Introduction 1 2 Since 2002 the nation has embarked on the provision of antiretroviral drugs for all its eligible citizens by implementing the MASA program.(The national antiretroviral therapy program was given the name MASA, the Setswana world for “dawn”). 3 4 5 7 8 9 The required high level of antiretroviral drug adherence in a poor resource setting remains therefore a serious concern. Assessment of adherence in HIV patients such as in this pilot-study may also provide tools to allow feedback and education on an individual health care provider–patient base. For this reason patients in the Botswana Infectious Disease Care Centers (IDCC) treatment program are urged at each visit to the IDCC facility to comply with the prescribed ART regimen. This occurs in three stages: (1) in group instruction sessions together with other HIV patients and individual counselling by a trained nurse or pharmacist, (2) by their individual health care provider (physician), and (3) by the pharmacist. In addition, patients are usually accompanied by a close family member who is asked to assist or remind patients of the pill intake (adherence partner). 10 11 12 10 13 14 11 15 21 This study was designed as a pilot study to evaluate electronic adherence monitoring in an HIV infected patient group that was put on antiretroviral medication for the first time. A secondary objective was to compare the adherence measured by electronic monitoring with that of self-reporting by means of a medication diary. Methods Patients 3 2 Design This was a trial in which treatment-naïve patients were monitored with regard to adherence to prescribed anti-HIV medication. Patients were not informed about the use, blinded of the electronic monitoring system, and were only asked to return their pill-bottles at each visit to the pharmacy when they returned for a refill and a consultation. Patients were also supplied with a self-reporting form. The study did not involve study related interventions and the subjects were not required to change behaviour in any way. The subjects were informed that the medication was supplied in special containers that had to be returned to the clinic but not about the monitoring system to prevent bias. The study was approved by the hospital management and the chief physician of the department of internal medicine. Treatment regimens The treatment regimen used consisted of 2-nucleoside reverse transcriptase inhibitors (NRTIs) plus 1 nonnucleoside reverse transcriptase inhibitor (NNRTI). Patients started the recommended first-line treatment with three different agents. These were zidovudine and lamivudine in a combination tablet Combivir (CBV) plus efavirenz (EFV) or nevirapine (NVP). Male and female patients who were not anaemic were prescribed Combivir medication whereas anaemic patients (Hb < 7.5 mmol) were given stavudine (d4T/Zerit) and lamivudine medication. The stavudine dose was adjusted according to bodyweight (30 or 40 mg). All patients except females in the reproductive age category were given once daily EFV. All females of the former category were prescribed NVP at a 200 mg once daily dose for 14 days which, after assessing the liver function parameters, was increased to 200 mg twice daily. The continuation of NVP or EFV was dependent on the absence of significant rise in hepatic enzymes (AST and ALT). Approximately 90% of those who started the NVP or EFV treatment are able to continue this medication. Patients returned to the IDCC after a month for a medical check and refill of their prescription unless clinical events dictated earlier visits to the clinic. The treatment starters were booked to see the doctor after the first 2 weeks of therapy. After seeing the doctor, the self recorded medication card and the electronic monitors were collected, and the data stored in the microprocessor were transferred to a database in the computer. Following this, each bottle was refilled and provided with a new label with medication instructions. Most patients received a refill for a period of 1 month, some however for a shorter period. The potential side effects were discussed with each patient. The results of the first analysis of the electronic monitors were not used in any counselling. The electronic monitors were MEMS IV Track Cap devices (Aardex, Zug, Switzerland) with a MEMS IV Communicator for read-out of the results. Study endpoints The primary study endpoint was the adherence level measured (over a minimal period of 6 weeks) by the percentage of days on which the patients took a correct dosing over the monitored period. Adherence was also expressed as the number of pill-intakes recorded on a self-reporting forms designed to reflect the intended schedule and timing of treatment. In the event a patient opened his/her bottle more than was prescribed (surplus opening), it was assumed that the patient correctly took the prescribed pills. However, occasions on which a patient opened less than the prescribed dose frequency were considered as adherence failures. Sequence and duration of trial period Each patient was immediately given