# Oral berberine alone or combined with letrozole × Cumulative live birth in infertility associated with PCOS Verdict 3070 · Evidence Check Stored-material review date: 2026-10-05 Original evidence review date: 2026-09-08 Review-criteria version: 1.0 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: The trial recruited 644 infertile women meeting Rotterdam PCOS criteria. The 644 describes the overall recruited population, not the analysis denominators of the selected two arms. Detailed diagnosis, age, infertility duration, other infertility causes, baseline prognosis and eligibility remain unverified. - Ingredient, formulation and route: The selected intervention adds oral berberine to common active letrozole. Double-dummy tablets are described, but exact salt form, plant source, product, standardization and batch remain unverified. The berberine monotherapy arm provides context from the same trial and is not inherited as an observation of this add-on comparison. - Dose: The planned added berberine dose was 1.5 g/day. Common letrozole was planned at 2.5 mg on cycle days 3–7, increasing to 5 mg if needed. Actual administration, dose escalation, adherence, discontinuation, exposure after pregnancy and the between-arm balance of actual background medication remain unverified. This is not an instruction for personal use. - Comparator: The comparator is the common active letrozole monotherapy strategy. The overall study is described as a 3-arm double-blind, double-dummy randomized trial. Double-dummy placebos are not reinterpreted as a pure no-treatment control. The two derived comparisons share the letrozole arm. - Measured outcome: The endpoint is cumulative live birth during treatment. Retained reported rates are 36% with letrozole, 34% with the combination and 22% with berberine alone; the selected add-on comparison is combination 34% versus letrozole 36%. The historical statement that adding berberine did not increase the rate is retained without converting it into 0 effect, equivalence or noninferiority. Event definition, arm-specific denominators/events and formal between-arm estimates/P/CI remain unverified. The original D grade, 34 points and Warning are historical records, not a new assessment. - Duration: Up to 6 months of treatment and pregnancy-outcome follow-up are described. Individual actual exposure, cycle counts, discontinuation and the length, end and completeness of follow-up through birth remain unverified. The maximum 6 months is not treated as every woman’s actual treatment duration or the end of birth follow-up. - Questions outside this scope: Ovulation, pregnancy rates, insulin and weight outcomes do not replace a live-birth effect.; Salt forms, plant extracts, combination products, dihydroberberine, other derivatives and topical products are not interchangeable.; Do not generate arm denominators from 644/3 or back-calculate events from percentages. No new differences, RR, OR, NNT, CI or P are calculated.; The same 3-arm trial and shared letrozole arm are not added as independent trials or independent participant populations.; Retain both the historical statement of no added benefit with the combination and the lower reported direction for monotherapy. Do not convert these into 0 effect, equivalence, noninferiority or formal between-arm precision.; Retain the original pregnancy and fetal/neonatal bilirubin warning. Actual harm incidence, causation, post-pregnancy exposure and long-term safety remain unverified; no new personal dosing instruction is created.; Distinguish ChiCTR-TRC-09000376 from the excluded NCT01116167 and retain the unverified entry in the original registration field.; Reading retained summaries is not represented as current reading of external abstracts, full texts, registration, protocols or analysis plans.; Joint synthesis or comparison requires checking shared-control dependence, covariance, joint estimation and multiplicity/analysis plans. A joint synthesis model is not always required to describe this add-on comparison alone. ## Checklist items ### N01 · Is the question bounded to cumulative live birth when berberine is added to the same letrozole treatment? 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; access was not attempted. - Current findings: The original claim concerns cumulative live birth in infertile women with PCOS. This sheet compares letrozole plus berberine with letrozole alone; the original detailed-claim list is empty. - Why unverified: The original claim and derived-comparison boundary match the current input and formal assignment. The derived ID was not turned into an index in the original detailed-claim list. - Reason for missing information: This is a boundary-record condition, separate from establishing a clinical effect. - Required materials and conditions: Linked records of the original claim, intervention, comparator and derived-comparison boundary - Materials to obtain first: Linked records of the original claim, intervention, comparator and derived-comparison boundary - Further research: If the boundary changes, redefine the actual intervention/comparator combination and review it in a separate version. - Conditions for verification: Retain the original claim and identify letrozole plus berberine versus letrozole alone as a derived comparison. - Why this condition is needed: The additional contribution on a common letrozole background must be distinguished from the effect of replacing treatment with monotherapy. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: This is a boundary-record condition, separate from establishing a clinical effect. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N02 · Does the text distinguish the combination arm, the common letrozole control and the roles of double-dummy placebos? 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; access was not attempted. - Current findings: The text describes a double-dummy, 3-arm RCT using berberine or placebo and letrozole or placebo. The selected comparison is the combination versus letrozole alone. - Why unverified: The selected arms’ roles can be distinguished in the current description. Actual success of blinding has not been established. - Reason for missing information: Detailed placebo implementation and successful blinding are separate gaps. - Required materials and conditions: Records of active ingredients, placebo allocation and actual comparator roles in the 3 arms - Materials to obtain first: Records of active ingredients, placebo allocation and actual comparator roles in the 3 arms - Further research: If the full protocol is obtained, compare active/placebo allocation by arm and keep different strategies distinct. - Conditions for verification: Select letrozole plus berberine and letrozole alone; do not interpret the double-dummy design as a pure untreated placebo control. - Why this condition is needed: A common active comparator must be retained to distinguish the effect of letrozole itself from the additional contribution of berberine. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Detailed placebo implementation and successful blinding are separate gaps. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N03 · Are the trial berberine’s exact salt form, source, product and standardization established? 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; access was not attempted. - Current findings: The original claim conditions leave the exact salt form and plant source unverified. The retained account also keeps berberine hydrochloride, plant extracts, combination formulas and other derivatives distinct. - Why unverified: The exact product form and standardization cannot be established from the retained material. - Reason for missing information: Product identifier, manufacturer, salt form and batch are absent. - Required materials and conditions: Berberine product name, manufacturer, salt form, batch and standardization material - Materials to obtain first: Berberine product name, manufacturer, salt form, batch and standardization material - Further research: If original product material is obtained, compare composition and standardization and review unmatched products as separate questions. - Conditions for verification: Establish the exact chemical form, product identifier, source, content standardization and application to the selected arm. - Why this condition is needed: Product matching prevents the result of adding berberine to letrozole from being transferred to the add-on effect of arbitrary salts or botanical combinations. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Product identifier, manufacturer, salt form and batch are absent. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N04 · Are the planned oral dose, actual intake and adherence for the added berberine established? 