Investigator/ Reviewer Guidelines

Please refer to this link for guidelines on writing a study protocol:

http://health.usf.edu/research/ocr/index.htm

Required Elements of the Scientific Review:

• Introduction, Background and Specific Aims
• Study Design/Study Outcomes
• Data, Statistical Analysis and sample size
• Research Procedures
• Eligibility Criteria
• Proposed Tests and Measurements
• Data and Safety Management
• Adverse event Reporting
• Control for Bias/Random Error

Additional Considerations in your study design:

For the section Control for Bias and Random Error, please review the following guidelines:

I:  Controlling bias

1. Choice of the control intervention

• Is the control group inferior?
• Does true uncertainty about relative effects of competing treatment exist?
• Are you in equipoise?
• If this is not a randomized controlled trial (RCT), why control arm was not employed. Articulate uncertainties/unknown, why this study needs to be done.
• Please cite a SYSTEMATIC REVIEW justifying your research. If there is no systematic review, should you do it? If not, why not?

2. How is potential selection bias handed? (is your study comparing like with like?)

• Describe a sequence generation (randomization); if the study was not randomized controlled trial (RCT) describe the plan for sampling/enrollment
• Consecutive patients vs. convenient sample, how will you make sure that  participants in the exposed and unexposed groups are similar in all important respects except for the exposure?
• If appropriate, describe case definition.
• How will you protect allocation concealment? i.e. conceal treatment assignment until after the treatment has been allocated

3. Do you need to control for performance bias i.e. take into account, co-intervention/contamination?               

• Does your study need to be blinded to a) participants, b) personnel, or c) collection of outcomes?
• How will you handle incomplete outcome data (attrition and exclusions)?
• Describe the plan to assure completeness of follow-up and account for all withdrawals and drop-outs
• Will analyses be according to ITT principles?
• How will you prevent detection/observer bias?
• Do you need a method to prevent inadequacies in outcome assessment (e/g. “blinding” of those who will assess the outcome, use standardized assessment e.g. RECIST criteria in evaluation of tumor responses etc)
• How will you avoid publication bias/selective outcome reporting?
• Have you registered your trial?
• Will you have control over publication (i.e. assure that publication will adhere to the protocol)?
• Other (including topic-specific, design-specific) (specify design: RCT, cross-over, single arm etc)

II:  Controlling for random error

a) Select realistic effect sizes (ideally informed by systematic review) b) specify alpha and beta errors c) sample size

III:  Risk-benefit analysis

Attached is the article that has a nice algorithm about ethical risk-benefit analysis

For any additional information related to the Scientific Review Committee, please contact

saslam@health.usf.edu