counselling and made familiar with the ART treatment. During the counselling session emphasis was given with regard to the need of strict adherence of the prescribed medication and to methods to prevent disease transmission. They were also informed about the self-reporting form and given a pen to mark taking a treatment with a cross. ART medication was started thereafter and the adherence to the pill intake schedule was monitored by means of using electronic monitors and self-recording for a period of 6 weeks. At the start of the treatment, electronic monitors containing medication for a period of 2 weeks were provided. After an evaluation by the doctor at day 14 of treatment the electronic monitors and self-reporting form were collected and the data in the microprocessor were entered in the database. The electronic monitors were subsequently refilled with medication for the next period of 1 month. A new self-reporting form was also given. Patients were given 2 (in case of Lamivudine/Zidovudine + NVP or EVF) or 3 electronic monitors (in case D4T, 3TC, and NVP or EFV). Patients recruited in the study were asked to return the electronic caps and the self-reporting form on each occasion of a visit to the IDCC. A self-reporting form was issued at the start of the study. This form contains rows where the patient had to mark with x each time they took the pill at the correct time. As some people in Botswana could not read or write, this form was kept very basic. A pencil was given to every patient who participated in the study. Results A total of 30 HIV infected adults were enrolled in the study. In five patients full data could not be obtained because of various reasons. This leaves an evaluable group of 25 (9 male, 16 female; average age 35.6 years, range 22–55 years). Twenty patients completed the 6-week monitoring period and the mean follow-up period was 49 days (range 42–72 days). 1 Table 1 n Patient number TC Follow-up (days) MEMS Self report Lamivudine /zidovudine (bd) EFV (od) Lamivudine (bd) NVP (od/bd) d4T (bd) Drop out/non-retrievel 1 1 11 100 100 100 100    3 2 2 70 95 100 91   100   3 2 14 100 No data 100   100  3 4 1 51 98 100 98 100     5 4 − − − − − − − − 1 6 1 44 71 100 71 95     7 3 45 73 70   92 81 96  8 1 44 100 100 100 100     9 2 44 98 100 100   98   10 2 14 100 100 100   100  3 11 1 44 88 100 88 98     12 2 43 93 100 93   93   13 1 44 98 89 98 100     14 1 − − − − − − − − 2 15 2 44 93 100 98   93   16 2 15 62 100 100   62  3 17 2 − − − − − − − − 2–1 18 2 44 84 100 84   86   19 2 45 98 100 98   100   20 1 9 100 No data 100 100    3 21 1 − − − − − − − − 1 22 1 72 75 100 75 94     23 2 44 21 100 21   29   24 2 42 93 100 93   95   25 2 66 97 100 100   97   26 1 43 100 100 100 100     27 1 − − − − − − − − 4 28 1 49 94 No data 96 96     29 1 56  100 100      30 1 47 20 100 20 100     avg  42 85 98 89 98 92 87 96  Treatment codes (TC) are: 1 Lamivudine/zidovudine, Efavirenz (EFV); 2 Lamivudine/Zidovudine, Nevirapine (NVP); 3 Lamivudine, Nevirapine and stavudine (d4T); and 4 Lamivudine and Efavirenz. MEMS indicates the data combined for all different treatments. Patient numbers are not consecutive because patients who died have been omitted. Reasons for drop out and nonretrieval: 1 death; 2 MEMS thrown away; 3 missed by investigator (patients showed up on another date than the investigator expected); and 4 lost to follow-up surplus 10 1 Fig. 1 Examples of data of from patients with good and respectively poor adherence  Adherence assessed by means of self-reporting p Discussion In this pilot study we assessed the use of MEMS monitors to study the adherence to antiretroviral medication prescribed for HIV patients living in a low resource health care system. We demonstrated that assessment of adherence with this technique is feasible and may provide useful results. There was an approximately 30% drop-out rate of the recruited patients due to inability to recover data or early mortality. This may seem unacceptable in a well-resourced health care system. It is a reality in many countries where patients present with much more advanced disease, when there are sometimes great difficulties coming to the hospital, and patients can often not be reached by telephone or mail as they do not regularly have a postal address. It is likely that some association exists between failure to return to the hospital and adherence to the drug regimen and the current patient set therefore may reflect an overestimation of adherence. 22 23 24 25 26 27 28 12 28 29 The outcome of monitoring pill intake by the electronic monitors may therefore assist in timely counseling of patients with regard to their medication intake and persistence with the prescribed regimen. Adherence measurements by means of using electronic monitors together with self and adherence partner recording of pill intake are likely to be useful in a much needed larger, long-term and more comprehensive adherence study in a low-resource health-care-system setting. Such a study can contribute to directed efforts to optimise the treatment of HIV-infected patients worldwide.