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; access was not attempted. - Current findings: The oral route and planned berberine 1.5 g/day are specified. Actual intake, adherence and discontinuation details are absent from current material. - Why unverified: The planned oral dose is established, while actual individual exposure is not. - Reason for missing information: Actual administration, adherence and the distribution of discontinuations remain unverified. - Required materials and conditions: Arm-specific prescriptions, adherence, actual daily doses and discontinuation records - Materials to obtain first: Arm-specific prescriptions, adherence, actual daily doses and discontinuation records - Further research: Use dosing records to distinguish planned from actual exposure and compare the effects of deviations with the predefined analysis. - Conditions for verification: Distinguish planned oral 1.5 g/day from actual intake, adherence, discontinuation and changes. - Why this condition is needed: Actual exposure to the added ingredient must be compared with the planned dose to interpret its contribution on the common letrozole background. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Actual administration, adherence and the distribution of discontinuations remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N05 · Are the letrozole plan, dose escalation and actual administration established as the same in the combination and control arms? 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; access was not attempted. - Current findings: The plan of 2.5 mg on cycle days 3–7, increasing to 5 mg if needed, is retained. Actual escalation, cycle counts and adherence by arm are unverified. - Why unverified: The planned common background medication is established, but the balance of actual administration between arms remains unverified. - Reason for missing information: Actual arm-specific escalation and cycle counts are absent. - Required materials and conditions: Cycle-specific letrozole prescriptions, escalation, actual intake and between-arm balance in the combination and control arms - Materials to obtain first: Cycle-specific letrozole prescriptions, escalation, actual intake and between-arm balance in the combination and control arms - Further research: Compare letrozole administration, escalation and deviations in both arms to examine whether the common background treatment matched. - Conditions for verification: Check the common letrozole plan of 2.5 mg on cycle days 3–7, increasing to 5 mg if needed, and the between-arm balance of actual escalation and administration. - Why this condition is needed: If actual escalation of the common background medication differed, the result is difficult to attribute solely to the additional contribution of berberine. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Actual arm-specific escalation and cycle counts are absent. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N06 · Are the actual population and Rotterdam PCOS/infertility eligibility criteria sufficiently established for the selected comparison? 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; access was not attempted. - Current findings: The recruited population is described as 644 infertile women meeting Rotterdam PCOS criteria. Detailed diagnosis, age, infertility duration and exclusions are absent from current material. - Why unverified: The principal population category is established, while detailed eligibility and arm-specific characteristics remain unverified. - Reason for missing information: The 644 describes the overall population, not the analysis denominator of either selected arm. - Required materials and conditions: Diagnosis, age, infertility duration, inclusion/exclusion criteria and arm-specific population characteristics - Materials to obtain first: Diagnosis, age, infertility duration, inclusion/exclusion criteria and arm-specific population characteristics - Further research: If eligibility criteria and baseline tables are obtained, compare the target population with the actual trial population. - Conditions for verification: Establish the description of infertile women with Rotterdam PCOS and the detailed diagnostic, age and eligibility criteria. - Why this condition is needed: The population boundary prevents the add-on result from being extended to general PCOS management or other populations. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: The 644 describes the overall population, not the analysis denominator of either selected arm. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N07 · Are baseline infertility causes, prior treatment and arm-specific prognostic characteristics established? 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; access was not attempted. - Current findings: Current material describes infertile women with Rotterdam PCOS but provides no tables of baseline causes, previous treatment or arm-specific characteristics. - Why unverified: Material needed to establish baseline prognostic balance in the selected arms is absent. - Reason for missing information: Exclusion of other infertility causes and treatment histories remain unverified. - Required materials and conditions: Arm-specific baseline tables, investigations of infertility causes, prior treatment and medical histories - Materials to obtain first: Arm-specific baseline tables, investigations of infertility causes, prior treatment and medical histories - Further research: Compare baseline material with planned handling of prognostic variables and actual balance. - Conditions for verification: Establish prognosis in the selected arms, including male/tubal factors, baseline ovulatory/metabolic status and previous treatment. - Why this condition is needed: Baseline prognostic balance in the berberine add-on comparison must be examined using actual data. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Exclusion of other infertility causes and treatment histories remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N08 · Are actual random-sequence generation and implementation of the allocation ratio established in this multicenter trial? 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; access was not attempted. - Current findings: A multicenter double-blind 3-arm RCT with 644 participants and 1:1:1 allocation is described. Sequence generation, stratification and site-specific implementation remain unverified. - Why unverified: The design and allocation ratio are described, but actual randomization procedures are not established. - Reason for missing information: Detailed sequence generation, stratification and allocation by site are absent. - Required materials and conditions: Random-sequence generation, stratification, site-specific allocation procedures and CONSORT flow material - Materials to obtain first: Random-sequence generation, stratification, site-specific allocation procedures and CONSORT flow material - Further research: Compare the protocol and allocation procedures to examine implementation of randomization for the selected arms. - Conditions for verification: Establish not only the multicenter RCT and 1:1:1 description but also sequence generation, stratification and site-specific implementation. - Why this condition is needed: Evaluating selection bias in the two add-on comparison arms requires more than an allocation description. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Detailed sequence generation, stratification and allocation by site are absent. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N09 · Are allocation concealment and recruitment procedures established for the two selected arms? 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; access was not attempted. - Current findings: The current summary describes an RCT but lacks details of allocation concealment, enrollment sequence and recruitment implementation. - Why unverified: The label randomized trial cannot substitute for actual implementation of allocation concealment. - Reason for missing information: Allocation-concealment documents and recruitment records remain unverified. - Required materials and conditions: Allocation concealment, enrollment timing, recruitment sequence and actual implementation material - Materials to obtain first: Allocation concealment, enrollment timing, recruitment sequence and actual implementation material - Further research: Obtain allocation/enrollment documents to examine concealment procedures in relation to recruitment timing. - Conditions for verification: Establish recruitment before allocation disclosure, concealment methods and actual implementation in the selected arms. - Why this condition is needed: Recruitment bias from predictable allocation must also be examined when comparing an addition to the same background medication. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Allocation-concealment documents and recruitment records remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N10 · Are actual double-dummy blinding and the independence of live-birth ascertainment established? 