DATA PRESENTATION AND DISCUSSION OF FINDINGS 4.1 Data Presentation
4.3 Discussion of Findings
There are a variety of complex factors which promote and/or hinder adherence to treatment.
As such, the effectiveness of any intervention is unpredictable. Understanding the predictors of adherence is the first step in trying to improve adherence to antiretroviral therapies. A detailed understanding of the possible factors that can contribute to non-adherence will
UNIVERSITY OF IBADAN LIBRARY
138 greatly aid in the development of interventions to improve adherence, particularly for susceptible patients. It was, therefore, necessary to conduct context-specific research to unearth issues that must be dealt with for the improvement in the level of treatment adherence for HIV and TB co-infections.
The study found significant differences in the proportion of male and female in the sample.
This difference may have been due to the fact that women are more at risk of HIV than men, as other studies have also indicated (Afolabi, et al., 2009; Olisah, Baiyewu and Sheikh, 2010;
Landman, Ostermann, Crump, Mgonja, Mayhood, Itemba, Tribble, Ndosi, Chu, Shao, Bartlett and Thielman, 2008; Floridia, Giuliano, Palmisano and Vella et al., 2008). A study in Zimbabwe found out that risk of HIV infection in women increased with increased number of sex partners, but did not in men (Gregson, 2006). Another possible explanation for the gender difference observed in this study is that women get tested more often than men. This is because they have to necessarily undergo HIV screening during antenatal clinical services and, consequently, are more aware that they are infected (Olisah, et al., 2010). Other studies have also found out that more women than men are tested for HIV repeatedly, thus explaining why women may be appear to be more infected than men (Le Coeur, Collins, Pannetier and Lelièvre, 2009; Venkatesh, Madiba, De Bruyn, Lurie, Coates and Gray, 2011). Subsequently, on competent advice on the benefit of taking antiretroviral drugs, more women take up treatment.
It is instructive, and of grave concern that a majority of the respondents (72.6 %) were within the labour force, that is, the ages at which people are expected to be economically productive.
Besides, they are ages at which most reproductive activities take place. The economic implication of the HIV and TB epidemics is devastating, especially, because of the age it affects most. The impact of HIV on the economically productive age has long been recognised and documented (Morison, 2001). Hilhorstaa, van Liereb, Odec and de Koningd (2006) examined the socio-economic impact of HIV in Benue State, Nigeria and argued succinctly about the costs. Hilhorstaa et al. (2006) reported high cost in terms of expenditures for health care, funeral and mourning, and time spent providing care for the infected, which places serious demands on income and productivity; while the diversion of resources have implications for investment and savings.
Results also suggest that more young women are exposed than young men. The older men usually target younger women for sexual pleasure, since they are the ones who have money,
UNIVERSITY OF IBADAN LIBRARY
139 and can afford to satisfy the needs of young girls for sexual gratification. This exposes the young women to sexually transmitted diseases including HIV. The fact that older men are more exposed to HIV and TB than older women suggests that, and is also confirmed by the data on marital status; when women grow older and get married, their risk level reduces to a smaller rate. On the contrary, men who still engage young girls in sexual intercourse, even after marriage, are at a higher risk.
The problem then is that as married men get exposed, their wives invariably get exposed as well. Although the proportion of singles in the sample is higher than the married, the proportion of the married is still high enough to raise concern and alarm. It must be noted that the proportion of those widowed is high (14.4 %). This reveals one of two things: that their spouses may have already died of the diseases (HIV and/or TB) or that because their spouses had died, they started some other sexual liaisons, thereby exposing themselves to infections.
More than half of the respondents had adequate knowledge of their treatment and the implication of not adhering to drugs. Knowledge was however modified by the length of time patients had been on HIV, but not TB, treatment. The longer the period on drugs the better understanding of treatment. However, this study found out that knowledge of treatment and the implication of non-adherence is not significant in predicting level of adherence. The presupposition that knowledge of treatment plays an important role in patients‘ adherence to treatment is therefore brought into serious question. On one hand, this finding revealed that there gap between knowledge and practice. On the other hand, it is contrary to studies that found knowledge to be predictive of higher level of medication adherence. For example, Kalichman, et al. (2008) had found a positive association between health knowledge and adherence to treatment.
