Participant selection.

We used purposive sampling to select key informants based on their knowledge and/or experience of the COVID-19 health system response in Gauteng, and the implementation of the overall strategy in the five health districts and the various health facilities. We selected individuals from the following six categories of key informants: executive or senior managers or officials from different levels of provincial or local government; senior clinicians or frontline health workers in hospitals and clinics; professional associations and/or unions; academics, researchers or technical experts; the private health sector; civil society/ non-governmental organisations or health advocacy groups.

Data collection tool.

We developed two interview schedules in English: one for executive managers and one for all the other key informants (S1 Appendix). The semi-structured interview schedule for executive managers consisted of five main subsections. The first section focused on the key informant’s role in the COVID-19 response. The second section focused on a description or overview of the actual health system response to COVID-19, and included probes related to LMG and the first three pillars in the conceptual framework. The third section focused on stakeholder involvement and communication, and included probes on the fourth pillar of the conceptual framework. The fourth section focused on key informants’ perspectives on the COVID-19 response and/or strategy, notably their perspectives on the innovations, contestations or weaknesses, unintended consequences (both positive and negative) of the COVID-19 response and its implementation. The last section focused on key lessons from the management of COVID-19 interventions in Gauteng and recommendations for future management of health crises (including a possible second wave).

The semi-structured interview schedule for all other key informants was a truncated version of the key informant schedule for executive managers, and excluded sections 2 (specific response) and 3 (stakeholder involvement).

The research team reviewed the interview schedules for content validity and clarity of questions.

Research team and preparation for data collection.

The research team consisted of four individuals. The principal researcher (LCR) holds a South African Research Chair, has a PhD in health systems research, and has more than three decades of health systems research, leadership and management experience. The second author (CM) is a retired public health medicine specialist and paediatrician, with four decades of management and health system experience. The third author (BM) is a PhD candidate, holds both an undergraduate health professional qualification and a Master of Public Health degree. The final author (SM) holds an undergraduate health professional qualification, a Master’s degree in Nursing Education and a PhD in health systems research. Cumulatively, BM and SM have two decades’ experience of working as health professionals.

The research team jointly developed and discussed the interview schedules prior to their finalisation. We also held a separate meeting to go through the questions and did two trial runs to ensure readiness to interview the key informants and to collect the data.

Data collection.

We conducted the interviews between 1 September and 15 October 2020. Each participant was contacted via email to request voluntary participation in the study. Following informed consent, we arranged the interview date and time with each key informant. The researchers emailed the information sheet, informed consent forms, and the relevant interview schedule to the study participants. All interviews were conducted in English and virtually via Microsoft Teams or Zoom. We conducted telephone interviews on two occasions when technology failed.

Each interview began with an introduction of researchers to the participant, assisting the participant to be familiar with the virtual platform and putting the participant at ease. This was followed by an introduction to the study, and an explanation of the voluntary nature of participation. Prior to the start of the interview, the researchers confirmed voluntary consent for participation and recording of the interview.

Following informed consent, the research team used the semi-structured interview schedule as a guide to explore each participant’s perspectives on the COVID-19 response and the management of the COVID-19 pandemic in Gauteng. The researchers used probes to obtain details and clarification of responses. Each interview lasted between 30 to 60 minutes, but the duration varied depending on the key informant’s responses. We made detailed notes during the interviews, and complemented these notes with the audio-recordings.

All interviews were recorded digitally and labelled with a key informant code. All audio-recordings are kept on a password-protected computer to ensure confidentiality.

Data analysis.

Following data collection, the principal researcher examined and collated all the raw data and submitted the collated data to the other researchers for review. The four researchers held several meetings to discuss progress, debrief on the completed interviews, identify the emerging issues, identify gaps and assess whether new information emerged from the interviews.

Each of the four researchers analysed the data independently, using thematic analysis [54]. Each researcher read and reread each transcript independently to familiarise herself with the data, and to get a sense of the whole interview. The transcripts were coded line by line, by writing key words on early impressions, using the direct words from the transcripts [54]. Each researcher made notes on reflections from the data. Following the coding stage, each researcher developed themes, interrogated and evaluated the themes for similarities and differences in meaning across different categories of key informants.