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; access was not attempted. - Current findings: Multicenter double blinding and a double dummy are described. Placebo appearance, actual maintenance of blinding and the persons/procedures confirming births remain unverified. - Why unverified: The planned design is described, but actual blinding and ascertainment implementation are not established. - Reason for missing information: Maintenance of blinding and outcome-verification procedures are absent. - Required materials and conditions: Placebo matching, maintenance/breaking of blinding, outcome assessors and verification procedures - Materials to obtain first: Placebo matching, maintenance/breaking of blinding, outcome assessors and verification procedures - Further research: Compare protocol and ascertainment documents to examine implementation of blinding and independence of outcome assessment. - Conditions for verification: Establish the double-blind/double-dummy description, placebo matching, unblinding or guessability and birth-ascertainment procedures. - Why this condition is needed: This condition examines deviations in dosing, subsequent care and outcome assessment caused by knowledge of the added ingredient. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Maintenance of blinding and outcome-verification procedures are absent. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N11 · Are concomitant treatment, rescue treatment and deviations beyond the common letrozole established in both arms? 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; access was not attempted. - Current findings: The current summary gives only selected-arm names and planned doses; actual concomitant/rescue treatment, crossover and reasons for discontinuation are absent. - Why unverified: Material needed to establish differences in other treatments within the selected comparison is absent. - Reason for missing information: Actual concomitant/rescue treatments and deviations remain unverified. - Required materials and conditions: Arm-specific concomitant/rescue treatment, crossover, other infertility treatment and reasons for discontinuation - Materials to obtain first: Arm-specific concomitant/rescue treatment, crossover, other infertility treatment and reasons for discontinuation - Further research: Reconstruct arm-specific treatment pathways to examine the definition of the effect of interest and appropriate handling of deviations. - Conditions for verification: Check concomitant, rescue and other infertility treatment and differences between the arms apart from the added ingredient. - Why this condition is needed: If concomitant treatment beyond the common letrozole differed, attribution solely to added berberine is difficult. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Actual concomitant/rescue treatments and deviations remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N12 · Are the maximum 6-month plan, actual individual treatment/cycle counts and exposure after pregnancy established? 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; access was not attempted. - Current findings: A maximum 6 months and pregnancy-outcome follow-up are described as a plan. Actual individual administration, cycle counts, discontinuation and exposure after pregnancy remain unverified. - Why unverified: The duration plan is established, but actual exposure length and discontinuation pathways are not. - Reason for missing information: The maximum 6 months is not read as every woman’s actual exposure or the end of pregnancy follow-up. - Required materials and conditions: Arm-specific actual treatment duration, cycle counts, discontinuation timing and dosing after pregnancy - Materials to obtain first: Arm-specific actual treatment duration, cycle counts, discontinuation timing and dosing after pregnancy - Further research: Distinguish the treatment calendar from pregnancy follow-up and examine actual exposure duration and discontinuation pathways. - Conditions for verification: Distinguish the maximum 6-month plan from actual treatment duration, ovulatory cycle counts, discontinuation and exposure after pregnancy confirmation. - Why this condition is needed: This condition prevents actual treatment duration for the added exposure from being substituted for the entire pregnancy follow-up period. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: The maximum 6 months is not read as every woman’s actual exposure or the end of pregnancy follow-up. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N13 · Are the length, end and completeness of pregnancy follow-up for live-birth ascertainment established? 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; access was not attempted. - Current findings: The original claim conditions state a maximum 6 months with pregnancy-outcome follow-up, but length through birth, arm-specific follow-up end and completeness are absent. - Why unverified: The exact observation window used to establish birth outcomes cannot be determined from current material. - Reason for missing information: Pregnancy follow-up length and completeness remain unverified. - Required materials and conditions: Arm-specific pregnancy/birth follow-up schedules, last contact, follow-up end and completeness - Materials to obtain first: Arm-specific pregnancy/birth follow-up schedules, last contact, follow-up end and completeness - Further research: Compare the observation window from pregnancy onset to birth confirmation with incomplete follow-up. - Conditions for verification: Distinguish treatment end from follow-up after pregnancy through birth and establish arm-specific follow-up start, end and completeness. - Why this condition is needed: The observation window needed to obtain birth outcomes in the add-on comparison may differ from the treatment duration. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Pregnancy follow-up length and completeness remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N14 · Are the event definition, cumulative method and measurement unit of cumulative live birth sufficiently established? 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; access was not attempted. - Current findings: Cumulative live birth during treatment is specified as the principal endpoint. Gestational-age criteria, woman/infant units, multiple births and cumulative counting rules remain unverified. - Why unverified: The principal endpoint label is established, but actual event definition and unit are not sufficiently established. - Reason for missing information: Event definition, unit and verification procedures remain unverified. - Required materials and conditions: Endpoint definition, birth-verification procedures, cycle-accumulation rules and handling of multiple/repeated births - Materials to obtain first: Endpoint definition, birth-verification procedures, cycle-accumulation rules and handling of multiple/repeated births - Further research: Obtain endpoint specifications to examine consistency of woman, cycle and infant units with cumulative counting. - Conditions for verification: Establish the cumulative-live-birth label, gestational-age criteria, woman/infant unit, accumulation across cycles, multiple-birth handling and verification procedures. - Why this condition is needed: Ovulation, pregnancy or infant counts must not be substituted for cumulative live birth per woman in the add-on comparison. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Event definition, unit and verification procedures remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N15 · Are the original reported rates of 34% versus 36% correctly attributed to the add-on 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; access was not attempted. - Current findings: The original results are letrozole 36%, combination 34% and berberine 22%. The selected add-on comparison is combination 34% versus letrozole 36%; no new difference or statistic was calculated. - Why unverified: The narrow attribution of arm names and reported rates matches the retained account. This does not establish formal between-arm effects or precision. - Reason for missing information: Arm-specific events, denominators and formal effect estimates remain separately unverified. - Required materials and conditions: Attribution records linking each arm’s reported rate, name, original statement and result location - Materials to obtain first: Attribution records linking each arm’s reported rate, name, original statement and result location - Further research: When new reports are obtained, first verify arm names for the same endpoint, denominator and time, then compare with existing attribution. - Conditions for verification: Link combination 34% and letrozole 36% to the selected comparison, while retaining berberine monotherapy 22% as original context. - Why this condition is needed: This condition preserves the observation for the add-on comparison without confusing the reported rates of the three arms. