From the discussion so far, it is obvious that stigma is the key underlying reason for the reluctance of many people to take up voluntary counselling and testing leading to antiretroviral therapy treatment. According to Niehaus [10], the main cause of stigma is the association of the disease with death rather than sexual promiscuity. Writing about South Africa, he acknowledges that this assertion is not entirely new but it shifts the emphasis for the cause of HIV-related stigma from immoral sexual behaviour or unprotected sex to fear of death. He observes that although immoral sexual behaviour is frowned upon by society, the sentiments against it are not strong enough to lead to the persistent stigma related to the disease and positive persons. In some African cultures, men are allowed to marry more than one woman and teenage pregnancies due to pre-marital sex are also common. According to Burn and Grove [9], this perception of HIV infection illuminates many aspects of people’s responses to the disease, including the (non) use of health care services. It also helps to explain why most clients preferred to use services in privacy and hide their status from relatives and others.

These views about the disease are common in Ghana. In a study on the disease among the Akan in Ghana, Crentsil [11] observes that positive persons are often stigmatised because the disease is believed to be incurable and the fear that the infected person is likely to die of it. Likewise, the pain and suffering the person goes through before death is generally considered as shameful and disgraceful, and in some cases lead to perfunctory funerary rites. This is often due to the fear of contagion associated with the corpse of person who died of HIV infection. Such death is described as ‘bad death.’ The concerns of clients in this study on the risks associated with the disease resonate Niehaus [10] and Crensil’s [11] findings that HIV is a deadly disease. The fear of death associated with the disease and the shame that often characterises HIV-related death could frighten relatives and others to distance themselves from its sufferers. It is therefore not surprising that most clients in this study said that their relatives would reject or ostracise them if they got to know of their status.

Confidentiality is central in the communication between health workers and clients in the provision of counselling, testing and treatment services. The professional ethics of health workers obliges them to keep information obtained in contacts with clients to themselves, as stipulated in the Ghana Draft Policy Document on HIV/AIDS by MOH/UG [12]. This was also the rule in the facilities where the study was carried out. Maintaining confidentiality by health workers in matters relating to HIV infection, care and treatment is therefore about trust. However, the trust clients have in health workers may be eroded where it is promised and not fulfilled. The possible lack of confidentiality in the health care setting, on the other hand, is a constraint on the use of services and discourages clients and potential clients from accessing counselling testing and treatment services.

Although health workers acknowledge that confidentiality is core to the provision of counselling, testing and treatment, they see the possible harm of strictly enforcing confidentiality with unintended risks for others. It places them in a dilemma as to whether they should maintain confidentiality or not when they know that a client is likely to infect others. As Nuwagaba-Biribonwoha et al. [13] point out, the dilemma of nurses regarding patients who feel unable to disclose their test results and who may go home to infect partners and unborn children is that they struggle with their professional ethical obligation to maintain confidentiality. In this study, health workers encountered the same problem. Health workers face the dilemma of either observing the rights of clients to maintain confidentiality as required or ignoring this right and breaching confidentiality. Maintaining confidentiality becomes more complicated when health workers are expected to help prevent the spread of the disease by providing information on behaviour change to the infected and uninfected.

It follows that some health workers in this study necessarily disclosed clients’ status to others such as spouses or relatives without their consent. Health workers did that in the interest of clients, particularly in the case of infected mothers. The spouses or relatives of these mothers were involved in counselling on the need for such mothers not to breastfeed their new-born babies. Such disclosures were motivated by the belief that those relevant others would help clients to adhere to treatment or also present for counselling and testing. However, some of these mothers held the view that the nurses did that to embarrass them for voluntarily disclosing their status to relevant others. Others also blamed the nurses for making such disclosures which negatively led to marital difficulties It is in view of these challenges that clients are often encouraged to go for couple counselling, shared confidential or arranged disclosure through which health workers helped many of them to inform their spouses and relevant others about their status. This, in most cases helped to reduce the negative repercussions associated with positive test results in the case of women and led to the use of treatment services. One study in Kenya by Farquhar [14] similarly showed that pregnant women who were offered couple counselling with partners were three times likely to report using ART for prevention of mother-to-child transmission. Another study by the South India AIDS Action Group (SIAAP) [15] in Tamil Naadu reported that couple counselling schemes have shown success in reducing the levels of violence and the numbers of abandoned women after diagnosis. Obviously, the fact still remains that in most of the cases, health workers disclosed the status of some clients in the interest of public good.