The level of adherence was found to be low with a mean of adherence of 51.7%. Only 38.1%
were in the high adherence category. Patients in the high adherence category were those that would fit into the category that the WHO defined as reaching 90% to 95% level of adherence.
The level of adherence found in this study is lower than that reported by Erah and Arute (2008), which noted a 59% adherence. In fact, another study in Nigeria that involved HIV and TB co-infected patients reported a 41% adherence to 95% of medication taking (Njepuome and Odume, 2009), which is closer than the rate observed in the present study.
The fact that the level of adherence in the present study is closely related to that of Njepuome and Odume (2009), which like this study, involved infected patients, suggest that
co-UNIVERSITY OF IBADAN LIBRARY
140 morbidity may play a significant role in adherence behaviour. Other studies had previously reported a higher level of adherence (see Kalichman et al., 2008; Osborn, Davis, Bailey and Wolf, 2010). For example, Osborn et al. (2010) reported a 71% mean level of adherence; the only drawback was that they conducted a phone-based pill-counting survey which can be fraught with bias. From the perspective of public health and service delivery, treatment non-adherence undermines the efficient distribution of scarce resources and represents wastage of public resources (Erah and Arute, 2008).
Different levels of adherence have been reported in earlier studies in Nigeria. For instance, the levels reported for studies conducted in Kano (northern Nigeria), Sagamu, Niger Delta and Benin City (Southern Nigeria) were 49.2% (Nwauche, Erhabor, Ejele and Akani, 2006), greater than 85% (Idigbe, Adewole, Eisen, Kanki, Odunukwe, Onwujekwe, Audu, Araoyinbo, Onyewuche, Salu, Adedoyin and Musa, 2005) and 80% (Mukhtar-Yola, Adeleke, Gwarzo and Ladan, 2006); while Afolabi et al. (2009) reported 44%. In several countries in sub-Saharan Africa and North America, varying levels have also been reported (Mills, et al., 2006). However, significant proportions of HIV-infected patients do not reach high levels of adherence, and this can lead to devastating public health problems. Getting patients to take drugs everyday without failure for the rest of their lives is one of the biggest challenges.
It is important to also note that patients may take the total number of prescribed doses, but may not take these at the appropriate times. Melbourne et al. (1999) in a previous study had found out that within a subgroup of patients who took more than 90% of doses, there was significant dosing fluctuation in 50% of patients during the first two months of treatment. The dosing fluctuation ranged from taking the medication within two hours of the prescribed dose time to greater than two hours of that defined time.
It is quite possible that adherence may be under- or over-estimated because of the difficulties involved in its measurement. Bell et al (2007) had stressed this reality when they concluded from their study that there were serious complexities in the measurement of adherence and probable overestimation of adherence by pill count and self-report. Of course, these are the main methods used in developing countries; this consequently raises concerns about the development of drug resistance.
Although gender of respondents did not predict adherence at the bivariate level, it was a significant factor at a higher level of analysis. Male respondents were more likely to be in the
UNIVERSITY OF IBADAN LIBRARY
141 higher level of adherence than females. This may be related to other factors like income and ability to afford their needs such as food and transportation to treatment centres for drugs. It was found that men have higher chances of being in the high adherence category than women. By access to resources and ability to afford health care, men are better placed than women, and as such, have the better chances of continuing in treatment-taking. Gender differences in adherence to treatment have also been reported by other studies. Similarly, Mirjam-Colette, Pisu, Dumcheva, Westfall, Kilby and Saag (2009), Salami, Fadeyi, Ogunmodede and Desalu (2010), Applebaum, Richardson, Brady, Brief and Keane (2011) and Hawkins, Chalamilla, Okuma, Spiegelman, Hertzmark, Aris, Ewald, Mugusi, Mtasiwa and Fawzi (2011) found out that the male gender adhere more to treatment than the female.