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Arm-specific events, denominators and formal effect estimates remain separately unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N16 · Are randomized, analyzed, completed and endpoint-observed denominators established for the two selected arms? 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; access was not attempted. - Current findings: Only 644 participants overall and 1:1:1 allocation are described. Arm-specific randomized, analyzed, completed and birth-observed denominators are absent. - Why unverified: The overall count and allocation ratio do not establish actual integer denominators for each arm. - Reason for missing information: Do not use a 644/3 calculation or denominators from the other sheet. - Required materials and conditions: CONSORT counts, analysis populations, birth-endpoint denominators and missing-data flow for the two selected arms - Materials to obtain first: CONSORT counts, analysis populations, birth-endpoint denominators and missing-data flow for the two selected arms - Further research: Compare definitions of each denominator to distinguish overall allocation from the endpoint-analysis population. - Conditions for verification: Distinguish the overall 644-participant allocation description from each selected arm’s randomized, analyzed, completed and endpoint-observed numbers. - Why this condition is needed: The combination and shared-control denominators must not be replaced by the overall 644 participants or by 644/3. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Do not use a 644/3 calculation or denominators from the other sheet. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N17 · Are live-birth event counts and person/infant/cycle units established for the two selected arms? 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; access was not attempted. - Current findings: The original results contain only three-arm percentages; original live-birth events, denominators and repeated/multiple-birth handling are absent. - Why unverified: Percentages alone cannot establish actual event counts or units. - Reason for missing information: No new back-calculation, rounded event counts or application of a 644 denominator is performed. - Required materials and conditions: Original live-birth events, denominators, units and handling of repeated/multiple births - Materials to obtain first: Original live-birth events, denominators, units and handling of repeated/multiple births - Further research: Obtain original result tables to examine the relation between events and person/cycle denominators. - Conditions for verification: Establish original event counts and the person/infant/cycle units of their denominators without back-calculating reported rates. - Why this condition is needed: Birth-event counts or numbers of women cannot be generated from the combination arm’s reported percentage alone. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: No new back-calculation, rounded event counts or application of a 644 denominator is performed. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N18 · Are the estimand, ITT/PP and handling of intercurrent events established for the add-on comparison on the common letrozole background? 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; access was not attempted. - Current findings: The current summary gives only the principal endpoint and arm-level reported rates; estimand, ITT/PP and specifications for intercurrent events or missing data are absent. - Why unverified: The effect of interest and actual analysis population for the selected comparison have not been established. - Reason for missing information: The distinction between monotherapy/add-on strategy effects and actual-intake effects remains unverified. - Required materials and conditions: Estimand specifications, ITT/PP populations and plans/results for intercurrent events and missing data - Materials to obtain first: Estimand specifications, ITT/PP populations and plans/results for intercurrent events and missing data - Further research: After establishing the estimand and analysis populations, examine how pregnancy, discontinuation and rescue treatment enter the definition. - Conditions for verification: Define whether the effect concerns the assigned add-on strategy or actual additional intake, and check handling of pregnancy, discontinuation, rescue, missingness and ITT/PP. - Why this condition is needed: The effect of assignment to added berberine may differ from the effect of actual additional intake. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: The distinction between monotherapy/add-on strategy effects and actual-intake effects remains unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N19 · Are the formal estimator, method, P and confidence interval for the selected between-arm effect established? 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; access was not attempted. - Current findings: Current material contains only arm-specific percentages; the formal effect estimator, model, P and confidence interval for the selected comparison are absent. - Why unverified: Reported rates are established, but material needed to establish formal comparison estimates and confidence intervals is absent. - Reason for missing information: No new difference, RR, OR, NNT, CI or P is calculated. - Required materials and conditions: Original between-arm effect estimates, model, P, confidence interval, denominator and endpoint time - Materials to obtain first: Original between-arm effect estimates, model, P, confidence interval, denominator and endpoint time - Further research: Compare original analysis tables and the SAP to establish comparison estimates and precision. Do not exclude an unadjusted analysis solely because it is unadjusted. - Conditions for verification: Establish original estimates, uncertainty, denominator and time from an appropriate unadjusted or adjusted between-arm model. - Why this condition is needed: A comparison of reported combination-arm rates must be distinguished from the precision of a formal add-on effect estimate. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: No new difference, RR, OR, NNT, CI or P is calculated. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N20 · Are clinical importance and interpretation criteria for superiority, noninferiority or equivalence established for the selected comparison? 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; access was not attempted. - Current findings: The text states that adding berberine to letrozole did not increase the birth rate. Importance thresholds, an equivalence hypothesis and confidence intervals remain unverified. - Why unverified: The original conclusion’s direction alone cannot newly establish clinical importance, equivalence or 0 effect. - Reason for missing information: Prespecified margins, confidence intervals and clinical-importance criteria are absent. - Required materials and conditions: Hypotheses, clinical-importance/noninferiority margins, effect confidence intervals and interpretation plans - Materials to obtain first: Hypotheses, clinical-importance/noninferiority margins, effect confidence intervals and interpretation plans - Further research: Establish original hypotheses and margins, then examine how confidence intervals address clinically important differences. - Conditions for verification: Compare prespecified hypotheses and clinical-importance criteria with the corresponding confidence interval; do not convert nonsignificance or original no-benefit wording into equivalence. - Why this condition is needed: The historical statement of no additional benefit differs from establishing an add-on effect of 0 or equivalence. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Prespecified margins, confidence intervals and clinical-importance criteria are absent. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N21 · Are losses, missing data, incomplete birth follow-up and the effects of their handling established for the selected arms? 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; access was not attempted. - Current findings: The current summary lacks counts of losses, missing data and incomplete birth follow-up, their handling and sensitivity results. - Why unverified: Material needed to establish how missing-data handling affects reported arm-level rates is absent. - Reason for missing information: Missing data are not treated as 0 events or complete follow-up. - Required materials and conditions: Arm-specific losses, missing/incomplete pregnancy or birth follow-up, reasons, handling and sensitivity material - Materials to obtain first: Arm-specific losses, missing/incomplete pregnancy or birth follow-up, reasons, handling and sensitivity material - Further research: Examine consistency between missing-data flow and planned handling and establish the scope of original sensitivity analyses. - Conditions for verification: Establish arm-specific missing data, reasons for loss, unconfirmed births, analysis handling and sensitivity analyses. - Why this condition is needed: The add-on strategy’s reported rate may also be affected by the handling of women who were not followed up. