Development↦Statistic Methodology↦Statistics in the Protocol↦Missing Data
What is it? Why is it important?
Missing study data is incomplete data. It can affect the study’s statistical power and/or create bias (i.e. distortion of statistical results).
Missing data can affect the reliability and credibility of study results. Therefore, it is important to:
- Take preventive measures to limit the amount of missing data
- Plan appropriate statistical strategies to handle missing data
Missing data may have different root causes, including:
- Incomplete questionnaires: participants may not know, or decide not to answer certain study questions (e.g. due to privacy concerns, topic sensitivity)
- Missing visit(s): in longitudinal studies, participants may miss follow-up visits voluntarily or involuntarily (e.g. for health reasons)
- Early study termination: participants may withdraw early or be lost to follow-up (e.g. due to health issues or death)
- Technical issues: malfunctions in study-related equipment may lead to data loss (e.g. failure of laboratory equipment during analysis)
What do I need to do?
As a SP-INV, define measures to minimize missing data in your study.
Potential strategies include to:
- Implement a study design that facilitates data collection (e.g. the use of an electronic Data Capture System (eCRF) with built-in validation checks)
- Train study staff on data collection and documentation procedures (e.g. correct and complete documentation during study visits)
- Implement supporting strategies for study participants to ensure compliance and minimize study drop-outs (e.g. clear study instruction leaflets, provide participants with a smart pillbox (i.e. a digital solution for monitoring and improve adherence to medication intake), use of e-mail reminder systems reminding participants of pending study visits)
- Plan on-site monitoring visits, and in particular central data monitoring
Discuss with a statistician on how to handle missing data. Plan statistical analyses that account for missing data
Describe methods for handling missing data in the study protocol and, if applicable, in the Statistical Analysis Plan (SAP).
More
Common statistical approaches used to address missing data are to:
- Complete case analysis – Analysing only participants with non-missing data
- Imputation – Estimating and replacing missing values (e.g. using multiple imputation)
Where can I get help?
Your local Research Support Centre↧ can assist you with experienced staff regarding this topic
Basel, Departement Klinische Forschung (DKF), dkf.unibas.ch
Lugano, Clinical Trials Unit (CTU-EOC), ctueoc.ch
Bern, Department of Clinical Research (DCR), dcr.unibe.ch
Geneva, Clinical Research Center (CRC), crc.hug.ch
Lausanne, Clinical Research Center (CRC), chuv.ch
St. Gallen, Clinical Trials Unit (CTU), h-och.ch
Zürich, Clinical Trials Center (CTC), usz.ch
References
ICH GCP E6 (R3) -see in particular:
- 3.11.4.5.4 Monitoring of clinical trial data - identify missing data
ICH Topic E9 – see in particular
- 5.3 Missing values and outliers
ICH Topic E8(R1) General considerations for clinical studies - see in particular
- 5.6 Statistical analysis