You've already learned about the importance of randomisation. However, it's also important to address reducing other biases, which might occur in trials. This can be achieved using a method called blinding. Blinding is a procedure whereby one or more parties in a trial are kept unaware of which treatment arm participants have been assigned to. It's an important aspect of any trial, and is performed to avoid or prevent conscious, or unconscious bias in the design and delivery of a clinical trial. The different parties involved in a clinical trial are all possible sources of bias, including the patient being treated, the clinical staff administering the treatment, the physician assessing the treatment, and the team interpreting the results. Wherever possible, blinding should be performed in trials. There are essentially two types of blinding: single and double. Single blinding occurs when only one party is blinded, usually the participants. If both the participants and study staff are blinded, it's called a double blind study. As you can imagine, if you don't know what the patient is taking, your prior prejudice and the patient's prior prejudice is removed. Double blinding is a particularly useful method for preventing performance and detection bias. Performance bias refers to systematic differences between groups in care that is provided, or in exposure to factors other than the interventions of interests. For example, if the investigators know that an experimental group have been given an active drug, they may focus their attention on that group. Detection bias refers to systematic differences between groups in how outcomes determined. For example, the participants might receive more frequent exams and more diagnostic tests. This could result in the experimental group having a greater chance of a positive outcome. Blinding of those who assess the trial outcomes may reduce the risk that knowledge of which intervention was received, rather than the intervention itself, affects outcome measurement. The patient or doctor who trusts in the effect of a specific intervention may unconsciously, or intentionally, perceive or detect an enhanced treatment effect. This can be especially important for assessment of subjective outcomes, such as degree of post-operative pain. Blinding can also help prevent withdrawals from studies, which can lead to incomplete outcome data. For example, in a trial of aspirin vs placebo, imagine you know which patients are taking aspirin, and they come back and say, "I have indigestion". And you say, "Oh hang on, you're on aspirin, you better stop taking it". However, if you didn't know he was taking aspirin, you might be more likely to say, "Don't worry, we'll monitor you, come back to me in a week's time", and then we may find it went away. Now, in general, you want to avoid people dropping out of your trial, and blinding help such unnecessary withdrawals. Likewise, if a patient knows what treatment they are receiving, and feel like it's an inferior treatment, they may be more likely to drop out, or not follow the directions of the trial. And as a result, this biases the study findings. Withdrawals and anticipated dropout rates should be considered when calculating a sample size. The sample size of a trial is extremely important, and must be determined long before the trial starts. To calculate the sample size, you need to decide what power your study will have, and power increases with increasing sample size. The power of a study is essentially its ability to detect an effect, or an association, if one truly exists. This should be determined a priori to be at least 80% and preferably 90%. For example, a power of 90% means if the true difference between treatments is equal to the one we planned, then there's only a 10% chance that the study will not detect it. Back to blinding. Unfortunately, blinding is not always possible. For example, if you're comparing a surgical technique it may be impossible to blind that to the people involved, in particular the surgeon. It may also be ethically impossible to blind the patients, and it makes it more costly in its broadest sense, because it's very complicated to keep things blinded. It's also difficult if drugs require titrating, that is, if a drug requires that you increase or decrease the dose, it can add difficulties. Now you know the importance of blinding for ensuring internal validity in randomised controlled trials, and where possible trials should be double blinded.