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Missing data are not treated as 0 events or complete follow-up. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N22 · Is consistency of preregistration, protocol and SAP with actual reporting established for the selected comparison? 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; access was not attempted. - Current findings: ChiCTR-TRC-09000376 and a historical statement matching the paper with WHO/ChiCTR are retained. Actual registration versions, SAP and change history have not currently been acquired or read. - Why unverified: The identifier and historical matching statement alone do not establish consistency between actual prior plans and reporting. - Reason for missing information: Registration timing, protocol, SAP and change history remain unverified. - Required materials and conditions: Actual ChiCTR/WHO registration versions, change histories, protocol, SAP and original result tables - Materials to obtain first: Actual ChiCTR/WHO registration versions, change histories, protocol, SAP and original result tables - Further research: If registration and SAP are obtained, compare registration timing, hypotheses, arms, endpoints and analyses with reporting. - Conditions for verification: Use the ChiCTR identifier as a lead to compare actual registration versions, timing, selected arms, endpoints and analysis plans with the original report. - Why this condition is needed: Establishing whether the add-on comparison was a prior question and analysis requires more than a registration number. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Registration timing, protocol, SAP and change history remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N23 · Are multiplicity handling and hypothesis hierarchy established for the two derived comparisons in the 3-arm trial? 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; access was not attempted. - Current findings: The original material describes a 3-arm RCT but lacks the two comparisons’ prespecified hierarchy, multiplicity handling and selection from all results. - Why unverified: Producing two sheets does not establish the trial’s hypotheses or handling of multiplicity. - Reason for missing information: Comparison hierarchy and multiplicity handling remain unverified. - Required materials and conditions: 3-arm hypothesis hierarchy, multiplicity handling, comparison-specific SAP and reported results - Materials to obtain first: 3-arm hypothesis hierarchy, multiplicity handling, comparison-specific SAP and reported results - Further research: Compare the original SAP with complete results to examine comparison selection and multiplicity handling. - Conditions for verification: Establish prespecified hierarchy and multiplicity handling for the monotherapy/add-on comparisons and examine reporting selection. - Why this condition is needed: Splitting a 3-arm trial into two sheets does not remove the need to establish the hypothesis hierarchy of the add-on comparison. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Comparison hierarchy and multiplicity handling remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N24 · Are both the lack of added benefit with the combination and the unfavorable direction for monotherapy 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; access was not attempted. - Current findings: The original three-arm rates, statements that monotherapy was lower and addition did not increase live birth, the separate metabolic counterview, D 34 and all axes are retained. - Why unverified: The current input’s unfavorable direction, counterview and assessment were not deleted or recalculated. This is not new certification of clinical effects. - Reason for missing information: Formal comparison statistics and precision remain unverified. - Required materials and conditions: Preservation records of original results, counterviews, original assessment, axes and correction history - Materials to obtain first: Preservation records of original results, counterviews, original assessment, axes and correction history - Further research: If subsequent reports become available, compare favorable, unfavorable and no-benefit material under the same attribution rules and record changes as separate events. - Conditions for verification: Retain all three-arm rates, original conclusions, D 34 and all axes without alteration rather than selecting favorable results only. - Why this condition is needed: The original statement of no added benefit must not be overwritten by metabolic possibilities or the monotherapy result. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Formal comparison statistics and precision remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N25 · Are the ChiCTR link, exclusion of the wrong NCT and original unverified registration-field wording kept distinct? 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; access was not attempted. - Current findings: The bibliography lists PMID 27336209, DOI 10.1016/j.fertnstert.2016.05.022 and ChiCTR-TRC-09000376. The original central point excludes NCT01116167, while the evidence-table registration field remains “not established.” - Why unverified: The established historical identifier link and the original field’s unverified label are retained separately. Current external registration verification is not complete. - Reason for missing information: Current comparison with original registration material was not attempted. - Required materials and conditions: Linked records of PMID, DOI, ChiCTR, excluded NCT and the original registration field - Materials to obtain first: Linked records of PMID, DOI, ChiCTR, excluded NCT and the original registration field - Further research: If actual registration material is obtained, compare identifiers, population, 3 arms and endpoints and record discrepancies as separate correction events. - Conditions for verification: Distinguish ChiCTR-TRC-09000376/PMID/DOI from the excluded NCT01116167 and retain the original registration field’s “not established” wording. - Why this condition is needed: This record condition prevents the add-on trial’s identity from being replaced with an incorrect registration for another trial. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Current comparison with original registration material was not attempted. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N26 · Is it explicit that the two derived sheets concern the same 3-arm trial and shared letrozole arm? 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; access was not attempted. - Current findings: The retained original set contains 1 3-arm RCT report. Both sheets share the same 644-participant trial and letrozole control and are not counted as independent repeated trials or two participant populations. - Why unverified: The two comparisons and shared-control relationship within the same report are explicit. - Reason for missing information: Exact comparison-specific analysis denominators and covariance remain separately unverified. - Required materials and conditions: Relationship records linking one study family, two comparisons, the shared control and report identifiers - Materials to obtain first: Relationship records linking one study family, two comparisons, the shared control and report identifiers - Further research: Compare report populations, recruitment and identifiers to link the same family and duplicated controls. - Conditions for verification: Group the same PMID/DOI/ChiCTR into one study family and do not add the two sheets/shared control as independent trials or independent participants. - Why this condition is needed: Even when the add-on and monotherapy comparisons are presented as separate questions, the control arm of one study must not be counted twice. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Exact comparison-specific analysis denominators and covariance remain separately unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N27 · Are independent replication of the same question and the scope of access established? 