This was a qualitative-methods exploratory, descriptive, and phenomenological study. Phenomenological studies aim to describe shared meaning derived from the lived experiences of several individuals.²¹

The study was conducted at the Chris Hani Baragwanath Academic Hospital (CHBAH) and Hillbrow Community Health Centre (CHC), which are public health facilities in Gauteng, South Africa. The CHBAH serves an immediate population of at least 1.5 million people living in Soweto and its surroundings while also serving as the tertiary referral centre for much of the Gauteng population (11.4m).²² Hillbrow CHC is located in the centre of the Johannesburg Central Business District.²² The CHC provides primary healthcare services to the local community including HIV testing and treatment. People who live within the catchment area of the Hillbrow CHC are primarily immigrants and migrant workers who have come to the city for work purposes.²³ The CHBAH is an enhanced surveillance site within the GERMS-SA laboratory-based surveillance network in South Africa.

The target population included people living with HIV (outpatients and inpatients), who were older than 18 years with laboratory-confirmed CM. Participants had to have a Glasgow Coma Scale score of 15 and living with a diagnosis of CM for more than 10 days and 30 days for inpatients and outpatients, respectively. We purposively selected 9 people living with HIV who met these inclusion criteria. Phenomenological studies typically only require data from a few individuals who have experienced the phenomenon to discover its core elements.²⁴

For inpatients, we screened the GERMS-SA database to identify prospective participants. Eligible patients were asked to participate, and their consent was sought. Outpatients were identified through pharmacy records; at Hillbrow CHC we retrieved the medical records of patients on fluconazole to assess their eligibility. Eligible patients were contacted by phone to confirm the date of their next clinic appointment. At this clinic visit, their consent to participate was sought.

Data were collected using a semi-structured interview guide during individual face-to-face, in-depth interviews. Interviews were conducted from September 2020 until November 2021. All interviews were audio-recorded, and field notes were taken. To maintain privacy, patients were interviewed in their cubicles and curtains were closed. Outpatients were interviewed in the consultation room. During the interviews, we adhered to coronavirus disease 2019 (COVID-19) safety protocols including face masking and maintaining a distance of 1 m. Interviews were conducted in the participants’ preferred language by the first investigator, who is conversant with the main local languages (Sepedi, IsiZulu, Sesotho and Setswana) spoken in the setting. Interviews lasted for 35 min to 60 min. The interview started by asking a broader general question, ‘What has it been like living with CM?’. Other open-ended questions were asked to provide a textural and organisational description of the experience of living with CM.

The interview recordings were transcribed verbatim by an independent provider. Interviews conducted in the participants’ home languages were translated into English by the researcher, and then transcribed.

Data were analysed manually by the researcher using the Moustakas phenomenological approach, which is a modification of the Stevick-Colaizzi-Keen method.²⁵ The four steps of the Moustakas approach are as follows:

Trustworthiness was maintained by upholding the principle of credibility, transferability, conformability, and dependability.²¹ Credibility was maintained by the extended period of the study.²¹ The researcher also read the transcripts while listening to the audio for quality checks. Transferability was ensured by describing the sampling methods and findings in detail so that they were transparent.²¹ Conformability was ensured by keeping audits trails of all actions via a Microsoft Word document as a means of opening the research process to outside inspection.²¹ Additionally, each step of the analysis was discussed with the co-authors, one of whom is experienced in qualitative-methods research. Further, the research process and findings were discussed with colleagues who are experts in the field as part of peer debriefing. Dependability was ensured by concurrently collecting and analysing data, which helped to refine some of the interview questions.²¹

Ethical approval was obtained from the Faculty of Health Science Research Ethics Committee of the University of Pretoria, reference no. 433/2020. Permission to conduct the study at the CHBAH was obtained from the Medical Advisory Committee and Hospital Chief Executive Officer. Permission from the Hillbrow CHC was sought from the Research Committee of the Johannesburg Health District, reference number 2020_06_005. Written informed consent was obtained before conducting interviews. With participants’ permission, interviews were audiotaped. Participants’ names were not used during interviews to maintain confidentiality during transcribing.

A total of 9 patients (5 men, 4 women) aged between 22 years and 47 years, and who were mainly black people (n = 8), participated in the study. One participant was mixed race.

Five themes and several sub-themes were identified (Table 1).

All participants visited healthcare facilities because they had severe headaches for more than 2 weeks. Other signs included painful neck or neck stiffness, vomiting, and painful eyes when looking into the light. Most participants had headaches that were persistent throughout the day and could not recall the exact time and place when the pain started. Participants initially thought it was just a headache that would go away, but they realised that the headache was not getting any better, and they had to seek help. Headaches did not start in any specific area, with some participants indicating the frontal part of the head and others from the back and the neck upwards. Participants explained that the pain could not be resolved at home and that the pain was only relieved once they were admitted to the hospital:

Participants described the pain as being more severe than any other pain they had ever experienced, even worse than labour pains:

One male patient who had been shot with a gun in his leg indicated that even the gunshot pain was much better than the pain he felt due to CM: ‘even a gunshot wound is less painful than this’ (Participant 5, Male, 22 years old).