Once the coding was complete, the four researchers held a meeting to discuss the independent codes and themes, and to reach inter-coder agreement. Following this, the research team held several other meetings to analyse the data, discuss and finalise the themes, their meaning and interpretation. We used an iterative process of inductive coding and theme development, followed by interrogation of these codes and themes in light of the study objectives and the selected conceptual framework.

Ethical considerations.

The Office of the Gauteng Premier commissioned the original case study. All researchers signed confidentiality and non-disclosure agreements. The principal researcher (LCR) requested advice from the Chair of the Human Research Ethics Committee (Medical) of the University of the Witwatersrand in Johannesburg on whether an ethics submission should be submitted prior to conducting the interviews. The Chair indicated that the confidentiality and non-disclosure agreements were sufficient to undertake the Gauteng case study.

During the data collection, the research team provided a detailed information sheet to all key informants, as well as consent forms for conducting the interview and for the audio-recording. Prior to the start of the interview, we provided a verbal explanation of the study, and informed participants of the voluntary, confidential, and anonymous nature of study participation.

The researchers obtained written approval from the Office of the Premier to submit this journal article based on the health case study for possible publication. Subsequently, the principal researcher also obtained an ethics waiver from the Human Research Ethics Committee of the University of the Witwatersrand in Johannesburg (#W-CBP-210331-03) to enable publication.

We complied with the principles of the Singapore Statement on Research Integrity [55], and use participant codes to ensure anonymity. All the data are stored on a password-protected laptop. The audio-recordings will be destroyed two years after publication of the case study.

Trustworthiness and validation.

The researchers ensured credibility by prolonged engagement with the documents, the key informant data and the audio recording of the interviews [56]. We ensured dependability by having a semi-structured interview schedule and a detailed description of the methods, data collection and analysis. Conformability was ensured through independent coding of the key informant interviews, discussion of, and agreement on the themes by the four researchers [56, 57]. We validated the data by doing two presentations to key stakeholders in the Office of the Premier, and by submitting a detailed written report.

As the interviews were conducted after the first wave of COVID-19 infections, and in light of a second wave of infections between December 2020 and 15 February 2021, we sent the interview transcript and an abstract of the case study to 15 key informants in February 2021. We requested these individuals to comment voluntarily on their original interview responses, and the summary of the results. Six of the key informants responded, confirming their initial responses, and support of the emerging themes. One key informant added more details to her original transcript, but without altering the core messages obtained during the first interview.

Theme 1: Innovative, multi-sectoral and comprehensive provincial COVID-19 response.

The first theme, derived from both the document review and the interviews, centres on the multi-sectoral and comprehensive response to the COVID-19 pandemic, and the innovations, whether driven from the top, or those that happened organically.

The National Disaster Management Act provides the legislative framework for the COVID-19 pandemic response in South Africa [34]. In concert with national developments, the Gauteng Provincial Government adopted a strategic response to the COVID-19 pandemic that aimed to address the dual challenge of saving lives and saving the economy [45]. The Gauteng COVID-19 Strategic Response incorporated a detailed risk assessment that took account of multiple levels of deprivation at the district level and consisted of six pillars: a comprehensive health response; food security and social relief; state capacity and adaptability; economic response; social mobilisation and human solidarity; and law enforcement and compliance [45].

Several high level governance structures were set up, the most important of which were the Provincial Coronavirus Command Council (chaired by the Premier) and the Provincial Disaster Management Command Centre (PDMCC) co-chaired by the Provincial Director-General and the Provincial Police Commissioner [45]. Each of the six pillars translated into a work stream that presented updates to the PMDCC, which in turn reported to the Provincial Coronavirus Command Council, and to the Provincial Executive Council (i.e. the Provincial Cabinet).

Several key informants commented positively on the visible and strong political leadership by the Gauteng Premier and the former Member of the Executive Council (MEC) for Health.

