Abstract
It is imperative to prove efficacy of tailored interventions and translate the efficacious ones into clinical strategies for achieving good ART adherence. ART adherence among registered HIV/AIDS cases at HIV treatment centre, Pakistan Institute of Medical Sciences, Islamabad was assessed through RCT. Study duration was 10 weeks; eligible subjects (N = 76) were randomly halved; Intervention Group (IG) received trial interventions i.e. subject involvement, weekly phone reminders in addition to routine counselling, while Comparison group received routine counselling only. Self-reported adherence (SRA) questionnaire and pill identification test (PIT) conducted at both baseline and follow-up in addition to CD4 count and viral load. ITT using ANOVA; McNemar’s test for variables with before-after assessments within a group. Results showed significant differences in ≥95% SRA, ≥95% Adherence on PIT, Viral load test of <50 copies per cubic mm. These interventions should be included in the overall treatment strategy for HIV/AIDS in Pakistan.
Keywords
Background
Since the launch of World Health Organization (WHO)’s “3 by 5” initiative in 2003, several countries have established antiretroviral therapy (ART) programmes worldwide. The unachieved WHO target was overtaken by G8 Gleneagles Summit commitment of “as close as possible to universal access to treatment for all those who need it by 2010” during mid 2005. 1 Being a major advent of the modern science, highly active antiretroviral therapy (HAART—commonly stated simply as antiretroviral therapy, ie, ART) has radically improved the clinical status of many patients with HIV infection by lowering viral load along with increase in CD4 counts and reducing HIV-related morbidity and mortality. The commonly prescribed ART regimens include 3 or more medications, most with 2 or more daily doses. 2 The large number of medications involved, the complicated dosing requirements, and the suboptimal tolerability make adherence difficult. 1 However, effectiveness of ART strongly depends on the patient’s ability to adhere closely to the complicated regimens.
Although considerable efforts had been dedicated to study various measures of ART adherence and to establish correlation between ART adherence and virologic outcomes, evidence on interventions to improve ART adherence among HIV/AIDS cases is very limited. In the region of South Asia, scientific studies in the domain of treatment, care, and support of HIV/AIDS cases are almost nonexistent. Essentially, the implications of poor adherence are substantial at both the individual and population levels. For the patient, the benefits of treatment are reduced, leading to undertreatment of their condition and limiting their therapeutic options. At the population level, nonadherence results in medication wastage as well as increase in health care costs and drug resistance in the case of incompletely treated infections, which may also lead to transmission of drug-resistant strains to other persons during high-risk activity.
WHO formulated a model to explain adherence to ART as interplay of 5 major factors including patient-related factors, therapy-related factors, socioeconomic factors, condition-related factors, and health care system−related factors. 3 Therefore, attempts to improve adherence need to be multifaceted, particularly a combination of 2 or more specific interventions. The treatment, care, and support programme for HIV/AIDS in Pakistan has been operational since 2003 and average ART adherence has been reported as 83% among HIV/AIDS cases in the report of Treatment Adherence Study: Self Reported versus Measured Adherence to Antiretroviral Therapy—2007 (nonpeer reviewed). 4 It was, therefore, empirical to study efficacy of specific interventions to improve adherence among HIV/AIDS cases in Pakistan for enhancing their quality of life and health status.
The current study conceptually addresses therapy-related factors and health care system−related factors through participant involvement by discussing issues around ART adherence, social support, reasons of missing dose, and their active complaints. Patient-related factors are being dealt through weekly phone call reminders that also indirectly enhance the provide-patient relationship targeting health care system−related factors as well. The study aimed to improve quality of life of HIV/AIDS cases in Pakistan. The specific objective of the study was to assess the efficacy of interventions for improving adherence to ART regimens in patients with HIV/AIDS in treatment centre at PIMS, Islamabad. Findings of this study will be helpful in modifying the service package for treatment & care of HIV/AIDS in Pakistan.
Materials and Methods
The present study was a controlled, 2-arm randomized trial including a 02-week baseline phase to assess the participant’s current adherence and health indices, a 04-week intervention phase in which participants will be randomized to receive or not receive the interventions for skill building for improved adherence, and a 02-week follow-up phase to measure the outcomes. Data entry, statistical analysis, and report writing was done in the last 02-weeks. The total duration of this study was 10 weeks (15 April-30 June 2008).
