The number of subjects in a clinical trial is usually determined by the primary objective of the trial. Steps in the sample size evaluation procedure:

• choosing the appropriate test statistic,
• determining the null hypothesis,
• determining the alternative (working) hypothesis at the chosen dose(s),
• the probability of erroneously rejecting the null hypothesis (the Type I error),
• the probability of erroneously failing to reject the null hypothesis (the Type II error),
• the approach to dealing with treatment withdrawals and protocol violations,

The method by which the sample size is calculated will be described in the protocol, together with the estimates of the parameters used in the calculations (such as variances, mean values, response rates, event rates, difference to be detected).

The biostatistics part of the study plan contains the following items:

• A description of the statistical methods to be employed, including timing of any planned interim analysis(ses).
• The number of subjects planned to be enrolled.
• Reason for choice of sample size, including reflections on (or calculations of) the power of the trial and clinical justification.
• The level of significance to be used.
• Criteria for the early termination of the trial.
• Procedures for handling missing, unused, and spurious data.
• Procedures for reporting any deviation(s) from the original statistical plan (any deviation(s) from the original statistical plan should be described and justified in the protocol and/or in the final report, as appropriate).
• The statistical and data handling software to be used.

• The procedure preparation will be done according to our SOPs and the actual study protocol.
• We generate the randomization code using SAS® software or nQuery Advisor®.
• Each randomization data set will be stored in a fire safe closed area.
• One paper copy in closed envelop for the sponsor.
• Preparation of Closed Randomization Envelopes.

• The CRF contains all the information comprised in the protocol (It should focus on the data necessary to implement the planned analysis, including the context information (such as timing assessments relative to dosing) necessary to confirm protocol compliance or identify important protocol deviations.)
• Important! - coding all information
• QA check to compare the information in protocol and CRF
• Preparation of CRF Completion Guidelines
• Organization of the printing and distribution of CRFs and the Completion Guidelines

• We use ORACLE® database or SAS® dataset to store the data
• For each continuous variable interval criterion, and dictionary for category or string data,
• Automatic check of each criterion
• CRF review before Screen Testing,
• Making Study Database or File and Study specific guideline,
• Security - individual password (no internet connection to the data entry and statistical equipments)
• Initial training to data entry staffs,
• Single or double data entry depending on the Sponsor's requirements.
• Data entry check according to predefined procedure

Standard chapters

• Demography and anamnesis - descriptive statistics
• Example of table header for continuous variables
• N Mean S.D. Median Min Max
• Example of table header for categorical variables
• Ntotal Cat1(N,%) Cat2(N,%) Cat3(N,%)
• Safety analysis - descriptive statistics and individual lists
  • AEs grouped by severity, outcome, relationship to study medication
  • If coding of adverse events was required by the sponsor then AEs grouped by body system and AE name

Non-standard chapters

• Efficacy analysis
• Primary efficacy analysis performed usually on both ITT and PP population
• Secondary efficacy analysis performed on ITT population only

Data Cleaning Plan deals with the process of detecting, diagnosing, and editing faulty data:

• Detecting:
  • Lack/Excess of data
  • Outliers/Inconsistencies
  • Strange patterns
  • Suspect analysis results
• Diagnosis:
  • Errors, missing data
  • True extreme
  • True normal
  • No diagnosis still suspect
• Editing:
  • Correction
  • Deletion
  • Leave unchanged

• SINGLE DATA ENTRY = typed into the computer database by a single individual.
• DOUBLE, BLIND DATA ENTRY = typed by one individual, then typed by a second individual, with the computer comparing each keystroke. Differences are noted in a report to the supervisor, so that all three can look at the original entry to come up with the best decipher of the letters in question.

The Scope of MedDRA we use for coding under medical control:

• Diseases
• Diagnoses
• Signs
• Symptoms
• Therapeutic indications
• Investigation names & qualitative results
• Medical & surgical procedures
• Medical, social, family history
• Terms from: COSTART®, WHO-ART®, HARTS®, J-ART®

• Each query checked by database coordinator and sent to CRA and/or monitor and/or investigator by fax, email, self-addressed envelop or personally.
• Modifications are made according to the answers.

• Preparation of database closing includes running dictionaries, processing lab normal, running final global validations and edits, cross checking the number of cases.
• Each correction is tracked by the system and signed by the investigator and/or monitor.
• Finally the hard database closing is performed by writing it on a protected CD, and the working file for statistical analysis.
• The CD will be stored in a fire protected closed area.

• The statistical analysis will be done according to SAP, after breaking the blind.
• We will electronically document each activity of the statistician in the log file and the tracking file (Audit Trail).

• We make the tabulated form using SAS®

• The statistical report contains (according to ICH E3 guideline) the following items:
  • Objective of the study.
  • Design of the trial,
  • Data quality controls,
  • Treatment Group Comparability (if applicable)
  • Efficacy Analyses (Sample Size Estimation, Endpoints, Patient Subsets, Evaluability Criteria, Handling of incomplete Data, Subset Analyses of Efficacy)
  • Safety Analyses (Patient Included, Endpoints)
  • Other Analyses
  • Deviations from the SAP
  • Statistical Methodology (Statistical Software, Treatment Group Comparability, Efficacy Analyses, Safety Analyses)
  • Results (Treatment Group Comparability, Efficacy Analyses, Safety Analyses)
  • Conclusions
  • Appendices (Summary, Individual data)

There is a minimum level of 20% checking for CRF, but as our partner requires it is possible to offer 100 % verification in the certain cases.

Referenciák