Provincial assessment of OSH for health workers' readiness for SARS-CoV-2
Four PDoHs—representing 170,686 health workers—participated in the OSH readiness assessment. The OSH directorate represented provinces A and B. The Employee Health and Wellness directorate and the Department of Public Health Medicine represented Provinces C and D, respectively. By June 12, 2020, the four PDoHs had 3,675 SARS-CoV-2 cumulative cases among health workers (2,670 in Province A, 599 in Province B, 207 in Province C, and 199 in Province D).
Provincial Occupational Health Services readiness for the protection of health workers
At the time of data collection, all four provinces had a provincial SARS-CoV-2 plan from which they were working, albeit not specific for OSH for health workers. In addition, provinces A and C had an IPC policy specific for SARS-CoV-2 in the health sector. Only Provinces A and D had OSH SARS-CoV-2 provincial coordinating teams and a dedicated budget for OSH.
At the PDoH level, only Province A had an occupational medical practitioner. Provinces A and B had an occupational health nurse. All PDoHs had an IPC manager and an environmental health practitioner, except for Province C, which lacked the latter. Only Province A had an occupational hygienist, and only Province C lacked employee health and wellness professionals. Statutory laws in South Africa require employers to have health and safety representatives and committees, but only Province A and D had functional committees.
At the time of data collection, all PDoHs had provided SARS-CoV-2 training to a proportion of their health workforce. Province A provided the seasonal influenza vaccine to its health workers. Province D had conducted some health risk assessments specific to SARS-CoV-2. Provinces A and D reported having rehabilitation plans for infected and affected health workers and regularly screen them looking for incident cases. All Provinces but Province B provided treatment and mental health services. However, provinces B and C did not have adequate personal protective equipment.
All four PDoH reported were collecting SARS-CoV-2 data from their health facilities using a tool (these differed in all the PDoHs). The PDoH reasons for collecting data included reporting to National Government Authorities, including the Department of Health and Department of Employment and Labour, and the Department of Public Service and Administration, the employer of public servants in South Africa. Except for Province D, all the PDoHs indicated that they had a server to store the data, and none of the PDoH had a monitoring and evaluation plan.
Health facility assessment based on principles of the HealthWISE tool
Health facilities and participants in the HealthWISE walkthrough assessment
Forty-five health facilities representing 34,192 health workers (i.e. 20% of the health workforce in the four PDoH studied) participated in the HealthWISE guided health facility SARS-CoV-2 assessment. The health workers who were trained and participated in the HealthWISE walkthrough assessment included: occupational medical practitioners, occupational health nurses, environmental health practitioners, IPC nurses, quality assurance nurses, and hospital managers (medical, nursing managers and finance). In addition, the sample included employee health and wellness practitioners and trade unions. Participation of occupational health doctors, occupational health nurses, EHPs, and IPC nurses depended on whether the hospital had such professionals on their staff establishment. As such, we used this indicator as a proxy for the availability of OSH services in that particular health facility.
HealthWISE walkthrough assessment findings for all facilities
The main entrance to the health facilities is the first contact between patients and the health facilities. It has a pedestrian gate or door-like structure and boom gate for vehicles which is manned by security guards. The walkthrough (Table 1) observed that the opportunity for close contact between security personnel and pedestrian patients as well as security personnel and drivers of vehicles was highly probable. Both the pedestrians and drivers had to sign in a security booklet; during this period, the opportunity of cross-infection increased due to the proximity of individuals, exchange of pens and booklets, and poor attention to IPC protocols.
While ventilation in the main entrance was adequate for health workers in the outdoors security area, security guard houses were often small and crowded and lacked windows or any other form of ventilation. During the walkthrough, a lack of administrative control was immediately apparent, as there was generally a lack of posters and signs. There were crowded chairs indicative of no social distancing, even in the absence of health workers in the area. The majority of security guards were wearing cloth masks but had them below their noses.
The accident and emergency areas in most health facilities were high-risk with the twenty-four-hour operation. We display the findings of walking through this area in Table 2. Ventilation was a considerable challenge in the majority of assessed health facilities. Natural ventilation was impaired by mostly poor facility design, while mechanical ventilation was not available in many facilities and when available was broken and had a poor maintenance record. In the health facilities assessed early on in the study, we noticed that administrative controls would only be followed and monitored during working hours. Afterhours, there is no extra staff assisting with the enforcement of IPC measures, leading to a lack of implementation and monitoring of IPC measures for evening and night shifts.