We were led properly in our response. The Provincial Command Centre had given us clear, accurate instructions on what needs to be done e.g. the setting up of quarantine centres. It showed bold leadership (KI 8, Central Hospital Manager).

Two key informants from civil society organisations also echoed the strong, visible political leadership. One said:

The Premier has given the impression of efficiency and transparency, and that has inspired confidence, and the belief that everything was under control. I was impressed with the slides presented at the media briefings, and the attempt at showing hard data (KI 23, Civil Society).

Several key informants reported that the pandemic provided an opportunity for innovations at various levels of the health system. In one health district, one key informant reported that contact tracing started early March 2020 when the country was in hard lockdown, with very few reported COVID-19 cases. The district health team developed written standard operating procedures. They established two types of tracer teams: physical tracers (primarily nurses) whose responsibility was to visit the homes of COVID-19 cases to test and educate other members of the households, and telephonic tracers (lay counsellors or other tracers), who were responsible for monitoring the cases and contacts every day for the 14-day isolation period.

The public sector hospital managers and clinicians interviewed were of the opinion that the success of the COVID-19 response in Gauteng was largely because of proactive managers at health facilities, the excellent teamwork, innovation or improvisation necessitated by the pandemic, and the contribution and sacrifice of frontline staff.

The response from the hospital was quite quick. By January [2020], we already started training on COVID-19. The team work was amazing–everyone pulled together to get the work done–they were paid overtime, the hospital hired agency nurses -the province only gave extra staff from 1 June 2020. Our psychiatrist set up an online psychosocial support for staff, which was then taken up by province. We did all the protocols- we gave them to province, not the other way. (KI 7, Central Hospital Manager).

Key informants reported that the pandemic highlighted the thousands of committed and responsive health workers, especially those on the frontline, who rose to the challenge of providing health care, in line with their ethical and professional obligations.

Clinicians from different departments organised themselves to work as a team and to cover departments that needed the most assistance. Psychologists provided debriefing for staff members (KI 8, Central Hospital Manager).

COVID-19 brought out the strength of each individual health worker. We have never experienced better teamwork than during this period… there was willingness to share resources and to work overtime (KI 9, Regional Hospital Manager).

In Gauteng, there was recognition of the need to ensure financial, human and other resources to meet the increased demand for hospitalisation and quarantine. The Gauteng Provincial Government allocated an additional R4 billion (~267 million US$; $1 = R15) to the health sector for the appointment of additional staff, the procurement of PPE, and the upgrading or building of additional infrastructure. Many key informants viewed this significant injection of finances in a positive light.

The response to COVID-19 was comprehensive. We were able to get additional finances. We also took the decision to revamp the current infrastructure, using alternative building technology (KI 4, Executive Manager).

The support was palpable on the ground…we were given R14 million (~$933 000) to beef up current staffing levels. The budget mitigated staff shortages especially with the opening of new wards (KI 8, Central Hospital Manager).

Key informants from the frontline expressed their appreciation for the innovative initiative to have a central warehouse and dedicated budget for PPE.

The PPE procurement was centralised, we fetched our PPE from the warehouse-the system was working well for us (KI 8, Central Hospital Manager).

Initially we had challenges [with PPE] but later, the district did their best to get us enough PPE, hence we had limited number of staff who were infected because we had enough PPE in my facility (KI 32, PHC Facility Manager).

Some of the key informants commented on the digital innovation in Gauteng, evidenced by the screening/tracing app, and the bed availability dashboard. Gauteng province also sponsored a consortium of modelling experts to identify potential hotspots, and to estimate the need for hospitalisation, critical care beds and resources.

Although public health measures such as handwashing, sanitisers, and mask wearing were national interventions, several key informants highlighted the widespread positive behavioural change and institutionalisation of these measures in the province.

Theme 2: Contestations and contradictions in the pandemic response.

The second emerging theme highlights key informants’ perspectives on the contestations and/or contradictions in the COVID-19 response in Gauteng.

