Self-reported adherence to HAART in South-Eastern Nigeria is related to patients' use of pill box

south america

13 pages

Please download to get full document.

View again

of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
The aim of this study was to assess levels of adherence and predictors of adherence to HAART in South-Eastern Nigeria. Selfreported adherence to HAART was assessed at 4-week intervals for a period of 3 months. A 10-item questionnaire was used to
    International Journal of Public Health and Epidemiology ISSN: 2326-7291 Vol. 3 (3), pp. 017-025, March, 2014. Available online at  © International Scholars Journals Full Length Research Paper    Self-reported adherence to antiretroviral therapy in sub-Saharan Africa: A meta-analysis   * 1 Charles E. Okafor and  1 Obinna I. Ekwunife * 1 Department of Pharmacy, National Orthopedic Hospital, Enugu. PMB 01294, Nigeria. 1 Mopheth Pharmacy, Lagos Nigeria. Accepted 02 October, 2013 As treatment of HIV infection with antiretroviral medications becomes a reality in sub-Saharan Africa, adherence to treatment regimen becomes a challenge. A meta-analysis was conducted to summarize the reported adherence rate in sub-Saharan Africa. Forest plot was used to visualize the extent of heterogeneity among studies. Following the random effect model, the combined adherence percent was 84.31% (95% CI = 79.48% - 88.60%). The Monte Carlo sensitivity analysis provided an alternative statistical method to evaluate pooled proportion and the analysis was similar to the random effect analysis. Identified barriers to adherence include: depression, centralized ART clinic, interruption in drug supply/procurement, stigma, absence of social support, cost of ART, complacency, forgetfulness and medication related problems. Cost of ART (OR = 2.19; 95% CI= 1.65  –  2.90), Complacency (OR = 5.25; 95% CI = 2.89  –  10.80), and medication related problems (OR = 1.68; 95%CI = 1.28  –  2.22) were the strongest barriers to adherence. This study showed a good level of adherence in sub-Saharan Africa. However, barriers to adherence identified in this study could be employed to improve adherence to a near perfect level. Key words:  Adherence, antiretroviral, Sub-Saharan Africa, HIV/AIDS, self-reported, meta-analysis. INTRODUCTION HIV is a major public health problem for developing countries especially those in sub-Saharan Africa. Sub-Saharan Africa is more heavily affected by HIV and AIDS than any other region of the world (UNAIDS, 2013). At the end of 2011, estimate of 23.5 million (69%) people were living with HIV globally in this region with 1.2 million death recorded (UNAIDS, 2012/2013). Antiretroviral therapy has been shown to improve patients’ therapeutic outcome  in several studies (Chi and Cantrell, 2009). With the help of donor agencies, the number of patients receiving antiretroviral therapy in sub-Saharan Africa has increased. In Africa, the number of patients on antiretroviral therapy increased from less than 1 million in 2005 to 7.1 million in 2012 (UNAIDS, 2013). As treatment of HIV infection with antiretroviral medications becomes a reality in sub-Saharan Africa, adherence to treatment regimen also becomes a challenge  *Corresponding Authors. Email: that needs to be addressed. Although, adherence to ART in some sub-Saharan African countries is over 70%, opportunities remain for improvement to a near perfect level. Adherence is defined as taking medications or interven- tions correctly according to prescription (Reda and Biadgilign, 2012). There are different methods of assessing adherence which include: Medication event monitoring (MEM); Self-report; pill count; biologic markers; body fluid assay; viral load monitoring etc (Landovitz, 2011). Pill count, MEM and self-reported method are the most reliable methods (Mweemba et al., 2010). However, there is no gold standard method for measuring adherence. The self-reported method is easier to perform and is cost-effective compared to other methods. A major disadvantage of the self-reported method is that the patients could be bias in their reports. Some studies have shown that antiretroviral regimens require 70-90% adherence in order to effective (Nachega et al., 2010), while most studies have shown that ART requires ≥95% adherence in order to be effective   (Byakika-Tusiime et al., 2005a; Diabate et al., 2007; Weiser      Okafor & Ekwunife 017 et al., 2003). Understanding the pathogenesis of HIV has suggested that adherence to ART of at least 95% or greater is required to keep the viral load at undetectable levels for as long as possible to prevent drug resistance and to maintain the functionality of the immune system. Knowing the percentage of adherence and understanding the predictors of non-adherence are the initial steps in a bid to improve adherence to ART. A detailed understanding of the possible factors that can cause non-adherence will greatly aid in the development of interventions to improve adherence in this African setting. This study aimed