CLE Data Collection & CRF Management 3 — Questions and Answers
Question 1: When collecting data on exclusive breastfeeding duration, which denominator is most appropriate for calculating prevalence at six months?
- All infants born during the study period
- Only infants who initiated breastfeeding
- All infants who reached six months of age in the cohort (Correct answer)
- Only infants whose mothers completed all CRF visits
Correct answer: All infants who reached six months of age in the cohort
Prevalence at six months should use all infants who survived to six months as the denominator to avoid selection bias.
Question 2: A CRF query asks the lactation educator to clarify whether a reported feeding was 'breast' or 'expressed breast milk.' The best response is to:
- Answer based on the most common feeding pattern for that mother
- Review the source document and respond with the verified information (Correct answer)
- Close the query without response if the visit was over 60 days ago
- Escalate to the IRB before providing any clarification
Correct answer: Review the source document and respond with the verified information
CRF queries must be resolved by reviewing the original source document and providing verified, accurate information.
Question 3: Which scale of measurement is used when recording a mother's breastfeeding intention as 'plans to breastfeed' vs. 'does not plan to breastfeed'?
- Interval
- Ratio
- Ordinal
- Nominal (Correct answer)
Correct answer: Nominal
Categories without an inherent rank order or numerical value are nominal (categorical) data.
Question 4: An educator discovers that 15 CRF entries from a previous month were never signed. The FIRST corrective step is to:
- Back-date signatures to the original visit dates
- Notify the study coordinator and follow the site's data correction SOP (Correct answer)
- Destroy the unsigned forms and re-create them with correct signatures
- Submit a variance report to the FDA immediately
Correct answer: Notify the study coordinator and follow the site's data correction SOP
Unsigned CRFs are a protocol deviation that must be escalated to the coordinator and corrected per the site's SOP; back-dating or destroying records is fraud.
Question 5: In a breastfeeding support program database, a unique subject identifier (ID) primarily serves to:
- Link subject records across visits while protecting PHI (Correct answer)
- Identify the educator responsible for each data entry
- Indicate the infant's gestational age at enrollment
- Distinguish between full-term and preterm feeding protocols
Correct answer: Link subject records across visits while protecting PHI
A unique subject ID links longitudinal records while replacing personally identifiable information to protect subject privacy.
Question 6: During a monitor visit, the clinical research associate (CRA) requests access to a mother's complete medical chart to verify CRF data. The lactation educator should:
- Refuse access because only the IRB may review medical records
- Provide access only to the pages referenced in the CRF
- Facilitate access to relevant records per the signed informed consent and authorization (Correct answer)
- Require the CRA to submit a written formal request to hospital administration first
Correct answer: Facilitate access to relevant records per the signed informed consent and authorization
Signed informed consent and HIPAA authorization typically permit sponsor monitors to review source documents necessary for verification.
Question 7: A lactation educator wants to compare average milk volume expressed by mothers at weeks 2, 4, and 6 postpartum. The most appropriate statistical test is:
- Independent samples t-test
- Chi-square test of independence
- Repeated measures ANOVA (Correct answer)
- Mann-Whitney U test
Correct answer: Repeated measures ANOVA
Repeated measures ANOVA is designed to compare means across three or more time points within the same subjects.
CLE Data Collection & CRF Management 3
When collecting data on exclusive breastfeeding duration, which denominator is most appropriate for calculating prevalence at six months?