Key informants from the frontline complained about a top-down approach to planning, with instructions issued, lack of or insufficient engagement with hospital managers, and little or no room for questioning decisions from the top. Those in the private health sector commented on the lack of engagement or discussion on contracts, with perceptions of rigidity and a heavy-handed, top down approach of the Gauteng Department of Health. In some instances, the highly centralised approach resulted in confusion or duplication of roles.

At the War Room [the centralised structure set up to coordinate and guide the COVID-19 response], there was duplication of roles, you would find three sets of teams doing the same thing. You would have a [provincial] public health specialist team analysing data, then you have a clinical health team in the department (of health) doing the same thing, and then the district people (health programmes) doing the same thing (KI 9, Regional Hospital Manager).

However, some informants were of the view that the province was at the mercy of instructions issued by the National Department of Health.

National including NICD [National Institute for Communicable Diseases], province, district–it was very unclear who did what. The War Room was good but it wasn’t clear at first what kinds of decision were taken there. This led to a certain paralysis in responding. For example, in early March 2020, we were contacted by a church group planning a big event in April, asking for advice. Province said they couldn’t advise, the NICD said they couldn’t–eventually the District had to give an answer. We emphasised that we had no authority to do so, but said that it would not be a good thing to go ahead with the event (KI 12, District Health Manager).

There were mixed perceptions on intergovernmental relations, and coordination and collaboration. Some key informants were of the opinion that COVID-19 facilitated improved collaboration across several levels, namely across provincial government departments, between the province and local government health departments, between the public and private health sectors, between tertiary hospitals and the surrounding clinics, and within hospitals, between management and staff.

In contrast, some key informants highlighted poor intergovernmental relations. One key informant reported that one of the municipalities only came on board in August 2020, while another said:

Intergovernmental relations were problematic-although the City owns clinics, the provincial chief director communicated directly with the clinics, and bypassed the executive director for health in the city. There was insufficient investment in good relationships, and at times, the province-local government politics played itself out (KI 1, Executive Manager).

Another key informant highlighted the fragmentation between national, provincial and local government, lack of or insufficient communication, and weak messaging.

The left hand does not know what the right hand was doing. There was a huge disjuncture between National and the Gauteng Department of Health, lack of trust, and the relationship has worsened. The national meetings were mostly tick box exercises (KI 6, Technical Support/ Researcher/ Academic).

All the key informants were familiar with the rationale and purpose of the Disaster Management Act, namely to ensure an integrated and coordinated response to the COVID-19 pandemic. Some highlighted the confusion created by the accompanying regulations, especially at the beginning of the lockdown, exacerbated by the numerous changes, the lack of a coordinating mechanism, and lack of user-friendly guidelines. Those key informants from the frontline complained about the rapidly changing regulations that were open to interpretation, and the impracticality of some of these regulations.

There were regulations about the numbers of people that have to be at work. In health, all employees are essential and we couldn’t keep to the 30% required by the regulations. The regulations stated that people above 60 and with co-morbidities, were supposed to stay at home. In health, these were the people with the wisdom to deal with certain complications that might arise due to COVID-19, such as highly skilled critical care nurses (KI 10, Central Hospital Manager).

Another key informant highlighted the problem of unclear guidelines.

We faced challenges with unclear guidelines. For example, when a facility has a COVID-19 case, it is not clear who should do the deep cleansing—the general assistants at the facility or the external service provider (KI 32, PHC Facility Manager).

Two private sector key informant also highlighted the difficulties of unclear guidelines or rapidly changing regulations.

The regulations were amended every 14 days, and it was problematic to adapt so quickly (KI 5, Private Sector Manager).

The guidelines were not for the entire nation, and were not applicable to the private health sector. There was no opportunity for private sector input. We reached out to National and assisted with the wording of the regulations for the prescribed minimum benefits, to make sure that people are covered for COVID-19 (KI 24, Private Sector Manager).

Although the provincial COVID-19 Strategic Response emphasised the involvement of different stakeholders in fighting the pandemic [45], some key informants criticised the lack of participation of communities, frontline health workers, and the private health sector in decision-making.