WHO Conceptual Framework on Adherence to ART (Figure 1)
Broadly, the various factors affecting intention to adhere and ultimately ART adherence can be summarized as Socioeconomic related factors, health provider/system−related factors, condition-related factors, therapy-related factors, and patient-related factors. Illustration of interplay of these various factors in modifying the ART adherence can provide a firm foundation for effective interventions to improve adherence. Evidence shows that interventions that address only one variable affecting adherence are less likely to succeed than are those that target multiple factors.
3
In this trial, the investigator is assessing efficacy of interventions in improving the medication adherence ≥95% among HIV/AIDS cases at PIMS, Islamabad. The trail interventions include the following: Patient involvement for cue-dosing, which will target the patient-provider relationship as well as the therapy-related factors. Phone calls and mobile reminder as memory aids to target the patient-related factors.
WHO Conceptual Framework on Adherence to ART
Study Population
The study universe comprised of the HIV/AIDS cases registered at HIV treatment center, Pakistan Institute of Medical Sciences (PIMS), Islamabad, since November 2005 to date. A sampling frame was prepared from the study universe including the medical record numbers of only those registered HIV/AIDS participants at HIV treatment center, PIMS, Islamabad, who were on first-line ART regimens for at least the last 3 months. The required sample size for study was calculated using the statistical formula to demonstrate a significant difference between 2 groups by comparing proportion. 5 At a 2-sided α value of .05 (1.96) and confidence interval of 95%, allowing the study power to be 80% (0.84) and to detect a difference of 30% improvement in ART adherence, the calculated sample size was 38 participants in each group. Thus the total sample size was kept about 76 participants. The sampling was done through simple random sampling technique using random number tables. A pencil was dropped on the page of random number table with closed eyes and the number right beneath the tip of pencil was taken as the starting point. The direction of movement along random number table was from left to right and later on when the table finished, top-to-down direction was used. Three columns of random numbers were taken for random selection of initial 38 medical record numbers into intervention group and the later 38 medical record numbers into the comparison group. Thus, the total sample size of 76 medical record numbers was randomly selected from a list of 88 medical record numbers of HIV/AIDS cases.
The sample selection was done based on the criteria of being registered at the treatment center in PIMS, Islamabad, and prescribed ART regimen for at least 03 months; aged ≥18 years; either sex; currently being treated with first-line ART; having a telephone contact (PTCL or mobile). The exclusion criteria included inability to give informed written consent; being admitted in a health facility where HIV medication is dispensed; pregnant and/or lactating female patients; participants participating in any other research.
Trial Interventions
The design of trial interventions was guided by the treatment protocols at HIV treatment center, PIMS, Islamabad. The HIV/AIDS participants were referred to the treatment center from other health care providers/facilities as well as local nongovernmental organizations (NGOs). The participants were registered at the center and were provided with medical consultation, baseline investigations, pretest and posttest counseling, HIV ELISA testing, CD4 counts, and HIV viral load testing at the center. To proceed with the trial interventions, initially a list of eligible participants was prepared with help of the Medical Officer at HIV treatment Center, PIMS. The total number of eligible participants fulfilling the selection criteria was found to be 88. Using a computer-generated list of random numbers, 76 of the 88 eligible participants were randomly assigned into 2 equal-sized groups. The intervention group (IG) was to receive the trial intervention (participant involvement and phone call reminders) and the other group was to receive the comparison intervention (routine counseling only) as the control group (CG).
The trial intervention included participant involvement in selecting dosing schedules that fit into their daily lives and discussing side-effects along with once weekly phone call reminders during the day timings (1000 hours to 1400 hours) to the patient by the research assistant as memory aids. During the phone call reminder, the participants were reminded of their interaction with the researcher at the clinic and were requested to strictly follow the ART dose schedule to avoid missing any dose.
For the purpose of participant involvement during counseling sessions, the participants were asked to generate a personalized time-of-day schedule for taking each prescribed antiretroviral (ARV) medication. The participants were also guided to associate medication doses with their daily routine activities performed at the times that the medication should be taken, which serve as cues for taking medication. For example, morning doses to be associated with morning rituals (like offering prayers, brushing teeth, or reading the newspaper) and evening doses to be associated with evening routines (like evening tea, children’s homework, or watching television news programmes).