In the accident and emergency department, the social distancing of staff was a great challenge due to insufficient space and high patient load. In some of the health facilities assessed, we found that they had or planned to have makeshift working areas using tents in some instances. This arrangement led to overcrowding, more pronounced in urban facilities. Health facilities in rural areas had high patient volumes in the mornings and almost no patients in the afternoons, allowing for social distancing. Health workers across facilities were often found congregating when engaging in administrative work or not attending to patients. There were many facilities with small or no rest areas leading to health workers crowding the administrative areas for their paperwork and resting periods. It was, however, concerning that there were very few or no SARS-CoV-2 posters or educational materials in most facilities. Waste management was generally good in the majority of assessed health facilities. The majority of the health workers were wearing the appropriate PPE, and most were wearing it correctly. In a few hospitals, health workers would come to work in their uniform, work with it and return home wearing it.
The findings for the outpatient departments in the assessed health facilities (Table 3) were similar to the accident and emergency department walkthrough. About half of health facilities assessed had some combination of natural and mechanical ventilation, with Province A facilities being most affected by poor ventilation. Social distancing was adhered to by the majority of the patients. Most health facilities aided this by marking brightly and visibly cancelling out some chairs or seating areas and rendering them unavailable for use. Patients standing in queues were assisted with markings of distances of between 1.5 meters and above to maintain social distance. Health workers had no or low access to often small dining and rest areas, leading to poor social distancing. Several health facilities had rearranged the furniture to allow for social distancing. A notable issue was the lack of posters and information on how health workers can protect themselves from being infected with SARS-CoV-2. There was access to water, soap and sanitizers for hand hygiene and proper management of waste from SARS-CoV-2 contaminated materials. On observation, workers were wearing PPE though it varied with a few wearing cloth masks and surgical masks and many using respirators, particularly N95 respirators.
The wards dedicated to suspected and confirmed SARS-CoV-2 patients varied considerably within and between provinces (Table 4). Some facilities did not have these wards or were still in the preparatory stages for these wards (i.e. renovations were taking place). Ventilation in these wards was mechanized and working. However, in Province A, there were a few health facilities with poor ventilation with non-functioning mechanical ventilation. In almost all health facilities assessed, the suspected and confirmed SARS-CoV-2 patients were socially distancing or kept in physically separated rooms. Health workers were social distancing in the main, except in a few health facilities in Province A, where we found health workers seated close to the main administration desk area when not attending to patients. Most of the health facilities had set up the furniture to allow for social distancing.
The posters in these wards were not specific for OSH or the protection of health workers, and their placement in the wards was not ideal for easy access by health workers. The facilities had water, soap and sanitizers except for one health facility with no water due to municipality failures for the whole town. Sanitizers, while present, were not strategically placed for easy access by health workers. Waste management was adequate, and health workers were able to discard with ease all contaminated materials. Health workers had access and were wearing PPE correctly with the exception of a few health facilities in Province A. Some health facilities in Province C displayed some elements of best practice as they had a suspect and confirmed SARS-CoV-2 patient ward separate from the rest of the hospital building, and the health workers for that ward were isolated from the rest of the hospital, stayed in accommodation provided by the hospital, and utilized hospital clothing and PPE during their shifts.
Table 5 presents the individual HealthWISE scores per health facility. The lowest was 14/88, while the highest was 73/88 (median 42, IQR 27–53.5), administrative scores were from 6/56–50/56 (median 28, IQR 18–36), ventilation scores were from 0/16–11/16 (median 4, IQR 2–6), and PPE scores were from 3/16–15/16 (median 9, IQR 7–11). None of the health facilities met the criteria for acceptable HealthWISE score, and 42% had an unacceptable HealthWISE score of below 50%.
The results of the logistic regression analysis assessing the availability of a provincial policy and hospital preparedness indicators showed that in the adjusted model, PPE and ventilation score were statistically associated with the availability of a COVID-19 provincial policy (Table 6). The presence or implementation of the provincial policy was associated with an increase in PPE score.
Hospital staff infection rate and compliance score
From March 5, 2020 to June 15, 2020, the cumulative hospital infection rate for COVID-19 at the 45 facilities ranged from 0–17.9%, with a median 0.3% infection rate (IQR: 0.0-0.9%). We applied nonparametric statistical tests because the data seemed non-normal (Shapiro-Wilk test : p < 0.001). We further adjusted by the total number of employees given that plotting infection rate against compliance score showed there might be some clustering by province (Fig. 1).
We classified each hospital into two categories: lower compliance or higher compliance, using the median score (42) to determine categories (Fig. 2). The lower compliance group (with a score < 42) had a median infection rate of 0.75%, and the higher compliance group (with score ≥ 42) had a median infection rate of 0.1%. While the infection rate is low for both groups, the Mann-Whitney U test suggested a significant difference between the medians of the two groups. Our adjusted Poisson Regression estimated an incidence rate ratio of 0.98 (95% Confidence Interval: 0.97, 0.98) for higher compliance vs. lower compliance.