Participants acknowledged that their lives changed once they developed CM symptoms. Some participants reported a loss of income due to loss of jobs while others reported being afraid of death. Additionally, participants perceived a reduced quality of life. Most participants were unemployed, but some had temporary jobs. Some of those with temporary jobs had to stop working to stay at home, spending their time in bed sleeping, and resting to relieve the pain, ‘I lost my job so really from here finish this … medication course, start the ARVs, get a job’ (Participant 5, Male, 22 years old).

Participants reported a poorer quality of life and an increased sense of being more dependent on others. Participants felt as if the pain had changed them and felt as if they were no longer themselves due to the changes affecting their daily routines and lifestyles. Some participants felt that they were a burden to their families because they could not perform certain tasks:

Some participants reported feeling depressed about their condition. Participants isolated themselves from friends to avoid being asked about their health, especially if they had notable weight loss:

Participants were afraid of succumbing to CM and its sequelae. Some of them mentioned they had heard about meningitis and someone who had died from it, although they did not know the type of meningitis. Others knew of family members or other people who had meningitis but never recovered fully and were still struggling with hearing loss, loss of vision, weakness of limbs, and cognitive slowness:

Regarding treatment, participants described their experiences with therapeutic lumbar puncture (LP) and their perceptions on the effectiveness of treatment. Most participants with serial LPs to decrease intracranial pressure mentioned that LPs helped to relieve headaches. Three participants noted that the LP procedure was painful and did not help; these 3 participants only had an LP at diagnosis. In contrast, 6 participants reported feeling better every time the doctor did an LP:

All participants indicated that they received better care at the hospital than that offered by general practitioners (GPs) before being admitted to the hospital. Inpatients felt they received effective treatment from primary healthcare clinics (PHCs) and the hospitals. Participants were referred from PHCs to the hospital. Few participants were aware of the medication they received in hospital. However, they acknowledged that they felt the medication was effective:

Health-seeking behaviour emerged as a strong theme. Participants initially sought help for their symptoms from other sources and not from health professionals. Six out of 9 participants self-medicated with over-the-counter analgesics such as paracetamol, aspirin, or combinations of these, while 3 participants were self-treated without medication. The spiritual or traditional healers and GP were seen as the next option after self-medication or self-treatment failed to improve their symptoms (Figure 1). The 3 participants who did not use any medication went to the traditional healers while the 6 participants who used over-the-counter medications consulted the GP (Figure 1):

Participants only sought help from PHC facilities (n = 7) or self-referred directly to the tertiary hospital (n = 2) when their headaches were unresolved. Seven participants who went to the local clinic were referred to the hospital at first contact, while the remaining participant self-referred to the hospital:

Notably, participants who consulted GPs were never referred to the next level of care despite their symptoms not improving.

Participants identified healthcare workers’ negative attitudes, their unhealthy lifestyles, and poor CM knowledge as barriers to effective HIV and CM care.

Some participants mentioned that they received terrible treatment from their local clinics. Participants felt that healthcare workers had negative attitudes, discouraging them from collecting their medication every month, especially for employed participants who could not take sick leave from work each month. One participant indicated that she had to send her child to collect her treatment as she could not take leave every month, but the nurse would not allow that:

Stigma emerged as a sub-theme. One participant indicated that he used to collect his treatment in the private sector using medical aid. He lost his job and income, and his medical aid was cancelled. He felt that he could not collect his medication at a public clinic because healthcare workers had negative attitudes toward people who are HIV-positive:

Participants also reported using alcohol, which caused them to default on ART. One participant indicated that he was consuming a lot of alcohol, especially during the festive season. Consequently, he forgot to take his medication because he was drunk by the time he had to take his medicine:

Poor CM knowledge stood out as a factor contributing to participants visiting unconventional places for treatment, including traditional healers. Participants’ understanding of CM was poor because some linked the symptoms with being bewitched, stressed, or being hit on the head. None of the participants linked CM to HIV or ART non-adherence. Some participants viewed the headaches as a spiritual attack or some form of witchcraft, and therefore they had to seek spiritual help from traditional healers. Some participants related that when they approached traditional healers, they were informed that someone was behind their headaches. Furthermore, the fact that the headache was persistent and severe was reason for some participants who did not believe in witchcraft to believe they had been bewitched:

This qualitative-methods study is the first in South Africa to explore HIV-associated CM experiences in routine care, filling an important knowledge gap. Recruitment of participants posed a significant challenge related to inability to trace outpatients on maintenance fluconazole at the PHC and the severity of CM in hospitalised patients. The use of a sample from one tertiary facility and one PHC setting also requires caution in generalising the results to other populations or settings.

CM-related health education should be integrated into post-HIV counselling for patients starting ART who are at risk of CM (e.g., CD4 count ≤ 200 (cells/µL), positive reflex CrAg test, and high HIV viral loads). Patients who are aware of the symptoms may be more likely to seek early treatment for CM. Further studies are needed to understand healthcare workers’ attitudes and assess the knowledge of GPs on the current guidelines for the diagnosis, prevention, and management of cryptococcal disease in adults.