In this systematic review we found that SES was not consistently associated with adherence to treatment among HIV infected patients. Since there was no study directly examining the association between SES and adherence in patients with HIV/AIDS, we evaluated the available data regarding the possible association between the major separate determinants of SES (income, education, occupation) and adherence. Although someone would have expected a clear association between SES and adherence to treatment based on data from studies on patients with chronic diseases other than HIV/AIDS infection, the evidence from the available studies does not fully support the existence of such an association in this patient population. However, a positive trend of association between levels of various SES components and levels of adherence to antiretroviral treatment is present among many of the studies.
By taking a close look at the data presented, it is noteworthy that among the reviewed studies that examined some of the main components of SES, most did not find a statistically significant association between these factors and adherence to antiretroviral treatment. It should be emphasized that a statistically significant association between income and education, two main determinants of SES, and adherence was found in only half and less than a third of the studies that examined income and education, respectively.
The existence of a possible association between income and adherence to treatment in HIV/AIDS patients was examined in 14 of the reviewed studies. Among the 7 studies in which income was found to be significantly associated with adherence, 4 concluded that the cost of antiretroviral treatment and/or poor living conditions were factors preventing patients from complying with treatment. If this financial obstacle was overcome, adherence was expected to reach considerably higher levels [23, 28, 30, 31]. In the remaining 3 studies, among patients having the economic ability to receive their medication, there was an association between the annual income and adherence [14, 21, 25]. It is presumed by the authors of one of the studies that patients with a higher level of income differ to those of lower/middle income, in terms of behavioral characteristics and hierarchy at the decision-making process, thus affecting their adherence to antiretroviral treatment . Furthermore, perceived economic support by a significant other was found to have a direct association with levels of adherence to antiretroviral treatment, in another of the reviewed studies . Such findings agree to the general idea linking stratification of income to disparities in health status and the will to adhere, placing the lower income patients on a deprivation scope, while allowing for higher income patients to adjust according to relative social status, possibly being influenced by other SES factors such as education and occupational status .
The existence of a possible association between level of education and adherence to treatment in HIV/AIDS patients was examined in 13 of the reviewed studies. Among the 13 studies that considered education as a probable factor affecting adherence to antiretroviral treatment, only 4 original studies [14, 16, 24, 29] proved a statistically significant positive association. Education, providing the basis of a stable future for each person, as well as altering the criteria used during the decision-making process and the knowledge to access health resources and information on disease and treatment, is a powerful implement and could possibly be influenced by policies targeted to enhance adherence among HIV patients [5, 6, 16, 29, 33, 34]. In 1 of the 4 studies, health literacy among those highly educated was also associated with higher level of adherence . Health literacy is related to educational level, but is influenced by other determinants as well, such as health care providers' supportive manner and instructional skills , should therefore be considered a sector in which external intervention – and further training – is applicable [29, 33, 35]. Of note, in 1 of the 13 studies that examined the level of education, a statistically significant reverse association between this variable and adherence was found, although this interesting finding was not elaborated further by the authors of the reviewed study .
Employment status was either not assessed or not found to be an independent factor associated with adherence, in the majority of the studies that we reviewed. Specifically, employment was found to have a significant impact on adherence in only 1 of 8 studies that examined this factor. The authors of that study postulated that having a busy workload might be an impediment to the patients' ability to adhere to antiretroviral treatment , therefore suggesting an adverse association between adherence to antiretroviral treatment and a demanding working schedule. Unemployment and lower occupational status have, however, been linked to lower levels of health status and increased mortality  and could be blamed for lower levels of adherence in terms of stress caused by job insecurity, physical exhaustion, and lack of control over one's working schedule (as was the case in the reviewed study) [13, 15], all of which could lead to a diminished intent and/or capability to follow antiretroviral treatment according to proper dosage and timetable . We feel that further research should be carried out in order to estimate the possible effects of employment and occupational status on HIV patients' tendency to adhere to antiretroviral treatment.
Our systematic review has several limitations. First, it was not possible to make a synthesis of the data using the principles of meta-analysis due to the fact that there was considerable heterogeneity among the reviewed studies. Adherence was measured by different methods in each of the studies and the cutoff percentage of adherence to treatment between 'adherent' and 'non-adherent' varied among the studies, depending on the authors' estimate. Furthermore, while most of the studies included patients generally following the model of life prevailing in the industrialized countries, some of the studies focused on populations having special economic, cultural, and social structures. Moreover, the studied patients received different antiretroviral regimens, ranging from monotherapy to HAART; the complexity of the treatment schedule affects the level of a patient's adherence. Second, SES was not focused upon as a homogenous group of specific factors in any of the reviewed studies, but was rather dispersed among its components, which were regarded as socio-demographic information. Therefore, we were forced to collect partial data regarding the association of such SES components, and adherence to antiretroviral treatment, where – and if – such an association was assessed. Occupation was only assessed in terms of employment status, as no data were given on status of occupation or working position of the patients. Additionally, we could not analyze the possible association between other SES proxy variables, such as the neighborhood, and adherence to treatment because the included studies did not report relevant data. Third, patients supposed to have lower SES, as perceived by the treating physician, are generally more likely to receive less complex antiretroviral regimens, and more information on how to maintain a satisfying adherence level. We cannot exclude that such an inequity could have occurred in the reviewed studies, as most studies were not set in a randomized controlled trial (RCT) environment, and include random HIV patients, therefore impeding our effort to find an association between levels of SES, and adherence to antiretroviral treatment.