The policy of maintaining confidentiality also has wider implications for a client’s therapy management group and the health worker-patient relationship. In most cases, especially in the Ghanaian tradition, relatives and sometimes neighbours often play a significant role in the care and treatment of a sick person as cultural responsibility. In fact, these individuals sometimes come together as a group and take all decisions regarding care and treatment on behalf of the patient. The group’s responsibility goes beyond the popular and folk medicine and extends to the hospital setting. It is for this reason that Gruskin et al. [16] report that in many sub-Saharan African countries, patients are often accompanied to health care settings by family members. Thus, ward environments and clinics tend to be characterized by over-crowding and lack of space for private discussions (see also Mulemi [17]). In his study on the search for therapy in Zaire, Janzen [18] refers to these people as the ‘therapy management group’ of the patient. He characterises the group as mainly a set of close kin members who help with the management of illness or therapy for a kin member. Members of the group thus know the type of sickness or disease the patient is suffering from in order to take appropriate decisions on cure and treatment, either in the hospital or in the folk system (see also Bossart [19]). It also implies that health workers cannot maintain confidentiality about the patient’s health problem since some relatives, as therapy managers of the patient must always be present during consultations.

The therapy management group could for example withdraw its support or assistance to the client for lack of information about his health condition to decide on how he can be helped. It may be for this reason that Mellvray and MacClean as cited in Janzen [10] exhort clinicians working in hospitals and medical centres in Africa to be sensitive to their patients’ therapy managers and to share information with them. Sharing information is likely to enlist the full support of the group towards the effort to find a cure or treatment for patients. The advice by Mellvray and MacClean also shows that in sub-Saharan Africa, it takes the combined effort of health workers and a client’s therapy management group to successfully cure or treat a patient of his sickness. The assumption is that when information about the patient’s health problem or sickness is shared with the group, rather than maintaining confidentiality, the patient gets a decent treatment beyond a clinical decision. The present study suggests that maintaining confidentiality in the care and treatment of HIV may complicate the work of care providers and also create problems for clients due to the crucial role of the therapy management group.

Another issue emerging from the data presented in this paper has to do with the debate of either providing HIV-related care and treatment services in isolation of or integrated with other services in the hospitals. The debate has been necessitated by the obvious preference of clients for centres and clinics sited in discrete or obscure locations in the hospitals due to the persistent stigma associated with the disease. In essence, the debate for and against integration or isolation by health workers is basically aimed at conveniently providing services to clients so they could use services in privacy and conceal their status from relatives and friends. Both arguments have merits and demerits which call for critical examination by health workers. It is therefore the responsibility of health care professionals to decide whether integration should be ‘partial’ or ‘complete’. In the case of partial integration, health workers would have to make it clear which aspects of counselling, testing and treatment need to be provided alongside other services, and in which specific units. With regard to total integration, it is important for health workers to identify which particular unit or units such services could be conveniently provided for clients to feel secure and to continue to access services. If isolation is chosen, it would be necessary to find locations in the hospitals which would be appropriate for such services. Perhaps, the locations for such services could be tucked away from open places to prevent other patients from identifying clients while they use services. This observation is in line with an assertion made by Odeny [20] in a study carried out in rural Kenya on integration of HIV care with primary health care services. The authors pointed out that stigma may be reduced if a patient’s HIV status cannot be determined by the general public simply by observing the physical location where a patient is receiving care. In the event that health workers decide to integrate HIV-related health care with other services, then they may consider putting up new facilities at obscure locations in spite of the perennial lack of funds. However, Odeny [20] opined that integration of services may lead to other opportunities for accidental disclosures of one’s HIV status which can be a barrier to using services. For instance, a clients’ status could be known to another patient who is not HIV-positive in the process of accessing care in a facility which has an integrated service delivery. In this regard, the effect of integration on stigma is unclear.

The debate on integration or non-integration of HIV counselling testing and treatment with other health care services finds meaning in a study conducted in Kenya and Swaziland by Population Council [21] and three other non-governmental organizations on the integration of sexual and reproductive health and HIV services. Conducted from 2008 to 2013, the Integration Initiative sought to generate rigorous evidence on feasibility, effectiveness, cost, and impact of different models for delivering integrated HIV/AIDS services in settings with high and medium HIV prevalence in sub-Saharan Africa. The study demonstrated that most clients prefer fully integrated services to save time and money and it also reduce the fear of stigma because clients trusted the providers of such facilities. Similarly, Baotran [22] indicated that the findings of their study on patient satisfaction with integrated HIV and antenatal care services in rural Kenya suggests that integration is highly acceptable to most HIV-infected women. According to them, integration improved HIV-infected women’s satisfaction with overall clinic experience. They attributed women’s satisfaction with integration to the ease of receiving care for both services in one single visit, removing the logistical difficulties of standing in line for services at two different clinics, and the possible disclosure of their status to other non HIV-positive persons. It is for this reason that some health workers in this study also advocated for the integration of HIV-related services into other care services in order to address the concerns of clients who in most cases are more worried about people knowing their identity as PLWAs than care and treatment they would receive in the facilities.

Whichever option health workers or planners decide to go for, one important thing that they must not forget is the interest and welfare of clients who are the ultimate beneficiaries of health care services. Integration or isolation of the services must ensure that clients are guaranteed privacy in the use of services to avoid identification in the health care setting and prevent stigmatisation. This is likely to encourage clients to continue accessing services and, also motivate potential clients to use services.