s bytes or full text. A report linked to retained summaries and citations. External gateways, full texts, registration, protocols and analysis plans remain unread, with access not attempted in this conversion. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | ### S02 Efficacy of berberine in patients with type 2 diabetes mellitus [PubMed](https://pubmed.ncbi.nlm.nih.gov/18442638/) Materials actually read: Retained Korean/English summaries, evidence and citations were read. External PubMed, DOI and full text remain unattempted, unacquired and unread. Findings recorded in the summary: 36 randomized; 15 berberine and 16 metformin participants analyzed. Berberine 9.5±0.5%→7.5±0.4%, within-arm P<.01; exact between-arm difference, CI and type of ± remain unverified. Record level: This is active-comparator/within-arm context. It is not converted into a direct result for the current placebo comparison. Independent replication: PREMOTE’s paper, registration and 4 arms are reports/gateways for the same trial. The 2008 pilot is a different active-comparator report. Unique direct replications under identical pretreatment, single-ingredient, double-placebo and week 13 HbA1c conditions, and total unique participants remain unverified. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | External PubMed bibliographic/abstract gateway | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | This is the retained PubMed link. Its current response, abstract provision/content and public access were not checked. A report linked to retained summaries and citations. External gateways, full texts, registration, protocols and analysis plans remain unread, with access not attempted in this conversion. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | External DOI/publisher gateway | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | This is the retained DOI link. Its current response, publisher provision, original-text version and public access were not checked. A report linked to retained summaries and citations. External gateways, full texts, registration, protocols and analysis plans remain unread, with access not attempted in this conversion. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | Original paper full text; direct location/version unverified | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | The original paper’s direct physical location, version, provision, public availability, acquisition and actual reading remain unverified. No direct URL is invented. A report linked to retained summaries and citations. External gateways, full texts, registration, protocols and analysis plans remain unread, with access not attempted in this conversion. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | Retained verdict summary | Retained verdict summary read | Retained copy acquired; summary read | Stored existing review summary read | Relevant retained summary, evidence and citation content was read. This is not the original clinical paper’s bytes or full text. A report linked to retained summaries and citations. External gateways, full texts, registration, protocols and analysis plans remain unread, with access not attempted in this conversion. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | ## Review criteria and methods These are Chamgap's own application conditions, informed by the relevant sections of official methodological guidance. They do not constitute official assessment-tool certification or a new efficacy score. - M03 [Cochrane Handbook 6.5 (2024), chapter updated 2023-08](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-03) · §3.1, §3.2.1–3.2.3, MECIR C5/C7/C13 · Clearly record the question, eligibility conditions and reasons for changes, and distinguish variations in interventions. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. - M04 [Cochrane Handbook 6.5.1 (2025), chapter updated 2025-03](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-04) · §4.5, §4.6.1–4.6.3, MECIR C42 · Distinguish studies from reports and link them; do not omit a study solely because results are unreported. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. - M08 [Cochrane Handbook 6.5 (2024), chapter updated 2019-10](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-08) · §8.1–8.2.3, Table 8.2.a · Define the outcome and effect of interest, then examine randomization, deviations, missing data, measurement and selective reporting. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. - M15 [Cochrane Handbook 6.5 (2024), chapter updated 2023-08](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-15) · §15.2.4, §15.3.1–15.3.3 · Distinguish effect estimates, intervals, importance and applicability. Sample size alone does not determine precision. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. - M19 [Cochrane Handbook 6.5 (2024), chapter updated 2019-10](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-19) · §19.1.2, §19.2.1–19.2.4 · Harm questions may require different designs, populations and follow-up from efficacy questions. Case reports do not establish incidence rates. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. ## What the statuses mean - Met: Limited verification of the retained identity, observation or scope required by this row. This is separate from clinical sufficiency or full-text certification. - Partly verified: Some facts needed by the question are known, while the stated analysis, population, formulation or denominator remains unverified. - Unverified: Direct information for this question is unverified in the retained scope. This does not mean an effect of 0, no studies, no harm or an unreviewed table. - Not applicable: A specific extension obligation lies outside the narrow question. This does not substitute for unread or unverified material. - Not applicable (alias): A grey-display alias of Not applicable. - Not read: The external material was not directly read in the current work. This is separate from reading the retained local summary. - Conflicting data: Actual disagreement between reports under the same conditions. Differences between other questions alone do not establish conflict. ## Change history 2026-10-27 · Public page/materials 1.0. The public view and materials were prepared from the existing 31 items. No new clinical search, verification-state change or scoring was performed.