Adherence is a complex, dynamic process that influences the outcome of HIV treatment and the patient's health status [6, 36]. It may change over time, as the health status or the patients' beliefs and attitude regarding the disease, the physician, and the treatment may alter, as well. As adherence does not concern only the patient, but the physician and the public health system too, it becomes evident that relevant factors cannot act independently, but instead they all interrelate [1, 6]. Lower level of adherence to antiretroviral treatment leads to recurrence of the symptoms, drug resistance, and increases the patient's viral load, thus affecting the patient-physician relationship in a negative manner and creating possible hazards for the community, in terms of transmission, viral resistance, social stigma, and financial and/or management problems within the public health system [1–4]. Predicting patients that are expected to have lower adherence, in an objective manner, could establish an individual approach to secure each patient's optimal response to antiretroviral treatment, according to each patient's specific characteristics [5, 31].
On the other hand, it has been noted before that physicians' choice regarding the medication they prescribe to their HIV patients is often influenced by their own estimates of expected level of patients' adherence to treatment, based on social stereotypes . In this way, HIV patients with a low SES are less likely to be prescribed triple therapy [34, 37]. However, the available evidence suggests that such estimates on expected patient adherence may have a limited accuracy and therefore should be treated with caution as they can result in harmful clinical consequences [30, 36]. Also, the time the physicians devote to their patients and the methods they use in order to educate them about the HIV infection/disease, and convince them about the importance of adhering to treatment, depends on their judgments about the sociodemographic characteristics of the patients [5, 36]. It is obvious that such an inequity in attention and instructions given by the physician, perhaps unavoidable in every day practice where patients gather in great numbers and time remains limited, results in uneven levels of co-operation and adherence between different patients.
Unlike SES, there were other factors, which were found to influence greatly and consistently HIV patients' adherence in the reviewed studies. Specifically, psychosocial factors such as depression [22, 24, 26, 28, 31], active drug [14, 22, 24, 26, 31] or alcohol use [14, 26], and lack of social support and stability were associated with suboptimal level of adherence [2, 3, 5, 8, 21]. Furthermore, cognitive factors such as self-efficacy and patients' beliefs and views regarding the disease and the effectiveness of medication (outcome expectancies) were found to be significant determinants of adherence [3, 4, 14, 27, 32, 38]. Also, adverse events were associated with lower level of adherence [4, 8, 30]. In general, complex schedule of drug therapy along with food restrictions were assessed as primary barriers to medication adherence [5, 6, 8, 9, 14, 21, 25, 27]. The quality of the patient-physician relationship played an important role as well. Acceptance, open communication, cooperation and trust in physicians were reported to be strong predictors of enhanced adherence [1, 2, 5, 6, 21].
In several studies it has been shown that SES is significantly associated with adherence to treatment in patients with chronic diseases [10–12]. Despite the fact that HIV infection is included among chronic diseases, it differs from all others. This is probably due to the fact that this infection is socially stigmatized, in grounds of transmission. It is not only a physical disease, but a psychological, mental, and social, too. In addition, this infection is connected with social discrimination, guilt, and prejudice [5, 28, 30]. HIV infection is a life-changing event, affecting the psychological status of the patient and results in his/her having to adjust again, in new conditions of life. It seems that during this process, cognitive and psychological factors are more important than SES for adherence to therapy.
In order for HIV patients to achieve higher levels of adherence to treatment, interventions regarding the patient, the clinician and the treatment have to be made [5, 6]. Specifically, helping patients to understand more about the HIV infection, as well as the antiretroviral treatment [5, 6, 16, 29, 33–35], coping with co-existing behavioral or psychiatric diseases [1, 3, 5, 6], and adjust medication schedules to the patients daily program or using memory helpers such as special pillboxes, reminders etc. [5, 6, 14, 15] are all important strategies. Additionally, the physician being consistent, vigilant, available, and explanatory can motivate the patient to adhere more to the antiretroviral treatment [1, 38]. Warning the patients about potential side effects and coping with them timely, checking the list of medications at each visit, giving written information or showing pictures so as to provide instructions, are alternative and effective ways to ensure patients co-operation and participation in the therapeutic process [5, 6, 34]. As for the health system, it has to be noted that having a medical insurance and easy access to primary care, receiving treatment by the same medical providers each time, receiving counseling by specialists, and not having to pay for the antiretroviral regimens, are factors that enhance adherence level [2, 4, 5, 9, 21, 30]. Improving a patient's financial and educational background is sometimes an impossible mission, however the aforementioned policies on educating and supporting the HIV patient can result in better adherence levels and should be investigated further, in terms of effectiveness.
Conclusively, the available evidence suggests that SES is not consistently associated with adherence to therapy among patients infected with HIV and it does not seem to be a major determinant of adherence to antiretroviral treatment. Many available studies suggest a positive trend among factors contributing to patients' SES and adherence to medical treatment among patients with HIV/AIDS, however such an association cannot be statistically consolidated throughout most of the studies included in our systematic review. It should be emphasized that it appears that there is a confusion regarding the accurate meaning of the term "SES" and thus it has been assessed in various ways. Future studies may further explain the different impact of SES to adherence to treatment between patients infected with HIV and patients suffering from other chronic diseases.