The 3 phases of the trial were as given below: Baseline phase: During the first 2 to 3 weeks, the researcher obtained initial information like name, address, and telephone contact numbers of the study participants in both groups from the clinical records at HIV treatment center, PIMS, Islamabad. In addition, the dates of next appointments of the study participants in both groups were also recorded. Those participants whose appointments were due during following 2 to 3 weeks time were to be enrolled in the study at the time of their visit at the HIV treatment center after taking informed verbal consent. However, telephone calls were made by the researcher to those participants (both IG and CG) whose next appointments were after more than 3 weeks time. Verbal informed consent to participate in the study was taken telephonically and then they were asked to visit the clinic during the following 2 to 3 weeks for baseline assessment of the sociodemographic, behavioral, and clinical characteristics of the study participants using a structured questionnaire.
All the study participants visiting the HIV treatment center at PIMS, Islamabad, received the routine adherence counseling sessions as per the clinical protocol without any other intervention for the 02-week baseline period. Intervention Phase: During the next 4 weeks time, the researcher applied trial intervention on the IG while CG was receiving routine counseling only. To implement the weekly phone call reminders for IG, a separate mobile connection was obtained by the researcher for convenience of contacting the participants at scheduled times and also maintaining confidentiality of enrolled participants. There was a telephone line available in the OPD room of HIV treatment center, PIMS, Islamabad, however, it was not practical to use the same connection for the research study.
Although the CG received only routine counseling sessions, the IG were given counseling sessions by the researcher in which the trial intervention was ensured including formation of personalized schedule for ART doses and also cue dosing. Follow-up phase: During the 02-week follow-up phase, the interventions were discontinued and participants from both IG and CG were assessed by the researcher for % self-reported adherence (SRA)
7
using self-report questionnaire for follow-up visits and pill identification test (PIT)
8
score. Those participants who did not come for their follow-up visit were assessed for % SRA telephonically using the same questionnaire. However, their PIT score could not be recorded (Figure 2
).
The Trial Profile
Data Collection Procedure
Measurement of ART adherence was done using a combination of subjective and objective methods. The subjective methods included SRA 6 and PIT. 7 The objective measures of adherence included CD4 counts and viral load test. 8
For SRA, a precoded-structured questionnaire was used adapted from the questionnaire on medication adherence and side effects for baselines and follow-up assessments by Adult AIDS Clinical Trial Group (AACTG)– Center for AIDS Prevention Studies, University of California, San Francisco (CAPS Tool).
There were 3 main parts of the questionnaire. The first part was to obtain information on the baseline sociodemographic, behavioral, and clinical characteristics of study participants. The sociodemographic variables included name, medical record number, gender, age at last birthday, education, occupation, number of children, if any, and place of residence. In addition, the information on addictive behavior was collected by using variables like smoking, alcohol, and drug dependence. Clinical characteristics at baseline included variables like duration of ART regimen and medically reported adverse effect of ART.
The second and third parts of questionnaire were for assessment of SRA at baseline and follow-up phases, respectively. These were adapted from the questionnaire on medication adherence and side effects for baseline and follow-up assessments adapted from AACTG-CAPS Tool. 6 Other variables at baseline assessment included knowledge and belief of participants about their ART medications, their social and family support in taking ART medications, the possible reasons for missing dose(s) if any, the existence of anxiety and hopelessness along with their coping abilities, and their self-reported symptoms. At follow-up assessment, variables on possible reasons for missing dose(s) and self-reported symptoms along with some specific questions about how many doses were missed and when. Also variables like adherence to timing schedule of ART doses, compliance with any special instructions, and missing doses over weekends are included in the follow-up questionnaire.
