html xmlns=”http://www.w3.org/1999/xhtml” xmlns:mml=”http://www.w3.org/1998/Math/MathML” xmlns:epub=”http://www.idpf.org/2007/ops”>
In the previous chapter we described the purpose of blinding and the different levels used. In this chapter we describe some techniques to implement and maintain the blind. These techniques may be used in single-blind, double-blind, and completely blinded studies as needed. Although blinding is usually associated with interventional studies, it should also be done for observational studies whenever possible. We describe methods for blinding observational studies in Chapter 28.
23.1 Masking the Intervention
We assume that a randomization schedule has been generated according to the methods in Chapter 21 by an individual not involved in the assessment of participants, and that participants accepted into the study are sequentially assigned a randomization number that is linked to the randomization schedule and that determines the allocation of the specific intervention. To protect privacy, each individual is given a study number when first identified as a potential participant (Section 29.1), but the randomization number is only assigned when the participant has passed all the screening criteria for the study, and it is different from and unrelated to the study number.
Blinding methods in randomized studies include not only physically masking the material that is used in the intervention but also ways of masking other information, such as laboratory results, that might reveal the intervention.
23.1.1 Masking a Pharmaceutical Intervention
There are many techniques that can be used for creating the material to blind an internal or topical pharmaceutical intervention. The simplest case is when there is a single intervention, which we will refer to as the active treatment, and a placebo, and the placebo can be created in the same form as the active treatment – as a tablet, capsule, cream, or liquid. The placebo should be identical to the active treatment in physical appearance and have similar taste, smell, and consistency. Often the manufacturer of the active drug will prepare matching placebos for use in randomized studies. In a similar way, a study involving different doses of the same medication can be blinded by giving each participant the appropriate mix of active drug and placebo tablets or liquids.
Participants will be randomized to one of three groups in a study to determine optimal dosage levels for a new use for an existing drug. The drug levels, 5 mg, 10 mg, or 15 mg twice daily have been shown to be safe. The drug is manufactured in 5 mg tablets. The manufacturer will provide the identical placebo tablets. All participants will take three tablets each time, but they will be given a combination of active drug and placebo tablets to achieve the correct dose.
If the protocol is a comparison of two drugs, such as a new treatment versus standard of care (SOC), and the two treatments are administered using equivalent vehicles, it may be possible to administer medications that contain either the test or the SOC treatment but are identical in appearance. If the two treatments have different forms or may be distinguishable for other reasons, such as taste or texture, then there should be matching placebos for each treatment, and a participant would be given the correct combination of placebo and active drug according to the randomization. This method can be used even if the method of delivery of the two treatments is different. For example, a participant could receive both a liquid medication and a tablet, one of which would be a placebo.
A study is designed to examine the additive effect of two different treatments; one is administered as a tablet and the other as a gel. The design calls for four groups; placebo only, the first treatment only, the second treatment only, or both treatments together. Each participant is given the appropriate combination of active or placebo tablet and active or placebo gel, depending on the assigned group.
If a treatment or treatments can be prepared in capsule form, then this may be very useful in maintaining the blind. A capsule can conceal some easily identifiable characteristics, such a strong taste, that might be hard to duplicate in a placebo. A capsule can also often be created to contain different doses of a drug without any visible change.
In Example 22B two treatments for depression were compared. The treatments were administered in capsules. The capsules were identical in appearance and could contain different dosages. This method also meant that if dosage changes were required, the pharmacist could simply change the content of the capsules without any visible changes to the capsules.
23.1.2 Maintaining the Mask When Dosage Levels Change
Sometimes the study protocol defines a periodic evaluation of the drug efficacy and dose changes, either up or down, for some participants. In this case, the blind can still be maintained. The evaluation and decisions about doses are made by an outside unblinded investigator and the treatment changes will also be determined by the same or another unblinded individual. If the testing and adjustment is on a fixed schedule, then each participant will have a change in the physical amount of medication, although the actual active drug may not change.
In a placebo controlled interventional study, the drug levels may be adjusted after three months. The manufacturer has supplied identical tablets with either 0 mg, 2.5 mg, or 5 mg of drug. Each participant begins by taking one tablet with either 0 mg (the placebo) or 5 mg of drug. The dosage may be adjusted either up or down based on clinical or laboratory results. After three months in the study the number of tablets each participant receives is increased by one. Participants in the placebo group will receive two placebo tablets. Participants in the active drug group may have their dosage increased by either 2.5 mg or 5 mg, but may also have their dosage decreased by replacing the 5 mg tablet with a 2.5 mg tablet and a placebo. Participants on active drug not needing a dosage change will simply be given an additional tablet containing placebo.
In Example 23D the testing could be repeated and the dosages further adjusted after a suitable period.
Some studies require that dosage modifications be done only as required, not on a regular schedule, and only for participants who require change. If only those participants receiving the active drug have dosage changes, then it will be clear that they are being treated with active drug. One way to maintain the blind in these circumstances is to create a schedule in which participants are randomly selected for an apparent dose change, usually the addition of a placebo. Depending on the study design, the investigators might also appear to reduce the amount of “medication” the participant is taking.
For a study similar to Example 23D, drug levels were monitored every four weeks and changes were made when necessary. The biostatistician generated a randomized schedule of changes. This identified which participants would get a dosage “change” in the form of an additional placebo tablet and the time when they would get the change. Real changes would be implemented as discussed in Example 23D. If the participant had had a real change in the prior interval, there would be no dummy change to avoid unnecessary anxiety, but real changes would be made as needed. Participants could receive up to two additional tablets. If more dosage changes were needed, these would be implemented by changes in the dose of the tablets given.
23.1.3 Masking for Non-Pharmaceutical Intervention
Blinding for studies that use other than oral or topical treatments is more difficult and tends to raise ethical issues. Nutritional studies may be blinded if all food is to be supplied by the investigators and the difference is not detectable, such as a tasteless supplement that is mixed with food. If the procedure is invasive, then it is usually not ethical to perform a non-active equivalent. Studies in which a placebo IV infusion, using the same carrier as an active infusion, have been allowed, but this might be problematic. The risks of a sham infusion are generally higher than those of an oral placebo therapy, particularly if the participant is unusually fragile. Studies with sham surgeries have also been done, but they are rare because of the risks to the participant. Most studies involving invasive procedures are single-blinded, but particularly for very aggressive therapy such studies are at significant risk of participant bias (Section 17.1.3).
Investigators wished to evaluate the efficacy of transcutaneous electric nerve stimulation (TENS) for treatment of chronic back pain. They wished to compare TENS alone to TENS with and without stretching exercise and no treatment. They were able to develop a sham unit with lights, dials, and wires that was identical in appearance to the TENS equipment, so that participants and most assessors were blinded to whether or not the participant actually underwent TENS. The assessors were also blinded to stretching exercises, although the participants could not be.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