Notwithstanding the innovative provincial initiative on modelling COVID-19 cases and resource requirements [45], the key informant interviews revealed intense contestations on the quality of information, and the results of the modelling, especially the estimated number of critical care and general hospital beds.

There was panic at the beginning, with over-estimates of the actual number of cases. For example, the 23 000 active cases look very different to projections of more than 100 000 cases. The additional [financial] allocation to health was made following their modelling and projections. The [Gauteng] Department of Health initially requested an additional R14.9 billion (~ $1 billion) for PPE, oxygen and infrastructure…we allocated R4 billion (KI 3, Executive Manager).

The initial 2020 modelling exercise produced an over-estimate of the expected number of COVID-19 cases at the peak of the pandemic in Gauteng [45]. This in turn led to an over-estimate of critical care and general hospital beds in the province, which influenced subsequent decisions or strategies that aimed to address the gap between bed needs and bed availability. Although the provincial government included the private health sector in the original modelling, the subsequent planning excluded the bed availability in the private health sector [45]. Consequently, the provincial government planned to overcome the gap in the need for additional beds through repurposing of existing wards, decanting wards in existing facilities and preparing them for the use of COVID patients, creating additional capacity within the existing hospital platforms, and creating temporary capacity outside existing facilities through field hospitals [45].

Key informant interviews also revealed contestation on the decision-making and actual decision on field hospitals. Some key informants were of the opinion that the decision on field hospitals was taken by the national Ministers of Public Works (Infrastructure Development) and Health, and handed down to Gauteng province, with no consultation or input from anyone. However, the research team was informed that Gauteng decided to invest in its own public health infrastructure, in line with the 2030 Growing Gauteng Together plan of action. Hence, only one field hospital was erected in the South of Johannesburg.

Some of the key informants thought that the field hospitals in preparation for the surge demonstrated quick decision-making in response to the pandemic, and created additional needed capacity.

We were able to develop additional capacity, using alternative building technology, which will remain the legacy of the pandemic. These facilities will be put to alternative use, e.g. for medical oncology (KI 21, Executive Manager).

Others were of the opinion that there was lack of transparency and that the decision on field hospitals was both controversial and ill-advised.

Field hospitals were very controversial from the start, as the experience from many places shows that they were not used. They are very badly designed and criticism or requests for change were ignored. Lots of money went into these. The decision-making was not agile enough; it was too compartmentalised, and unable to change direction (KI 11, Central Hospital Clinician).

There was no information in the public domain about field hospitals, whether they were used to their capacity, how other hospitals were coping and managing COVID-19 numbers (KI 13, Civil Society).

The key informants also expressed mixed opinions on isolation or quarantine facilities. Some key informants were of the opinion that the setting up of the quarantine centres showed bold provincial leadership. Others pointed out that there was poor utilisation of quarantine facilities, because the amenities were basic and/or community members refused to leave their homes. This was exacerbated by the long period of quarantine (14 days) in the early lockdown stages.

Lastly, the interviews revealed contradictory perspectives on the suitability of communication strategies used in the province. Many key informants pointed to good high-level communication from the Premier and former MEC on COVID-19 as a public health crisis, and government prioritisation of the pandemic. There were regular updates on the strategy and COVID-19 updates in the mainstream media (television, radio, newspapers). They felt that the explicit support of national decisions by the Premier and MEC was critical in getting a unified response. The visits of politicians to health facilities and communication with staff also strengthened Gauteng’s response to COVID-19.

However, key informants from civil society were of the opinion that the Premier adopted the same model of communication as the President, and that this was inappropriate in Gauteng, as the small size of the province provides a platform for greater interaction with various stakeholders. Although there was recognition of the Premier’s leadership, individuals from civil society organisations lamented government’s non-responsiveness to their offers of support or collaboration.

Some senior government officials, individuals from civil society and the private health sector highlighted sub-optimal communication to communities.

The lack of involvement or briefing of SALGA (= South African Local Government Association), trade unions, SANCO (= South African National Civics Organisation)-I think it was a big error at provincial level. The communication with the community was very, very, very poor (KI 1, Executive Manager).