to evaluate the percentage of adherence to antiretroviral therapy (ART) in sub-Saharan Africa from 2003-2012 using the self-reported method, to identify barriers to adherence and strategies to improve adherence. METHODS Study Objective PUBMED and Cochrane library were searched for similar meta-analysis to avoid replication in case the study has been conducted. The Problem-intervention-comparism-outcome (PICO) was used to formulate the research objective but in this analysis, there was no comparism. Only ‘problem -intervention- outcome’ were use d to formulate the research objectives. The objectives of the meta-analysis were: to estimate the percentage of adherence to ART in sub-Saharan Africa based on self-report; to determine and analyze predictors of non-adherence to ART; and to suggest strategies to improve adherence. Data Sources The search for primary studies was done by searching published literature using seven general databases and search engines which include PUBMED, MEDLINE, AJOL (African Journal Online), JAIDS (Journal of Acquired Immune Deficiency Syndrome), JIAPAC (Journal of the International Association of Physician in AIDS Care), IJSA (International Journal for STDs and AIDS), AIDSONLINE and Google scholar. The key words (self-reported, adherence, HIV, ART, sub-Saharan Africa, predictors of adherence, barriers of adherence) combined differently were used in the search. Full texts and abstracts of relevant studies were collected. Study Selection After identification of relevant studies, each study was assessed for eligibility. The authors searched all the abstracts and full texts independently against criteria for inclusion. Each article was analyzed individually based on percentage of adherence, objective of study, area of the study, predictors of non-adherence, method, result, implication and recommendation. Full texts or abstracts were retrieved for those that met the inclusion criteria. Studies were included if the study assessed self-reported adherence to ART; if the study was carried out in sub-Saharan Africa; if the study explored factors of medication adherence; If the study was carried out between 2003  –  2012; If the study dichotomized adherences at ≥  95%. For example: an inclusive study is one that dichotomized adherence at ≥ 95% and reports that 52 patients out of 105 patients adhere to their medication, which implies that 52 patients took at least 95% of their medications. Medication adherence assessed by pill count or MEM only was excluded except where self-reported adherence was inclusive in the study. Studies that dichotomized adherences at < 95% were excluded. Abstracts and full texts with implicit methodology were also excluded. Data Extraction Reported findings in the selected studies were extracted using the review’s extraction form. The extraction was done in duplicate. Information extracted from each study include the following: country of study, author(s), study type, sample size, adherence measure used, reasons for non-adherence, contributing factors to adherence, and percentage of adherence. Data Synthesis and Analysis Adherence estimates were presented in a table. A formal meta-analysis was conducted to summarize the reported adherence rates in the individual studies. The primary aim of the meta-analysis was to determine the overall proportion of patients in sub-Saharan Africa that adhere to their ART. Forest plot was used to visualize the extent of heterogeneity among studies. Since heterogeneity was expected, a measure of the degree of inconsistency (I 2 ) across studies was conducted using Cochrane Q, moment-based estimate of between studies variance and I 2  measure. The I 2  statistic was calculated as a measure of the proportion of the overall variation in adherence that was to between-study heterogeneity (Higins and Thomas, 2002). The forest plot shows the individual study proportion with Clopper-pearson confidence intervals (CIs) and the overall (combined) Dersimonian-Liard pooled estimate. Analysis was conducted using Stats Direct statistical software (Stats Direct Ltd, version 2.7.8). Sensitivity analysis was done using Monte Carlo Markov chain simu-    018 Int. J. Public Health Epidemiol. Figure 1.  Flow Chart of Study Selection Process. lation of variability (Briggs et al., 2006; Robert and  Casella, 2004). In the sensitivity analysis, the proportions were made probabilistic using beta-distribution. 1000 iterations were used in the simulation. Multivariate logistic regression model was used to analyze the predictors of non-adherence. A column was made for the total popula- Abstracts and Full Texts Identified n= 97 Studies Identified by Searching from Reference Lists n= 12 Excluded against Inclusion Criteria n= 30 Unable to Provide Information for Assessment n= 5 Articles Assessed for Eligibility n= 62 No Quantification of Adherence n= 24 Publications Meeting Inclusion Criteria n= 41 Excluded: n= 6 Result Implicit to Interpret n= 4 Outcomes by different methods not reported separately n= 2 Number of Studies included in the Review n= 35    Okafor & Ekwunife 019 Table 1.  