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; access was not attempted. - Current findings: The current retained set contains 1 direct 3-arm trial report, and no new clinical search was performed. Material establishing independent replication is absent. - Why unverified: Having one retained report does not prove that only one human study, or no human studies, exist worldwide. - Reason for missing information: Independent replication of the same question and comprehensive search scope remain unverified. - Required materials and conditions: Identities, reports, search scope and original material for independently recruited studies - Materials to obtain first: Identities, reports, search scope and original material for independently recruited studies - Further research: If new-material review is authorized, assess question-specific eligibility and compare independent recruitment and report overlap. - Conditions for verification: Establish separately recruited studies of the same population, exposure, comparator and live-birth endpoint and the actual reading/search scope. - Why this condition is needed: Replication of the add-on comparison is not obtained by splitting the same 3-arm report into two sheets. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Independent replication of the same question and comprehensive search scope remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N28 · Is there an applicable set for examining heterogeneity between distinct independent trials in the current retained material? 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; access was not attempted. - Current findings: The current retained report set contains only 1 direct 3-arm trial, and the two derived sheets belong to the same family. Multiple families needed for between-independent-trial heterogeneity are absent from this set. - Why unverified: There is no narrow applicable set for between-independent-trial heterogeneity in the current retained material. Unverified replication, denominators or precision are not changed to not applicable. - Reason for missing information: Whether other studies exist remains separately unverified. - Required materials and conditions: Differences in populations, interventions, endpoints and analyses if additional independent families are established - Materials to obtain first: Differences in populations, interventions, endpoints and analyses if additional independent families are established - Further research: If a separate family is obtained, compare question matching and design, population and estimator differences to examine whether synthesis is appropriate. - Conditions for verification: Do not treat two comparisons in the same study family as between-trial heterogeneity; examine differences between trials when another eligible family is established. - Why this condition is needed: The single family retained for the add-on comparison must not be synthesized as multiple independent trials. - Application exceptions: With only one family in the current retained set, between-trial heterogeneity is not applicable. Differences between comparisons within the study and dependence from the shared control remain subject to examination. - Remaining uncertainty: Whether other studies exist remains separately unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N29 · Are additional dose-response, long-term and subgroup effects beyond the tested dose and duration established? 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; access was not attempted. - Current findings: Current material describes berberine 1.5 g/day, a maximum 6-month plan and results in specific infertile women with PCOS. Direct material for other doses, longer durations and subgroups is absent. - Why unverified: Current material does not establish extended effects. This gap differs from the formal effect of the narrow selected comparison. - Reason for missing information: Extensions beyond the trial dose, duration and population remain unverified. - Required materials and conditions: Direct comparisons and interaction data for additional doses, durations, products and subgroups - Materials to obtain first: Direct comparisons and interaction data for additional doses, durations, products and subgroups - Further research: When an extension question is needed, review its material as a separate question; do not read simple subgroup differences as interactions. - Conditions for verification: Establish extensions beyond actual trial dose, duration and population using direct material and prespecified subgroup/interaction evidence. - Why this condition is needed: A trial adding 1.5 g/day must not be extended to added effects at every dose, formulation or long-term exposure. - Application exceptions: The narrow selected comparison does not always require evidence for every dose. Direct evidence is needed only for the extension being considered. - Remaining uncertainty: Extensions beyond the trial dose, duration and population remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N30 · Is the boundary between live birth and ovulation, pregnancy, metabolic or weight outcomes maintained? 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; access was not attempted. - Current findings: The original explanation and summary state that only live birth is assessed and that ovulation, pregnancy rates, insulin and weight are separate. This sheet maintains that boundary. - Why unverified: The narrow record condition of keeping the principal birth endpoint distinct from other intermediate outcomes is established. - Reason for missing information: Actual effects on other outcomes are not newly assessed in this sheet. - Required materials and conditions: Records of endpoint hierarchy and attribution boundaries for other outcomes - Materials to obtain first: Records of endpoint hierarchy and attribution boundaries for other outcomes - Further research: If material on other outcomes is obtained, review those questions separately and interpret their relationship to live birth only where direct linking evidence exists. - Conditions for verification: Fix the selected comparison to cumulative live birth; do not replace it with possible effects on ovulation, pregnancy rates, insulin or weight. - Why this condition is needed: The combination arm’s birth outcome must not be replaced with another intermediate endpoint. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Actual effects on other outcomes are not newly assessed in this sheet. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N31 · Are funding, product support, author relationships and conflicts specifically established? 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; access was not attempted. - Current findings: The original funding field says “public research network; details of product support to be checked.” Exact institutions, product providers, author relationships and conflict details are absent from current material. - Why unverified: The support category is described, while product-support and conflict details are not established. Generic English wording does not replace actual declaration content. - Reason for missing information: Detailed supporting institutions, providers, author relationships and declarations remain unverified. - Required materials and conditions: Funding bodies, research funding, berberine/letrozole supply, author relationships and conflict declarations - Materials to obtain first: Funding bodies, research funding, berberine/letrozole supply, author relationships and conflict declarations - Further research: Compare original support and conflict statements to establish research funding, product provision and author relationships separately. - Conditions for verification: Distinguish the public-research-network description from actual supporting institutions, product supply, author relationships and conflict declarations. - Why this condition is needed: Funding and product influence on the add-on result must also be examined using the actual support provided to each arm. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Detailed supporting institutions, providers, author relationships and declarations remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N32 · Are actual maternal, pregnancy and fetal/neonatal harm collection and results established for the selected comparison? 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; access was not attempted. - Current findings: The retained account contains warnings about trying to conceive and fetal/neonatal bilirubin. Actual harm events, person denominators, timing, collection, discontinuation and comparative risks for the selected arms are absent. - Why unverified: Warning statements do not replace confirmation of actual trial harm collection and incidence. - Reason for missing information: Actual harms, incidence, exposure after pregnancy and long-term safety remain unverified; no new calculation is performed. - Required materials and conditions: Arm-specific maternal/pregnancy/fetal/neonatal harm definitions, active collection, event/person denominators, timing, exposure, discontinuation and concomitant-treatment material - Materials to obtain first: Arm-specific maternal/pregnancy/fetal/neonatal harm definitions, active collection, event/person denominators, timing, exposure, discontinuation and concomitant-treatment material - Further research: If original safety material is obtained, compare harm definitions, collection, follow-up, denominators and discontinuation/concomitant treatment to distinguish signals from comparative risks. - Conditions for verification: Establish arm-specific harm definitions, collection methods, exposure/follow-up timing, event/person denominators, discontinuations, concomitant treatment and actual results. - Why this condition is needed: Separate harm information is needed because the live-birth add-on comparison alone cannot establish safety during pregnancy. - Application exceptions: The principal question concerns live-birth efficacy. Harm incidence is not imposed as an efficacy condition, while safety explanation and personal application require separate direct material. - Remaining uncertainty: Actual harms, incidence, exposure after pregnancy and long-term safety remain unverified; no new calculation is performed. