Study Design
It is imperative that a research study be designed properly so that the conclusions are sound and valid. Basic study design (e.g. case-control, randomized), study objectives, and the appropriate subject population must be considered at this stage of development. We can help you design your clinical research project so that it is analyzable and the results are meaningful. We can also help write parts of your clinical research protocol such as:
- Study design and population
- Define primary, secondary, or safety endpoints to be analyzed
- Power and sample size calculations
- Schematic diagram of the trial design
- Statistical analysis plan
Sample Size and Power Calculations
We can provide sample size and power calculations to ensure that the study will have adequate power to detect the effects of interest.
Statistical Analysis Plan
We can define and write the statistical analysis plan section of your protocol for you. For most investigator initiated studies a statistical analysis plan is a brief summarization of the statistical approaches to be used when analyzing the data. For complex studies, an additional more detailed statistical plan can be developed as a "stand-alone" document that can undergo revisions outside of the protocol, and therefore not trigger an IRB re-review.
Randomizations and Sampling Techniques
Depending on the study design, we can provide you with the following:
- Randomizing treatment groups - the randomization process is done by using a random number-producing algorithm (for randomized studies).
- Selecting a random subject sample - the same algorithm can be applied to chart review type studies for the purpose of identifying subjects.
- Perform the matching process in a matched case-control study.
Statistical Analyses
We prefer to be involved with your study from the planning stages if we are to perform the statistical analysis, but it is not required. A formal error analysis to confirm the integrity of the information may be needed. We can provide general consultation about analyzing your clinical research data and can perform the analysis for you. Specific tables and figures can be developed to meet your needs.
Data Listings, Tables, and Figures
Data listings, statistical tables, and figures can be developed along with your analysis results. They support the written analysis results and clinical findings of the study. Data listings are usually needed for formal submissions to regulatory agencies such as the FDA, and show subject data in an organized way such as patient demographics, vital signs, medical history, adverse events, and efficacy data. Statistical tables usually show the data summarized in a comparative way, and usually include descriptive statistics and probabilities (p-values). We can also generate summary tables and figures for your publications and presentations.
Publication and Presentation Preparation
We can assist in writing the non-clinical sections of your manuscript or presentation. Statistical writing requires specific terminology that compliments the clinical aspects, and should describe the statistical results accurately. We have co-authored papers and can help you with yours. We can also provide assistance with abstract submissions and presentations.
For biostatistical consultations, assistance with study design, analyses, manuscripts or presentations, please contact:
Henry SweeneyManager, Research Support Services
1301 Punchbowl Street, UH Tower 619
Honolulu, Hawaii 96813
Phone: 808.537.7825
Fax: 808.547.4055
Email: hsweeney@queens.org Kathleen Baker, Ph.D.
Biostatistician
1301 Punchbowl Street, UH Tower 621
Honolulu, Hawaii 96813
Phone: 808.585.5198
Fax: 808.547.4055
Email: kbaker@queens.org
Hours: MWF 4pm-7pm only.