Characteristics and Proportions of the Studies. Source Country Study Type Age (y) Sample Size Adherent Proportion Adherent (95% CI) Nachega et al, 2004 South Africa CS ≥18  66 58 0.88 (0.78-0.95) Musiime et al, 2011 Rwanda LS ≥18  389 354 0.91 (0.88-0.94) Iroha et al, 2010 Nigeria CS ≤18  212 183 0.86 (0.81-0.91) Diabate' et al, 2007 Ivory Coast LS ≥18  591 439 0.74 (0.71-0.78) Weiser et al, 2003 Botswana CS 15-49 109 59 0.54 (0.44-0.64) Okonji et al, 2012 Kenya LS ≥18  434 366 0.84 (0.81-0.88) Nabukeera-Barungi et al, 2007 Uganda CS 2-18 170 152 0.89 (0.84-0.94) Davies et al, 2008 South Africa LS 1-5 115 84 0.73 (0.64-0.81) Rougemont et al, 2009 Cameroun LS ≥18  238 178 0.75 (0.69-0.80) Chabikuli et al, 2010 Uganda CS ≥18  100 71 0.71 (0.61-0.80) Bajurniwe et al, 2009 Uganda LS ≥18  175 149 0.85 (0.79-0.90) Byakika-Tusiime et al, 2005a Uganda CS ≥18  304 207 0.68 (0.63-0.73) Muyingo et al, 2008 Uganda/Zimbabwe LS ≥18  2957 2785 0.94 (0.93-0.95) Senkomago et al, 2011 Uganda CS 18-65 140 140 1.00 (0.97-1.00) Watt et al, 2010 Tanzania CS ≥19  340 320 0.94 (0.91-0.96) Laurent et al, 2004 Cameroun ORCT ≥18  60 59 0.98 (0.91-1.00) Adedayo et al, 2005 Nigeria CS ≥18  579 498 0.86 (0.83-0.89) Brown et al, 2004 South Africa CS ≥18  50 38 0.76 (0.62-0.87) Ferris et al, 2004 South Africa CS ≥18  74 57 0.77 (0.66-0.86) Darder et al, 2004 South Africa LS ≥18  192 168 0.88 (0.82-0.92) Karcher et al, 2004 Uganda LS ≤18  76 52 0.68 (0.57-0.79) Byakika-Tusiime et al, 2005b Uganda LS < or > 1 44 43 0.98 (0.88-1.00) Hosseinipour et al, 2004 Malawi LS ≥18  141 134 0.95 (0.90-0.98) Omes et al, 2004 Rwanda LS ≥18  95 88 0.93 (0.85-0.97) Tu et al, 2004 DR Congo LS NA 30 30 1.00 (0.88-1.00) Traore et al, 2004 Burkina Faso LS ≥18  80 24 0.30 (0.20-0.41) Ramadhani et al, 2006 Tanzania CS 19-69 150 127 0.85 (0.78-0.90) Abbreviations: CS, Cross Sectional; LS, Longitudinal study; NA, Not available ORCT, Open Randomized Controlled Trial. tion of each study. Another column represented the number of non-adherent patients. The remaining columns represented the barriers to adherence. A barrier identified in each study is represented by the value ‘1’  or else, zero (‘0’).  The barriers to adherence were analysed and summarized with the aid of the software. RESULT Quantification of Adherence A total of 27 articles (full texts and abstracts) that met the inclusion criteria out of the 41 relevant studies were used in the meta-analysis. Some studies met the inclusion criteria but had no report on percentage of adherence. Rather, they had reports on predictors of non-adherence. A flow diagram of the studies included in this analysis is shown in figure 1. 13 countries in sub-Saharan Africa had studies on self-reported adherence that were found in electronic data bases. The adherence rate ranged from a minimum of 30% to a maximum of 100% as shown in table 1. The combined adherence in the analysis showed an adherence of 88.11% (95% CI = 87.39% - 88.81%) in the fixed effect model. For the heterogeneity of studies conducted, the Q    020 Int. J. Public Health Epidemiol. Figure 2. Forest Plot of adherence rate with 95% confidence interval (n = 27). statistic was very large (Q = 705.50, df = 26, P< 0.0001; I 2 = 96.3%). Since the variation in the studies were very large (I 2 = 96.3%), the random effect model was followed for data analysis. Following the random effect model, the combined percentage of adherence was 84.31% (95% CI = 79.48%  –  88.60%) as shown in figure 2. The Monte Carlo sensitivity analysis provided an alternative statistical method to evaluate the pooled propor-   0.0    0.2    0.4    0.6    0.8    1.0    Combined    0.84 (0.79, 0.89)   Ramadhani et al, 2006    0.85 (0.78, 0.90)   Traore et al, 2004    0.30 (0.20, 0.41)   Tu et al, 2004    1.00 (0.88, 1.00)   Omes et al, 2004    0.93 (0.85, 0.97)   Hosseinipour et al, 2004    0.95 (0.90, 0.98)   Byakika et al, 2005    0.98 (0.88, 1.00)   Karcher et al, 2004    0.68 (0.57, 0.79)   Darder et al, 2004    0.88 (0.82, 0.92)   Ferris et al, 2004    0.77 (0.66, 0.86)   Brown et al, 2004    0.76 (0.62, 0.87)   Adedayo et al, 2005    0.86 (0.83, 0.89)   Laurent et al, 2004    0.98 (0.91, 1.00)   Watt et al, 2010    0.94 (0.91, 0.96)   Senkomago et al, 2011   1.00 (0.97, 1.00)   Muyingo et al, 2008    0.94 (0.93, 0.95)   Byakika-Tusiime et al, 2005    0.68 (0.63, 0.73)   Bajurniwe et al, 2009    0.85 (0.79, 0.90)   Chabikuli et al, 2010    0.71 (0.61, 0.80)   Rougemont et al, 2009    0.75 (0.69, 0.80)   Davies et al, 2008    0.73 (0.64, 0.81)   Nabukeera-Barungi et al, 2007    0.89 (0.84, 0.94)   Okonji et al, 2012    0.84 (0.81, 0.88)   Weiser et al, 2003    0.54 (0.44, 0.64)   Diabate' et al, 2007    0.74 (0.71, 0.78)   Iroha et al, 2010    0.86 (0.81, 0.91)   Musiime et al, 2011   0.91 (0.88, 0.94)   Nachega et al, 2004    0.88 (0.78, 0.95)   Proportion (95% confidence interval)  
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks