Background: A variety of patterns of imperfect adherence have been reported including deviations in timing of doses as well as non-consecutive missed doses and drug holidays . Forgiveness is a drug specific property that arises from the relationship of the duration of action and the dose interval of the drug. When the duration of action greatly exceeds the dose interval then the drug is considered forgiving .
Aim: To develop a criterion to quantify forgiveness.
Methods: A criterion, relative forgiveness (RF), was developed to quantify forgiveness based on the probability of successful attainment of a treatment target. RF is defined as the number of times more likely that a target is successfully attained under perfect adherence compared to imperfect adherence. Values close to one indicate that a drug is forgiving to imperfect adherence and values close to zero indicate that the drug is particularly sensitive to imperfect adherence behaviour (i.e. not forgiving). Evaluation of the criterion was performed using simulation using MATLAB®. A 1-compartment instantaneous input multiple dosing pharmacokinetic (PK) model linked to an immediate effects Emax pharmacodynamic (PD) model was used. A target, for each individual of at least 90% of doses must have an effect at trough greater than 0.35 (units), was determined to indicate success. Between-subject variability was incorporated on the PKPD parameters and additive residual variability was added to the response. Random profiles of suboptimal adherence were generated. These profiles comprised timing variability, non-consecutive missed doses and drug holidays. 1000 individuals with each having a unique adherence profile were simulated. The influence of different types of imperfect adherence on RF was evaluated. Theoretical modifications to the drug were considered.
Results: The probability of success in achieving drug response was 0.4 and 0.67 for imperfect and perfect adherence, respectively, which resulted in a RF of 0.33. This means that therapeutic success was three times less likely with imperfect adherence. The RF of missed doses only and perfect timing was 0.42 whereas for imperfect timing but not missed doses was 0.88. The RF of a drug with longer half-life and greater potency, compared to the original drug with perfect adherence, were 3.03 and 1.21, respectively, indicating that increasing the half-life increased therapeutic success by 3 fold even in the presence of imperfect adherence.
Discussion: The influence of missed doses on RF is stronger than that of timing variability. Although patients are perfectly adherent, drug response may not be achieved. RF can be used as an index to determine how forgiving a drug is. This forgiveness index can be clinically useful with regards to the selection of forgiving drugs for patients with suspected or known suboptimal adherence.
1. Blaschke et al, Annu Rev Pharmacol Toxicol 2012; 52:275-301
2. Urquhart, Clin Pharmacokinet 1997; 32(5):345-56