Investigators must minimize the risks faced by subjects to the best of their abilities. IRBs review protocols and procedures and are responsible for ensuring that risks are minimized. Risks are minimized when the investigator and study team are highly competent; studies are well-designed; selected procedures present the least probability of harm of those that are adequate to achieve the goals of the study; subjects are monitored during the course of the study; adverse events are quickly identified, managed, and reported; and subject privacy and confidentiality are protected. More information on these topics can be found in other modules.
To further reduce risk, the investigator should always enroll the minimum number of subjects needed to yield the requisite statistical power, expressed as the probability of a ß or Type 2 error. Typically, investigators aim for a statistical power of 80 percent, which means that the probability of missing a “true” difference is 20 percent. This ensures that the number of subjects exposed to risk is minimized. Investigators should also design the subject inclusion and exclusion criteria so that individuals least susceptible to harm are included in the study (to the degree possible). The study design should also include methods for dealing with potential changes in the risk status or circumstances of individual subjects or groups of subjects. Whenever possible, research should be conducted using the same procedures or interventions being utilized for other purposes, such as diagnosis or treatment. For example, a biomedical investigator might use residual blood, or “leftovers,” originally drawn for a therapeutic purpose, or capitalize on procedures already being done for clinical purposes (e.g., collect an additional small amount of spinal fluid in the course of a clinically required spinal tap rather than doing another spinal tap for research purposes). An investigator conducting educational, social, or behavioral-science research, might utilize the school’s standard academic measures (tests) instead of having students undergo an additional assessment.