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N33 · Are historical warnings, actual trial harms and personal-use decisions kept distinct? 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; access was not attempted. - Current findings: The original safety warning and advice to consult a clinician are retained as historical text. This sheet does not turn them into new incidence estimates, established causation, safety at every dose or personal dosing instructions. - Why unverified: The narrow record condition distinguishing historical warnings from this sheet’s material-verification scope is established. - Reason for missing information: Actual risk estimates and personal-application decisions remain unverified. - Required materials and conditions: Records distinguishing the roles of original safety warnings and actual event material - Materials to obtain first: Records distinguishing the roles of original safety warnings and actual event material - Further research: If original safety material is added, establish the warning’s basis, population and events and examine these separately from comparative risk. - Conditions for verification: Retain original warnings without promoting them to actual risk estimates, established causation, long-term safety or new personal dosing instructions. - Why this condition is needed: The roles of the add-on strategy’s live-birth result and the original pregnancy warning must be distinguished. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Actual risk estimates and personal-application decisions remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N34 · Are the original review date, correction date, current sheet-review date and search/reading limits 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; access was not attempted. - Current findings: The original verdict review date is 2026-09-08 and the correction date is 2026-09-14. This sheet’s record-review date is 2026-10-05; no new clinical search or external reading was performed. Original search-cutoff and report-specific access-date fields are absent. - Why unverified: Different dates and the actual current reading scope are recorded separately. This does not establish a completed comprehensive current search. - Reason for missing information: Individual external access dates and search cutoff are not inferred where original fields are absent. - Required materials and conditions: Original review date, correction date, current review date, search scope and actual access records - Materials to obtain first: Original review date, correction date, current review date, search scope and actual access records - Further research: When new evidence updating is authorized, record search dates, scope and access events as new records without backdating historical dates. - Conditions for verification: Retain original review and correction dates, do not infer absent cutoff/individual access dates, and distinguish whether current new searches or external reading occurred. - Why this condition is needed: The currency of the add-on comparison cannot be established by changing the original review date into a new search date. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Individual external access dates and search cutoff are not inferred where original fields are absent. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N35 · Are English correction history, remaining generic wording and original null/empty values 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; access was not attempted. - Current findings: The record describes a 2026-09-14 correction of 1 generic English field while retaining grade, score and date. Remaining generic sample-size, funding and result English wording, empty detailed-claim lists/review notes and the complete original material are retained. - Why unverified: Specific translations remain distinguished from generic verification wording, without generating absent numbers, confidence intervals or funding declarations. - Reason for missing information: Remaining generic wording does not establish the corresponding actual clinical values. - Required materials and conditions: Preservation records of original corrections, bilingual fields, empty values, absent fields and types - Materials to obtain first: Preservation records of original corrections, bilingual fields, empty values, absent fields and types - Further research: If a new bilingual correction is needed, establish corresponding original values and the correction event and preserve the entire text before change. - Conditions for verification: Do not fill the 1 correction or remaining generic verification wording with specific clinical values; retain empty detailed-claim lists/review notes, original axes and original types. - Why this condition is needed: Generic English wording must not be read as new denominator, CI or funding facts in the add-on comparison either. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Remaining generic wording does not establish the corresponding actual clinical values. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N36 · Are current retained-summary reading, unread external materials and historical matching statements separated by material type? 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; access was not attempted. - Current findings: The complete bilingual account, evidence, bibliography and corrections in the current input were read. The original research-review statement about matching the paper with WHO/ChiCTR is historical; all current external access and reading remain unattempted. Individual access dates/levels are null because the original fields are absent. - Why unverified: Actual current local reading, unread external materials and historical statements are separated by physical material type. - Reason for missing information: Individual external access levels/dates and current full-text versions remain unverified. - Required materials and conditions: Material-specific current acquisition/reading, URL, version, historical access date/level and original-location records - Materials to obtain first: Material-specific current acquisition/reading, URL, version, historical access date/level and original-location records - Further research: If external material is actually obtained, verify its own access event, version and original locations and update only the current reading record. - Conditions for verification: Distinguish actual reading of retained summaries from current unattempted/unread PubMed, DOI, paper, registration and SAP materials; do not inherit access levels/dates from a containing document. - Why this condition is needed: Reading the add-on comparison’s summary must not be represented as rereading the paper’s full text. - Application exceptions: This condition concerns actual material for the selected comparison. Information from another known comparison, a non-live-birth endpoint or a different product does not substitute for it; no arbitrary sample-size threshold is imposed. - Remaining uncertainty: Individual external access levels/dates and current full-text versions remain unverified. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ### N37 · Are joint-estimation and dependence-handling materials established for the two comparisons using the shared letrozole arm? 