The communication was very sophisticated, but it did not filter to community level (KI 23, Civil Society).

I think the provincial government should use existing community forums to communicate key aspects of the pandemic, adopt a non-prescriptive approach and use the existing capacity to the maximum (KI 24, Private Sector Manager).

Theme 3: Fragilities, complexity and vulnerabilities of systems.

The document analysis revealed detailed plans on epidemiology and surveillance, case management, health infrastructure, laboratory services, emergency medical response and research [45]. The June 2020 report on PPE included statistics on the different types of PPE, existing stock, as well as stock ordered. Although the central warehousing and procurement of PPE were a Gauteng innovation, during the second half of 2020, widespread reports surfaced on allegations of corruption in the procurement of PPE [59, 60]. The subsequent referral of the PPE corruption to the Special Investigating Unit illustrated that greed appeared to trump ethical conduct and accountability [59].

Many key informants highlighted the reported PPE corruption scandal as a painful experience, which overshadowed the achievements and strengths of the COVID-19 response in the province. They were of the opinion that the alleged corruption was entirely preventable, and that the weaknesses of existing systems, with lack of checks and balances, created the perfect opportunity for corruption. They pointed out that a combination of centralised procurement, lack of transparency, lack of accountability, failure to prioritise societal interest above self-interest, non-compliance with existing legislation, and a failure to identify and anticipate risks contributed to the alleged corruption.

Several key informants pointed out that the COVID-19 pandemic exposed the fragilities, and in some instances, the failures of the health system. These fragilities were exacerbated by the single-minded focus on COVID-19, with its prioritisation resulting in a de facto health system lockdown, especially of the public health sector. This had many unintended negative consequences, including a reduction in access and/or provision of essential health services (e.g. maternal and child health services, HIV and tuberculosis, non-communicable disease care), scaled-down hospital services (e.g. no elective surgery), and loss to follow-up of some patients. Some key informants pointed out that the impact of COVID-19 prioritisation on adverse events e.g. maternal and/neonatal deaths is unknown.

Collateral damage to patient care was and is significant. The deaths of non-COVID patients probably surpassed those of COVID patients for the first 3 months although no formal system was set up to measure this. Many patients died without proper care while waiting for their test results (KI 11, Central Hospital Clinician).

There has been a 25–50% reduction in [childhood] vaccinations–this is recovering but there is a large unvaccinated cohort and we run the risk of a massive measles outbreak with increased mobility. There were major disruptions to HIV and tuberculosis treatment. We raised concerns about collateral damage very early on (KI 15, Technical Support/ Researcher/ Academic).

Some key informants pointed to the collateral damage in terms of the quality of care provided.

The quality of care was not the best—we had to erect tents. In the beginning, medications were not administered as needed because of limited personnel (KI 10, Central Hospital Manager).

The quality of care was also affected by the appointment of junior staff and/or the reliance on agency nurses.

Our effort to circumvent overcrowding made us to send the newly appointed staff to casualty. This was a serious problem—you find that there were 15–20 patients with three or four nurses who didn’t understand the functional mandate of a particular department. We discovered many clinical service gaps (KI 9, Regional Hospital Manager).

Several key informants were of the opinion that the perceived collateral damage was exacerbated by fearful and apparent contradictory messages to communities.

The document analysis revealed the establishment of a separate Public Health Stream within the health department that presented detailed reports to the Provincial Advisory Council, which in turn reported to the overall governance structure, namely the Provincial Coronavirus Command Council [45]. On paper, there were detailed strategies and procedures that dealt with epidemiology and surveillance, community mobilisation, education and advocacy, and case management and public health training [45].

However, key informant interviews demonstrated a lack of clarity or knowledge on the provincial COVID-19 testing strategy.

There was failure to share the Provincial Testing Strategy, especially on the employee testing. There was also non-compliance with testing guidelines e.g. random testing of communities without following the guidelines, leading to conflict with our laboratory team (KI 20, Technical Support/ Research/ Academic).

Frontline staff highlighted the challenges experienced with laboratory testing.

