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”>
You, the investigator, want the participants to comply with your study procedures, which involves coming back for study visits as scheduled and participating fully in every assessment, and may involve taking medications at home in an appropriate manner for an extended period of time. The participant, however, decides whether to follow your requests or not. Following these requests is called “adherence to the protocol.” You have no real control over whether or not the participant adheres to the study protocol, or even ever returns for another study visit. Thus, you need to define what participant adherence actually means for the study, how you will measure it, and what is “good enough.” Then you need to implement procedures to maximize participant adherence. If you cannot obtain reasonable adherence to the protocol, you may be unable to interpret the results of the study. We focused on a critical component of adherence, retaining participants, in Section 13.4, and here we focus on other aspects of adherence. As retention is such a critical component of adherence, however, we also briefly discuss methods to improve retention of participants in this chapter as well.
24.1 Efficacy versus Effectiveness
In the simplest possible case, you give a single dose of an intervention under direct observation and wait to observe the change in some measure within a participant. With a single dose of an IV infusion, what happens if a participant wants the infusion stopped in the middle, or has a reaction to it? First, you have to stop the infusion immediately. Second, you have to include the adverse reaction in the results of the study, since all safety problems need to be identified and reported. But what do you do about any outcome measures? If you exclude this participant’s results on the outcome measures from the results of the study, you would be assessing efficacy: the effect of the intervention in those who followed the study protocol. If you include the participant’s results, despite getting only a partial dose of the infusion, you would be assessing effectiveness: the effect of the intervention in the entire population in the study, whether they received the full infusion or not.
Efficacy assesses the actual biological effect of the intervention on the outcome. Thus, participants who do not get the full intervention are excluded from the analysis. Before an intervention can be thought to be potentially useful, efficacy has to be demonstrated. If there is not a biological effect, then the intervention cannot have a real benefit. However, efficacy alone is not enough. Even when an intervention is efficacious when used properly, if participants and later patients will not use the intervention because of tolerability or convenience issues, then the intervention will not be effective. Effectiveness measures how useful an intervention will be in actual practice.
You are studying an existing therapy for the treatment of a new indication. You recruit 10 participants with the condition for your study. One problem, already known about the therapy, is that it tastes awful. Among your 10 participants, 3 take the medicine with every meal, as they are supposed to, and all 3 of them show substantial clinical improvement of the condition. This suggests that the therapy is efficacious for the condition. The other 7 participants, however, take the medicine irregularly, or stop after one or two doses, because of the horrible taste. These 7 participants show little or no clinical improvement. Overall, then, the effectiveness of this medicine is rather poor.
This has implications on how you consider describing your intervention. For efficacy, you would be focusing on those who are adherent with the intervention. These participants use or take the intervention reliably, follow the instructions for use reliably, and in general would be considered good participants. You would use what is sometimes termed a “per protocol” analysis, summarizing the outcome only in those participants who followed a set of prespecified criteria. These criteria do not mean that the participants have to be perfect, however. For many oral medications, missing no more than 1 in 5 doses, so adherence is at least 80%, is considered good adherence. However, this level of adherence might be considered very low for some illnesses.
You are studying a new medication regimen for patients who are HIV positive who have failed multiple previous therapeutic regimens and are currently not virally suppressed. The new medication regimen needs to be taken twice a day, based on pharmacokinetic data, and is very intolerant to missed doses. In an early study seeking to show that the medicine can work, you would focus on the efficacy of the regimen in participants who adhere to the protocol. Thus, you might require at least 95% adherence for a participant to be considered adherent, so you would only include participants who took the medicine every day, twice a day, missing less than 1 dose every 10 days, when you analyze the results.
The initial studies were very promising, with remarkable efficacy in this population. Thus, an effectiveness study was begun.
Another study with the same medication regime as in Example 24B evaluated the effectiveness of the medication. Therefore, outcomes were analyzed for all participants, whether or not they took the medication regimen regularly or reliably. Furthermore, as is common in many studies of effectiveness in HIV medications, if participants did not complete the study, they were assumed to be failures, so that effectiveness would be conservatively measured.
24.2 Adherence: How Is It Assessed?
Adherence is a relatively new term, replacing the older term of “compliance” to describe participants’ behavior with respect to the protocol requirements. Compliance seemed to indicate a passive role for the participant in letting the study personnel do things to them. Adherence includes the actions and responsibilities of the participant as part of the study team. Adherence is assessed for all aspects of the study protocol, to assure that the participant is using the intervention as it is supposed to be used and participating in other protocol activities. As the investigator, you would like participants to be perfectly adherent. This includes
following timing and dosage regimens faithfully;
recording information as required, such as recording reactions to the treatment, food intake diaries, exercise activities, etc.;
participating in specific activities, e.g., exercise classes, during the treatment period;
attending all clinic visits; and
participating fully in all evaluations designed to assess the effectiveness of the treatment.
As you realize that no one is perfect, you need to set minimum standards for what would be considered acceptable adherence. The first aspect of adherence would seem simple to assess: Are participants adhering to the study treatment or not? But even in the simplest case, when you are directly observing whether the participant follows the treatment, defining acceptable adherence is often quite difficult.
