In this study, we have estimated the level of adherence to ART in paediatric HIV/AIDS patients using two commonly employed methods (caretaker report, and medication return) and one seldom used method (drug plasma concentration). Our data are similar to what has been reported in adult patients and suggest that care taker report and medication return may overestimate the true level of adherence. Adherence to Antiretroviral therapy in paediatrics is critical in order to maximize the benefit of medication. Inadequate adherence is associated with immunological, and virological failure; drug resistance, and treatment failure [4
]. In our study, high adherence levels to ART by caretaker report (98%) and medication return (97%) were observed. However, adherence level by nevirapine plasma concentration was significantly lower (85%) than the level estimated by caretaker report and medication return. Our findings are in agreement with those obtained in Cameroon in which a comparable proportion of patients were found to be adherent to ART by using nevirapine plasma concentration method [11
In addition, there was a weak agreement in estimating adherence between nevirapine plasma concentration method and care taker report (kappa 0.09) as well as medication return (kappa 0.131). This means, non-adherent patients were classified as adherent by caretaker report and medication return methods whereas in fact, were non-adherent.
Using medication returns and care taker report, our data show that only 3% of the patients were non adherent to antiretroviral medication. These findings are similar those reported from Uganda where the level of adherence assessed using medication return and caretaker reporting was relatively high [20
]. Nevirapine plasma concentration seems to give a better estimation of adherence to ART than caretaker report and medication return. Our study found a strong association between nevirapine plasma concentration (< 3 μg/ml) and immunosuppression. In addition, there was no association between medication return and immunosuppression. Similar findings have been reported from South Africa [13
]. Furthermore, our study shows that caretaker report and medication return have low sensitivity to detect non adherent children compared to nevirapine plasma concentration, which could detect a higher proportion of non adherent children. At the moment, medication return and caretaker report, are methods used routinely for measuring adherence in paediatric patients on ART in HIV care and treatment clinics. This means that, a great deal of non-adherent patients receive suboptimal treatment and this has clinical implication in terms of treatment outcome and the prevention of the emergence and spread of drug resistance.
Our data show that nevirapine plasma concentration may be a better predictor of adherence and correlates well with the immunological response compared to medication return and caretaker report. However, lack of ability to detect missed dose through determination of drug concentration by HPLC, for drugs with a long half-life, such as nevirapine is one of the limitations of our study. Failure to detect missed doses would underestimate non adherence to ART in some patients. In addition, only a single blood sample was taken, and the time of last dosing was dependent on the authenticity of the caretaker’s reporting, which may not be guaranteed. It is possible that some of the caretakers or older children damped medication due to reasons previously reported [7
] so as to create the impression of good adherence on the day of refill. Damping of pills to create false positive adherence has previously been reported [21
In this study, non-adherence by nevirapine plasma concentration was found to be 15%, which is alarming when considering the management of HIV/AIDS in children.
However, it is important to note that, blood drug concentration can be affected not only by the level of adherence but also drug bioavailability, individual’s metabolic capacity, timing of drug intake and the time the blood sample for determination of drug concentration is taken as well as having diarrhoea or vomiting immediately after drug administration. In our study, no patient was reported to have vomited or experienced diarrhoea after drug intake. It is unlikely that the low plasma concentrations observed in some patients have been due to poor nevirapine bioavailability since, in all patients, we used same paediatric fixed-dose combination tablets obtained from WHO prequalified manufacturers. In this study, all patients who were using other drugs, which could interact with nevirapine, such as ketoconazole and rifampicin were excluded precluding the possible influence of enzyme inhibition or induction on the observed plasma concentration. However, in this study, the patients were not genotyped with respect to CYP 3A and 2B6 to rule out the role of indivudual’s metabolic capacity that could influence the plasma concentrations of nevirapine. A recent Pharmacogenetics study conducted in Tanzanian adults indicated the frequencies of CYP2B6 genotypes to be 37%, 45% and 16% for homozygous wild type (fast metabolizers), heterozygous (intermediate) and homozygous mutated (poor metabolizers) repectively [22
]. This implies that the obtained nevirapine plasma concentrations in some of our patients could have been influenced by the patient’s 2B6 activity. It is possible that observed high or low nevirapine concentrations were due to delay of blood sampling or drug intake, we recorded the last time of drug intake prior blood sampling and the time of blood sampling. We noted that the mean sampling interval was 4
2 hours post intake of the last dose, which did not significantly deviate from the study protocol.
Despite the above limitations, our findings support the use of drug plasma concentration as an objective method for determining adherence to ART where feasible. This would complement the information obtained from care taker report and medication returns particularly in groups of patients showing poor treatment response. Determination of plasma drug concentration to estimate adherence is often not feasible but nevirapine plasma concentration may be useful and cost-effective in the setting of a pharmacovigilance program with adherence monitoring where this may be limited to patients at sentinel sites and at regular intervals. Apart from estimating adherence, nevirapine plasma concentration may be used in the analysis of non-response that may lead to detection of early signs of drug resistance particularly in patients with well characterised genotypes of important metabolizing enzymes.