While we were managing the pandemic, we had a lot of areas that were not in sync, so this is an area we can improve e.g. the laboratory tests, I don’t know where the problem was, but we had a lot of instances where people would be reported as negative or positive, but there would be missing information—no addresses and contact numbers, so it made contact tracing difficult (KI 34, District Health Manager).

Both the document analysis [61] and key informant interviews pointed to the vulnerabilities of health information systems. Key informants highlighted staff shortages, lack of capacity to analyse and/or interpret epidemiological trends, initial reluctance to participate in the national hospital surveillance system, and the fragmentation of the existing information systems, some of which remained paper-based. They pointed to poor data quality, which meant that sometimes the information provided did not make sense. The consequences of poor health information systems were the inability to plan pro-actively, use evidence to inform decision-making, respond quickly, and/or to implement appropriate action. These problems necessitated the setting up of new or parallel systems, creating a disjuncture between the data from the NICD and the provincial health information system, duplication of efforts, and lack of reliable information to strengthen the pandemic response.

Theme 4: Under-investment in the health workforce.

Although many key informants expressed appreciation for the additional funding and staff appointments as part of the COVID-19 response, a recurring theme in the interviews was the perceived under-investment and insufficient focus on the health workforce, especially on frontline staff. Some key informants were of the opinion that there was lack of prioritisation of human resources and that the health department did not demonstrate appreciation of staff, despite health workers risking their lives to do their duties.

Doctors are burnt out, nothing you do really is recognised, morale is low, and there is poor support, lack of equipment, unfilled trolleys, staff shortages, having to run around for stock …it a very draining emotional experience (KI 36, Central Hospital Clinician).

Health care workers contracted COVID 19 and some lost their lives. We should have given a little more attention to frontline health workers e.g. buses provided for travel so that they don’t have to contract COVID, more emotional (psychological) support (KI 32, PHC Facility Manager).

They lamented insufficient involvement of frontline staff in the pandemic response, or in certain decisions, such as putting up separate tents for COVID-19 positive patients.

We should have been involved in discussions around how oxygen will be supplied to the tent, and that is a problem….sometimes you run out of oxygen during a resuscitation (KI 36, Central Hospital Clinician).

Poor or sub-optimal human resource management exacerbated the perceived under-investment in the health workforce.

COVID-19 exposed poor managers who often took out their frustration on ordinary health workers. This was in contrast to those managers who were able to respond to pressure (KI 22, Professional Association).

Some of the key informants commented on the proactive response in Gauteng to deal with the serious threat of the COVID-19 pandemic, reflected in a dedicated focus on infection prevention and control (IPC) and occupational health services. Although the COVID-19 pandemic underscored the importance of IPC principles (e.g. handwashing, IPC training and infrastructure), which have been neglected historically, several key informants pointed out that the large number of infections among health workers reflected weaknesses in IPC.

We did not prepare the health care workers enough; most of the infections were on the so-called green side [non-COVID wards]. In accident and emergency units, there were infections caused by insufficient or poor quality PPEs, but utilisation [of PPE] was also a challenge (KI 4, Executive Manager).

Central hospital ABC was actually overwhelmed, and ran out of space for patients. There were mini-outbreaks happening in the wards. The infection rates among health care providers are an indictment on preparedness and PPE provision. For example, 55% of health workers in the Internal Medicine wards housing COVID or persons under investigation became infected over a period of 6 weeks. This happened due to poor IPC practices, but also poor PPE (KI 15, Technical Support/ Researcher/ Academic).

The comments from key informants on COVID-19 infections among health workers were borne out by an August 2020 Ministry of Health media statement [62]. By 4 August 2020, there were 27 360 COVID-19 cases among health workers, the majority (78%) were from the public health sector, while 22% were from the private sector [62]. At the time, the overall infection rate amongst health workers was 5% [62]. The media statement also reported 240 COVID-19 deaths among health workers, the majority (84.6% or 203) from the public health sector [62].

Several key informants pointed out that the provincial health department failed to deal with the fears and anxieties of frontline health workers.