ART Adherence Measurement at Baseline
The baseline value of percentage adherence was assessed through participant’s self-report on ART adherence 6 and PITest. 7
Participant’s self-report on adherence (SRA)
The baseline questionnaire was used for assessment of baseline and follow-up %SRA of the study participants. Using the data from specific questions about number of doses missed during last 4, 7, 15, 28, 90, and >90 days, %SRA was calculated for 4, 7, 15, and 28 days by dividing the number of doses missed in a period by the number of prescribed doses for that period and multiplying by 100 then subtracting this figure from 100. However, if any dose was missed during last 90 days, it was taken as 95% SRA and if any dose was missed during last >90 days, it was taken as 97% SRA. 6 To increase the validity of SRA, a preamble was used before asking adherence questions to reassure patients that the information would not be held against them and that it was not unusual to have problems with adherence.
Pill identification test (PIT)
All study participants were on first-line ART for at least 3 months. They were asked to identify the ARV pills they had been prescribed on a board containing 6 ARV pills with 1 similar-appearing pill (referred to as twin pill) for each ARV pill. The PIT score was calculated as sum of misidentifications weighted according to the degree of resemblance of the pills (0.5 for the ART pill and 1 for omission). The participant’s knowledge of his/her ART was considered satisfactory if PIT score was 0.5 and it was taken as 95% adherence. A score of 1 was taken equivalent to <95% adherence. 7
The data collection tool in English were translated into Urdu and pretested in the Prevention from Parent to Child Transmission (PPTCT) Clinic at PIMS, prior to the baseline phase, and modified accordingly before application to the study sample.
For the objective adherence measurement, there were 2 important laboratory determinants of the rate of progression of HIV infection as well as the response to ART, CD4 count and the plasma viral load tests.
9
CD4 counts: According to the treatment protocol at the HIV treatment center at PIMS, Islamabad, the CD4 counts of the HIV/AIDS cases were done before deciding to initiate the ART using the FACS Calibur Flowcytometer by Becton Dikinson. Then the CD4 counts were only repeated on annual basis since the average rate of decline of CD4 cells (CD4 slope) is about 50/mm
3
per year.
10
Therefore, for the current study, the baseline CD4 counts of all the participants included in the study were taken from the clinical records for the last 1 year. Moreover, since the HIV treatment center was out-of-stock of the CD4 kits, it was not possible to carry out any follow-up CD4 testing of the study participants. Therefore, none of the study participants was asked for a fresh CD4 test to be done for the purpose of the study. Viral load (VL): Similarly, based on the treatment protocol at HIV treatment center, PIMS, Islamabad, the viral load test of HIV/AIDS cases is done before deciding when to initiate the ART and later on repeated on a biannual basis to monitor the response to treatment using Ampliprep and Amplicor RNA-PCR Assay by Roche. Therefore, the viral load results were collected from clinical records for the last 6 months and those test values were taken as the baseline for them. However, the participants whose viral load test had not been done during the previous 6 months time were asked to get a fresh viral load test as the baseline. Similarly, the participants whose viral load test was due were asked to get their tests done for the follow-up assessment.
Ethical Considerations
An informed verbal consent form was translated into Urdu and it was read out to the eligible participants and their confidentiality was reassured. While making the telephone calls during the baseline, intervention, and follow-up phases, it was strictly observed that the addressee over telephone is the participant himself/herself and no other person attending the phone call is conveyed any information regarding HIV status of the participant. Moreover, for the sake of participant’s satisfaction and to sever any possible apprehensions being aroused by telephone calls, the researcher refrained from making the phone calls and one of the medical staff at the HIV treatment center, PIMS, Islamabad, was recruited for making the phone calls at all the 3 phases of the study.
The researcher also required a letter of facilitation from the Director, Health Services Academy, Islamabad, to assist in the field work at HIV treatment center, PIMS, Islamabad. The research proposal was approved by the Ethical Review Board of Health Services Academy, Islamabad. There was no conflict of interest.
Statistical Analysis
The data was entered and analyzed in Statistical Package for Social Sciences version 15. The data was analyzed on an intention-to-treat basis, to provide a realistic indication of effectiveness of the intervention. Analysis of variance (ANOVA) test of significance was used to detect differences within and between the intervention group (IG) and comparison group (CG) group at baseline assessments. McNemar’s test for paired data was used to test the significance of differences in key outcome variables at the follow-up phase. P value of less than .05 was considered as statistically significant.