Contrary to this study‘s findings, Daniel and Oladapo (2006) reported more defaulting among males than females in a survey of TB patients in Sagamu. Daniel and Oladapo argued that the role of men as breadwinners of the family, where they are expected to leave the house so early in the morning in search of work to provide for the family makes men more likely to default from daily clinic appearance for their DOT medication.
The unemployed and patients who were farmers were more likely to default, that is, be found in the low adherence category than those who had better or less strenuous jobs. Similar findings have been reported that indicated livelihood security to predict adherence to treatment (Rachlis, Mills and Cole, 2011).
Moreover, both marital status and income per month also significantly predicted patients‘
adherence to treatment behaviour. However, instead of a positive relationship, the study found an inverse association, that is, the higher the income the more likely the patient will be in the low adherence category. This is in contrast to studies by Brinkhof, Dabis, Myer, Bangsberg, Boulle, Nash, Schechter, Laurent, Keiser, May, Sprinz, Egger and Anglaret (2008) and Falagas, Zarkadoulia, Pliatsika and Panos (2008) review. Although Falagas et al.
(2008) found no conclusive support for existence of a clear association between socio-economic status and adherence; they found income, occupation and education to be positively associated with adherence. In Biadgilgn et al. (2008), a number of variables were found to be significantly associated with adherence, which touched on financial security.
Biadgilgn et al. (2008) found out that children whose parents did not pay a fee for treatment and children who had ever received any nutritional support from the clinic were less likely to adhere. The implication of Biadgilgn et al.‘s (2008) findings, especially the fact that children
UNIVERSITY OF IBADAN LIBRARY
142 whose parents did not pay for the treatment are less likely to be adherent, is that whereas financial hardship is a strong factor in non-adherence, completely free provision of medication can also become a negative factor. People usually presuppose that free medications are substandard and not important, and are wasteful of it. Although this issue could not be examined because all patients were receiving free medicines for HIV and TB, financial burden came up as a serious reason why respondents missed their medication.
This study also found out that patients who were resident in urban areas were more socially isolated than those in the rural areas. This finding was contrary to expectation. The rural community is where patients should have felt more isolated because everybody knows everybody else and the news that one is HIV positive can spread quite easily leaving the patient at the scorn of the entire village. The rural areas strong community spirit, which may work positively to counter the shame that would otherwise be felt by infected person.
However, the broader sociology of urban life holds that there is generally more social isolation in the urban than the rural areas. Given that social isolation did not significantly explain patients‘ adherence level, this rural-urban difference was of no effect in the explanation of adherence behaviour.
It also appeared that the shorter the time patients spend on medication for HIV, the more isolated they feel. This is because having newly enrolled in the treatment regime, the patient will still be very sensitive and ashamed of their status, but gradual counselling and the realisation that being positive is not the end of life, and that drugs can sustain and maintain one‘s life as long as any other person, the patients will feel less troubled and associate more.
Nevertheless, patients newly placed on treatment were more likely to be adherent than those who have spent longer time. This might be as a result of the zeal and hope with which a newly diagnosed patient may have compared to those who have been on the treatment relatively longer, and who have witnessed the effects and frustration of constant and continuous medication. Thus, it can be argued that the perceived benefits of taking medication, a basic postulate of the health belief model, works for HIV patients in the earlier stages of treatment, but the effects wane with time.
Given that appointments at the treatment centres are on a two-week interval, the cumulative cost that respondents bear on transportation in only a month will be more than N1000.00.
This is more evident, especially for those whose monthly income ranged from five thousand naira and below, who showed a mean transport fare of 496.7 naira per visit to the treatment
UNIVERSITY OF IBADAN LIBRARY
143 centre. However, it was discovered that patients who lived within walking distance of the treatment facility, and those who pay more than 500 naira were more likely to be non-adherence than those who pay a maximum of 500 naira. By implication, patients who live sufficiently far from the treatment facility but not too far to be financially too difficult to become an access problem, are more likely that those who live close to the facility to adhere to treatment. This finding corroborates the findings of Charurat, Oyegunle, Benjamin, Habib, Eze, Ele, Ibanga, Ajayi, Eng, Mondal, Gebi, Iwu, Etiebet, Abimiku, Dakum, Farley and Blattner (2010) who also found this to be the case in a Nigerian study. Although the number of treatment facilities in Nigeria continues to increase, patients may continue to avoid accessing care from facilities within their communities, because of stigma (Charurat et al.
