Services include statistical statements for study protocols, sample size calculation, data management, double data entry, quality control, statistical evaluation and reporting, prospective and retrospective analyses, meta-analysis and cost benefit calculations.

The statistical analysis plan is intended to be a comprehensive and detailed description of the methods and presentation of data, proposed for a clinical trial.

The statistical analysis plan should be included in the protocol. It may be contained in the body of the protocol, or it may be summarised and detailed in the appendix. When MediServ is involved in the development of the protocol, the statistical analysis plan still is reviewed/updated before the analysis begins.

All reviews/adaptations/extensions of the statistical analysis plan are performed before the blinding code is broken (or before analysis begins in an “unblinded” trial). Any changes in the statistical analysis plan are justified and fully documented in the statistical report.


The statistical analysis plan:

– States the variables that are considered in the analysis. A detailed definition is given of these variables. Primary, secondary, and supportive endpoints are clearly identified.

– Specifies the hypotheses to be tested and/or the parameters that are to be estimated in order to  meet the trial objectives.

– States the populations on which the evaluations are based. A detailed definition is given of the subject populations and the type of primary analysis is stated (e.g. intention to treat, ‘as randomised’, efficacy evaluable, etc.). Procedures of handling withdrawals and protocol deviations are mentioned.

– Gives a full and detailed description of the methods of analysis and presentation provided for each type of data.

Consideration is given to the following:

 Methods for handling multiple observations
– Methods for handling multicentre data
– Rules for derived variables-use of baseline values
– Use of covariate data
– Analysis of subgroups
– Interim or sequential analysis
– Rules for stopping the trial
– Methods for handling missing data
– Use of tabulation and/or other presentation methods
– Levels of statistical (and clinical) significance
– Use of confidence intervals.