Study Results
A total of 76 eligible and interested participants were enrolled and randomized into the 2 groups, that is, intervention group (IG) and control group (CG). During baseline phase of the study, 27 (71.0%) participants from IG and 33 (86.8%) participants from CG were interviewed by the researcher in person at the HIV treatment center, PIMS, Islamabad. The remaining 11 (28.9%) participants from IG and 5 (13.1%) participants from CG were interviewed telephonically. During the follow-up phase of the study, 4 (10.5%) participants from IG and 2 (5.2%) from CG were contacted on telephone for the follow-up questionnaire.
Thirty-four (89.5%) participants each from IG and CG responded both to the baseline and follow-up questionnaires. Two (5.2%) participants from IG and 4 (10.5%) from CG did not respond to the follow-up questionnaire as they could not be contacted on telephone numbers given to the researcher. Furthermore, in the IG, one (2.6%) participant died and another (2.6%) went to prison. Therefore, a total of 4 (10.5%) participants from IG did not receive the trial interventions during the intervention phase of the study. The proportion of nonrespondents at follow-up phase was thus similar in both the groups.
Characteristics of the participants in the 2 groups are detailed in Table 1 . The 2 groups, that is intervention group and comparison group, were comparable in terms of the sociodemographic characteristics. There were no significant differences between the groups in any of the baseline measures including gender (P value .304), age at last birthday (P value .985), education status (P value .528), occupation (P value .066), marital status (P value .162), having children (P value .580), and place of residence (P value .806).
Sociodemographic Characteristics by Group Assignment at Baseline Assessment
Abbreviations: CG, comparison group; CI, confidence interval; IG, intervention group.
a P values calculated by using ANOVA.
The majority of participants in both groups were males (68% in IG and 79% in CG) having age ranging from 35 to 44 years (42% in both groups). About one third of participants in both the groups were illiterate, however, 15 (39%) participants of IG and 18 (47%) participants of CG were having education status of secondary or above. Almost all the females in both groups were housewives, whereas 18% participants of IG and 32% participants of CG were unemployed. The marital status of about 76% participants of IG and 66% participants of CG was married and living together and majority were also having children. Table 2 shows the comparability of both groups in terms of their self-reported addictive behaviors, their knowledge and belief about ART, and their social/family support.
Self-Reported Addictive Behaviors, Knowledge, and Belief About ART and Social/Family Support by Group Assignment at Baseline Assessment
Abbreviations: ART, antiretroviral therapy; IG, intervention group; CG, comparison group; CI, confidence interval; DU, drug user; IDU, injecting drug user.
a P values calculated by using analysis of variance (ANOVA).
The mean duration of ART in IG was 14.8 months (SD = 8.07) and that of CG was 19.2 months (SD = 12.60), and the difference was not statistically significant (P value = .611). The proportion of patients who had medically reported adverse effects from ART was similar in both groups (P value = .461). All the comparisons were statistically tested using ANOVA test of significance. All participants were on regimens that contained at least 1 drug that was dosed twice a day (data not given).
The mode of HIV acquisition of participants in the 2 groups was also comparable and almost similar proportions of participants from IG and CG reported HIV acquisition from HIV-positive man (32.4% vs 31.6%, respectively) and HIV-positive woman (56.8% vs 52.6%, respectively). These differences were not found to be statistically different among the 2 groups (P value = .390) using ANOVA test of significance. The self-reported modes of HIV acquisition in both the groups are summarized in the Figure 3.
Self-reported modes of HIV acquisition among intervention group (IG) vs control group (CG)
Regarding their feelings of anxiety, hopelessness, and their ability to cope with depressive moods, participants in both IG and CG were comparable since the differences were found to be not statistically significant (P values > .05 for all relevant variables). The aggregated results are shown in Figure 4.
Self-reported status of anxiety, depression, and coping abilities by group assignment at baseline assessment
Key Variables at Baseline Assessment
At baseline assessment, the results of key variables including ≥95% SRA, PIT, viral load test, and CD4 count of ≥250 cells were also found to be comparable. The P values were calculated using ANOVA test of significance, and they were statistically insignificant ranging from .231 to .764 (Table 3 ).
Results of Baseline Assessment of Key Variables by Group Assignment
Abbreviations: IG, intervention group; CG, comparison group; CI, confidence interval; SRA, self reported adherence.
a P values calculated by using analysis of variance (ANOVA).