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; access was not attempted. - Current findings: The original report concerns the same 3-arm trial; the two derived comparisons share the letrozole monotherapy arm. Joint-estimation, covariance, shared-control handling and multiplicity specifications are absent from the current material. - Why unverified: The shared-control relationship is known, but analytical material for joint interpretation of the two comparisons remains unverified. - Reason for missing information: Joint estimation and covariance remain unverified; no new pooling, NNT or difference calculation is performed. - Required materials and conditions: Joint-estimation, covariance and multiplicity/SAP material accounting for the shared control - Materials to obtain first: Joint-estimation, covariance and multiplicity/SAP material accounting for the shared control - Further research: If joint interpretation is needed, compare the original SAP and comparison-specific estimates and establish a method accounting for shared-control dependence. Do not pool the effects in the two sheets as independent. - Conditions for verification: Check the original analysis for joint-estimation, covariance, shared-control handling and multiplicity specifications for the two comparisons. - Why this condition is needed: The add-on and monotherapy comparisons are linked through one study’s shared control, so their two effects cannot be pooled as independent observations. - Application exceptions: A joint synthesis model is not always required to describe this add-on comparison alone. Joint synthesis or relative interpretation of the two comparisons requires material accounting for dependence from the shared arm. - Remaining uncertainty: Joint estimation and covariance remain unverified; no new pooling, NNT or difference calculation is performed. No new difference/CI/denominator/incidence/grade calculation or reassessment. - Referenced methods: Existing method principles distinguish population, active treatment, actual exposure, cumulative live birth, follow-up, denominators, analysis, shared controls, funding and harms. Detailed conditions are Chamgap’s own application. - Related sources: S01 ## Evidence and sources ### S01 Wu 2016 PCOS double-dummy 3-arm cumulative live-birth report [PubMed](https://pubmed.ncbi.nlm.nih.gov/27336209/) Materials actually read: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread; access was not attempted. Findings recorded in the summary: Described as a multicenter double-blind 3-arm trial with 644 participants overall and 1:1:1 allocation. Reported cumulative live-birth rates are 36% with letrozole, 34% with the combination and 22% with berberine. This sheet concerns the add-on comparison: combination 34% versus letrozole 36%. The historical statement that adding berberine did not increase the rate is retained, while arm-specific n/events, formal effects/P/CI, pregnancy follow-up length, product support, actual harms and joint analysis accounting for the shared arm remain unverified. The original planned berberine 1.5 g/day, common letrozole 2.5 mg on cycle days 3–7 increasing to 5 mg if needed, and maximum 6 months are not converted into actual individual exposure or the end of birth follow-up. The between-arm balance of actual background medication also remains unverified. Record level: Arm-level reported rates and the historical direction of conclusions in the retained summary. The paper, registration and analysis plan are currently unread; no separate effect calculation or reassessment was performed. Independent replication: The retained set contains 1 3-arm report linked to PMID 27336209, DOI 10.1016/j.fertnstert.2016.05.022 and ChiCTR-TRC-09000376. The two derived comparisons share the same study family and letrozole arm. Comparison-specific analysis denominators, covariance and unique participant totals remain unverified. This does not establish independent replication or a new sum of participants. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Retained Korean/English bibliographic and research summary | Retained copy acquired; summary read | Retained copy acquired; summary read | Stored existing review summary read | The currently retained summary was read. It is distinguished from the paper’s full text. Retained bilingual bibliographic and result summaries were read. Historical statements about matching the paper with WHO/ChiCTR are not inherited as current full-text or registration reading. The original review date is 2026-09-08, the historical correction date is 2026-09-14 and the case-record review date is 2026-10-05. Historical access level and access date for individual external materials are not recorded; current access was not attempted. | | PubMed bibliographic/abstract entry | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | Only a bibliographic entry is linked; current availability, abstract version and text were not checked. Retained bilingual bibliographic and result summaries were read. Historical statements about matching the paper with WHO/ChiCTR are not inherited as current full-text or registration reading. The original review date is 2026-09-08, the historical correction date is 2026-09-14 and the case-record review date is 2026-10-05. Historical access level and access date for individual external materials are not recorded; current access was not attempted. | | DOI/publisher entry | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | This DOI entry comes from retained bibliography. The current publisher page, full-text availability and version were not checked. Retained bilingual bibliographic and result summaries were read. Historical statements about matching the paper with WHO/ChiCTR are not inherited as current full-text or registration reading. The original review date is 2026-09-08, the historical correction date is 2026-09-14 and the case-record review date is 2026-10-05. Historical access level and access date for individual external materials are not recorded; current access was not attempted. | | Original paper full text | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | The physical full text’s direct URL, version and retained original remain unverified. A bibliographic link does not mean the full text was read. Retained bilingual bibliographic and result summaries were read. Historical statements about matching the paper with WHO/ChiCTR are not inherited as current full-text or registration reading. The original review date is 2026-09-08, the historical correction date is 2026-09-14 and the case-record review date is 2026-10-05. Historical access level and access date for individual external materials are not recorded; current access was not attempted. | | Original WHO/ChiCTR registration material | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | Only the identifier and historical matching statement are retained. The actual registration version, timing and change history remain currently unverified. Retained bilingual bibliographic and result summaries were read. Historical statements about matching the paper with WHO/ChiCTR are not inherited as current full-text or registration reading. The original review date is 2026-09-08, the historical correction date is 2026-09-14 and the case-record review date is 2026-10-05. Historical access level and access date for individual external materials are not recorded; current access was not attempted. | | Original protocol, analysis plan and supplementary material | External access not attempted; not acquired; unread | External access not attempted; not acquired | Not read | Identity, existence, direct URLs and versions of the protocol, analysis plan and supplementary material remain unverified. Retained bilingual bibliographic and result summaries were read. Historical statements about matching the paper with WHO/ChiCTR are not inherited as current full-text or registration reading. The original review date is 2026-09-08, the historical correction date is 2026-09-14 and the case-record review date is 2026-10-05. Historical access level and access date for individual external materials are not recorded; current access was not attempted. | ## 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; distinguish intervention variants. · 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 · Link studies and reports separately; do not omit studies 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 · Fix the result and effect of interest when examining randomization, deviations, missingness, 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. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. ## What the statuses mean - Met: The row’s narrow record, attribution, boundary or actual fact was checked. This does not certify efficacy, precision or safety. - Partly verified: Some required information is available, but actual exposure, definitions, denominators or other necessary material remain missing. - Unverified: The criterion cannot be established within the retained scope. This does not mean 0 effect, 0 studies or failed access. - Not applicable: The condition has no applicable subject within the current retained set and narrow question. This does not replace a material gap or unread status. - Not reported: Limited to the particular material or section actually read. It is not extended to the entirety of an unread external paper. - Review incomplete: A status for external original-material comparison not yet performed, distinguished from review of locally retained content. ## Change history 2026-10-27 · Public page/materials 1.0. The public view and materials were prepared from the existing 37 items. No new clinical search, verification-state change or scoring was performed.