2010). As a result, scale up of treatment facilities must be coupled with support from the communities.
Family support is expressly seen by patients as central to medication adherence. One of the main drawbacks to its maximal utilisation is that the infected persons tend to deprive themselves of it by becoming withdrawn and cynical about expressed or given support-related behaviour. The fear of condemnation and stigma from family members and friends, and for the family as a whole, make them to hide their status. Even when they disclose their status, they lack the emotional ability to receive and appreciate support. They have mixed feelings about the support promised and received. Just as in Roberts and Mann (2003) and Edwards (2006), this study found out that difficulty with taking medicine in the presence of loved ones (family and friends) is a strong reason why patients default in their treatment regimen.
However, those who are open about their problem with their family, and have received support thereof found family support very useful in attaining the required level of adherence.
In some cases, family support may be difficult because both the family and community can become stigmatised as a result of an infected individual member of a family. When families are stigmatised, uninfected members experience unhappiness, making it difficult for family support to be provided and received with love. This study confirmed the finding of Shin et al.‘s (2009) that low social support is associated with non-adherence. Adherence interventionsmay be unsuccessful unless they target the underlying psychosocial and social challenges faced by patients living with Tb and AIDS.
UNIVERSITY OF IBADAN LIBRARY
144 Similarly, Mavandadi et al (2009) found out that social relationships play a significant role in the well-being of HIV positive people. As is the case with other chronic health conditions, individuals co-infected with HIV and TB often experience challenges that place a lot of demands on coping resources and impact their quality of life, including a reduced ability to participate in daily activities (O‘Dell, 1996), adherence to complicated treatment regimen (Tsasis, 2000) and changes in social network composition (Shippy and Karpiak, 2005). This last factor is very significant because in cases where people suffer from diseases which are stigma-related, they tend to lose that network of social relationships, a resource which otherwise should be useful to them in coping with the condition.
Family support for people who should be the breadwinner of the family can produce negative results if considerable understanding is not applied. In such cases patients have a heavy feeling of irresponsibility because they cannot carry out the requirement of providing for their families. Support can, however, help to counter the effect of the personal guilt and shame they feel.
Another problem that could hinder the effective deployment of family support in improving adherence is the feeling that the infected person is going to die irrespective of the support the family gives. Both infected persons and their family have a fatalistic belief at the back of their minds that once somebody has HIV with a TB co-infection, they will die even if not immediately. Fatalism creates a feeling of hopelessness for both family and the patient, leading to resignation, which is not conducive to improved medication taking. This finding is in tune with those of Wrubel, Stumbo and Johnson (2008) that discovered that discordant couples have this fear that their infected partner will pass away.
Moreover, it appears that what patients really need from family is emotional connectedness more than material or instrumental support, although both are necessary. In a study on perceived social support and medication adherence among African American women, Edwards (2006) had found out similarly that emotional and instrumental support were important, but the former is more in terms of expressed love, care and commitment is necessary in scaling up adherence. Ciambrone (2002) found that where families give instrumental support without emotional support, its effectiveness as a means of encouraging sustained treatment adherence is diminished.
UNIVERSITY OF IBADAN LIBRARY
145 Fear of perceived and actual side-effects of medicines was also an issue with many patients.
This is in tandem with a research on the factors hindering HIV positive people from taking up antiretroviral therapy and remaining in it. In that study, Rogowska-Szadkowska et al (2009) demonstrated a significant degree of prejudice regarding antiretroviral therapy among asymptomatic patients, which contributes to the decision of HAART refusal. The implication is that most HIV positive people would not want to adhere to treatment, because of the fear they have of the antiretroviral. When HIV is complicated by other infections, such as TB) as is the case with patients in the present study), the fear of the combined drug may become morbid, and act as a determining factor in patients‘ observance of their treatment.
UNIVERSITY OF IBADAN LIBRARY
146