Key Variables at Follow-up Assessment
A total of 68 participants (34 participants each from IG and CG) of the initially recruited 76 participants were available for the follow-up assessments. Moreover, only the results of SRA were available for those participants in both groups who were contacted telephonically for the follow-up assessment including 4 (10.5%) participants from IG and 2 (5.2%) participants from CG. Although, none of the participants had undertaken CD4 count test at follow-up assessment, the viral load tests were carried out for 23 (67.6%) participants from IG and 19 (50%) participants from CG. The differences between the 2 groups in terms of ≥95% SRA, ≥95% adherence on PIT, and viral load test of <50 copies/mL were all found to be statistically significant with P values ranging from .023 to .000 using McNemars test of significance for paired data. A summarized account of the primary outcome measures at follow-up assessment is given in Table 4 .
Results of Primary Outcome Variables at Follow-up Phase by Group Assignment
Abbreviations: IG, intervention group; CG, comparison group; CI, confidence interval; SRA, self reported adherence.
a P values were calculated using McNemar test of significance for paired data.
The follow-up assessment further showed that adherence to the timing schedule of ART doses was better among IG than CG and lesser proportion of IG missed any dose over weekend than CG; however, the differences were not statistically significant. Moreover, the compliance with any special instructions was higher among the IG (P value .052; Table 5 ).
Specific Questions on ART Adherence a
Abbreviations: ART, antiretroviral therapy; CG, comparison group; CI, confidence interval; IG, intervention group.
a P values calculated by using analysis of variance (ANOVA).
Discussion
This study assessed whether receiving the trial interventions was associated with improved ART adherence in the patients with HIV/AIDS at HIV treatment center, PIMS, Islamabad, in a randomized, controlled design. Findings of the study have demonstrated the possible use of phone call reminders coupled with participant involvement for improving ART adherence and virologic outcomes. Although the possibility of overestimation of percentage adherence cannot be excluded using the subjective SRA measure, the simultaneous use of objective measures like PIT and HIV viral load tests support its validity.
Relative to the comparison group, the participants randomly assigned to the intervention group showed better ART adherence (≥95%), measured by SRA questionnaire and PIT. The findings conform to that of Pradier C et al 11 with similar tool for assessment of SRA. However, the duration of their trial was about 6 months, which would have allowed them to detect the difference with greater validity. Moreover, the meta-analytic review of 18 randomized controlled trials on adherence interventions reported significant aggregated effect size for 95% adherence (OR = 1.50, 95% CI = 1.16-1.94) indicating that, overall, the likelihood of achieving at least 95% adherence was higher in the intervention arm than in control arm.
The ultimate goal of intervening to improve adherence is to improve medical outcomes of treatment. This study was designed to test this possibility using the body weights in kilograms and the viral load test result. The differences in viral load and body weights were found to be statistically significant. These findings are supported by the ones showed by Sorenson et al 12 who found a positive change in viral load associated with medication adherence. In addition, the study by Sorenson et al has not found significant changes in the body weights of participants in intervention arm. However, the inference in the current study from these findings might probably be constrained by the short follow-up duration and the relatively low baseline viral loads as well as the body weights. Moreover, in many cases, maximum antiviral effect may have been achieved over the period since the ART regimens were prescribed. Future studies of adherence interventions should address the issues of efficacy of regimens used and the timing of the intervention as it relates to the initiation of therapy.
Rigsby and colleagues 2 have shown that interventions focusing on adherence to dose intervals rather than total daily dose lead to improved adherence and virologic outcomes. However, in the current study, there was a lack of significant differences between the 2 groups in variables like adherence to the timing schedule of ART doses, compliance with special instructions while taking ART medication, and the adherence behavior over weekends. The possible reason for this limitation would be a shorter intervention phase (06 weeks) and relatively smaller sample size; whereas a larger sample, a longer term intervention, and a multisite trial might uncover significant differences. Nonetheless, this kind of intervention can also be applied to address medication adherence issues among patients with diseases such as tuberculosis or hepatitis C, which involve time-limited medication regimens.
The sample size of current study was 76 (38 participants in each group), the sample size in other studies found in literature was in the range of 55, 2 24, 13 350 14 ; thus, the sample size of 76 participants was fairly reasonable for this study.