We thought that we are strong, but we were all paralysed with fear in spite of infection prevention and control training. In June/July 2020 when the surge was really high, even though we had a guideline on which protective PPE should be worn, everyone wanted to wear the triple layer (disposable scrub suit, gowns, coverall, plastic aprons and N95 masks) of PPE even in non-COVID wards because of fear and not because of scientific evidence (KI 8, Central Hospital Manager).

This issue of staff fear wasn’t recognised- this was an unintended consequence of the communication campaign to try and get the population to follow guidance but had the effect of creating fear and panic amongst staff that this [COVID-19] was like Ebola- that they would die. The perception that there was a high mortality associated with [COVID-19] infection was never undone, even when reality was showing something different (KI 11, Central Hospital Clinician).

The majority of key informants in both the public and private health sectors highlighted that there was no clear strategy for employee assistance or psychological support. They noted that there was insufficient acknowledgement of the fears and anxieties of frontline health workers. A senior executive in the private health sector noted the following:

My sense is that staffing was a kneejerk reaction, rather than planned. Staffing could have been managed better, including sharing of human resources between the public and private health sectors. The two sectors operated separately. At our hospital group, we realised later on in the pandemic that the health workforce is an absolute priority (KI 5, Private Sector Manager).

Another private sector key informant noted that the early preparation of staff should have included psychological support, as there was a lot of uncertainty and the impact of COVID-19 on staff was massive.

Theme 5: Missed opportunities.

The interviews with key informants revealed that there were various missed or wasted opportunities. The Comprehensive Health Response was a hospital-based strategy primarily, with the Ward-Based Battle Plan developed much later, seemingly after COVID-19 infections had peaked in the province. Some key informants criticized the hospi-centric pandemic response rather than primary health care and/or community-based response.

We need to move away from a curative system and invest in primary health care (PHC). You cannot talk about PHC and continue to invest in hospitals. COVID-19 was a missed opportunity to optimise and focus on PHC (KI 18, Civil Society).

There were also perceptions of missed opportunities to engage and partner with communities and/or civil society. Several key informants pointed out that the COVID-19 response did not seem to take into account the vulnerability of communities in informal settlements, who did not have clean running water and adequate sanitation. They were of the opinion that the hard lockdown regulations of social distancing, confinement to residences and hand hygiene in informal settlements were almost impossible to adhere to. Although there was recognition of the Premier’s leadership, individuals from civil society organisations lamented government’s non-responsiveness to their offers of support, or collaboration.

There was a lot of goodwill on the part of civil society, but it does not have the power or the resources to intervene. Government did not take advantage of civil society or capitalise on the experience of extensive community involvement in combatting the HIV pandemic. To a large extent, the response has been biomedical and bureaucratic (KI 23, Civil Society).

Some of the key informants highlighted the missed opportunity of government in regulating the private health sector to enhance equity, and to ensure access to health care and to scarce resources, such as ventilators.

We had a two-sector response to the pandemic, contributing to inequity. There was a different turn-around time based on where you tested e.g. in private sector quick access, but this was not the case in the public sector. We did not have regulations on the distribution of resources e.g. ventilators between the public and private sectors (KI 18, Civil Society).

These key informant perspectives on the missed opportunity to strive for greater equity between the private and public health sectors were supported by the document analysis [41, 49]. Although the private health sector provides care to less than 20% of the South African population, it accounted for the majority of COVID-19 tests during the first wave of infections [49]. Evidence suggests that there is a lower threshold for admission to private hospitals, thus exacerbating inequities in access to COVID-19 treatment and care [41].

In contrast, those key informants from the private sector felt that they had been excluded, and this was both frustrating, and a missed opportunity for collaboration.

The biggest disappointment is that they [public health sector) thought they could do it themselves. There was a major missed opportunity to engage the private health sector in service provision. The guidelines were not for the entire nation, and were not applicable to the private health sector (KI 24, Private Sector Manager)

I have tried through three parallel pathways to get the private sector involved…it has been difficult to know how to assist. There has been no engagement or discussion [in Gauteng] on contracts with private medical practitioners (KI 26, Professional Association).