The retention rate in current study was about 90%, with no difference between the 2 arms. Similar findings were shown by the meta-analysis of randomized controlled trials on adherence interventions. The retention rates of studies were found to be ranging from 40% to 100% (median = 80%) and not differing significantly between the 2 groups for any study.
Explaining possible reasons for the higher effect size for >95% adherence than for the viral load <50 copies/mL, the measurement factors of both primary outcomes might be attributing to it. The viral load tests were obtained through blood samples drawn at baseline or follow-up phase, whereas the 95% adherence was based on the SRA estimate supplemented by the PIT. However, a meta-analysis indicates that studies using more objective assessments of adherence (eg, electronic medication monitoring and pill counts) have shown larger effect size for 95% adherence than the studies using SRA estimates. 15
Some limitations of this study have to be acknowledged. First, the follow-up period was limited to 06 weeks (45 days); sustainability of intervention effects over longer period beyond the time of the sessions needs to be confirmed. Another limitation of the study was the general methodological problem related to assessment of adherence based on participant’s self-report, which may be affected by social desirability and recall bias. Nevertheless, the instrument also contained a set of additional categorical questions that minimized the risk of overestimation. Moreover, there is evidence to confirm that SRA is reasonably reliable and correlates well with the virologic outcomes. 12
A third limitation was the nonavailability of CD4 test kits at the HIV treatment center, PIMS. Also, the clinical protocols along with the time duration of study did not allow viral load testing of all participants at the follow-up phase. Furthermore, the participants enrolled in this study were representative of the population served by the HIV treatment center, PIMS, Islamabad. However, the study group included relatively few men having sex with men. Although mode of acquisition of HIV is not reported as an important predictor of adherence, it is possible that interventions aimed at improving adherence will need to consider the cultural makeup of the target group. However, the controlled design of this study has guaranteed that the sociodemographic and behavioral characteristics were similar in both randomized groups.
Finally, quality assurance methods for the trial intervention are also important. Considering the ethical requirement of establishing contact with the HIV/AIDS cases registered at HIV treatment center, PIMS, Islamabad, a staff member of the center was recruited as research assistant to make the telephone calls so that the participants are not apprehended by phone calls from the researcher who is not a regular health care provider at the center. Although considerable efforts were made to deliver standardized intervention through use of written scripts, some variability due to the staff that made the weekly telephone calls cannot be overlooked.
Last but not the least, a lesson learnt was that any research with people living with HIV (PLWH) should not overlook the importance of linkages with community-based organizations working for PLWH to obtain accessibility to and acceptability from the target group. These organizations generally have the mandate to provide support groups, peer-counseling, and other assistance in socioeconomic as well as health-related issues, therefore, the PLWH have an expected tendency to route their contacts with health care providers and more so with researchers through them.
Conclusion
Although, the generalizability of findings of this study may need further confirmation, the findings present evidence in favor of the feasibility and efficacy of participant involvement and phone call reminders to increase ART adherence that could be easily implemented with limited additional resources in most of the clinical settings providing treatment, care, and support for HIV/AIDS in Pakistan. However, meeting the challenge of translating interventions that are efficacious in research trials into effective clinic-based strategies that can also be used in resource-poor areas require an ongoing operational research agenda.
The study findings provide basis for adapting the trial interventions to achieve ≥95% ART adherence and gain the possible clinical as well as public health advantages for prevention and control of HIV/AIDS in the country. Therefore, these interventions should be integrated into the current protocols of HIV/AIDS treatment, care, and support programme under the National AIDS Control Programme, Ministry of Health, Islamabad.
Footnotes
This article is based on an original work by the authors and is available for distribution and reproduction in any medium, provided proper citation is ensured.
Acknowledgements
I acknowledge the support of all the staff at the HIV treatment Center, PIMS Hospital, Islamabad, Pakistan and particularly the patience and interest of the registered HIV patients to participate in the study. I also acknowledge the logistic support from Health Services Academy (HSA) and the National AIDS Control Programme (NACP), Ministry of Health, Pakistan.
The authors declared no conflicts of interest with respect to the authorship and/or publication of this article.
The authors received no financial support for the research and/or authorship of this article.