INTRODUCTION
The outbreak of a novel coronavirus disease 2019 SARS-CoV-2 (COVID-19) was recognised as a global pandemic on 11-Mar-2020 by the World Health Organisation (1). Previous experience with severe acute respiratory syndrome (SARS) and Ebola has highlighted the risk of infection and death for front line health care workers (HCWs) who experience workplace exposure to infectious diseases (2). During periods of infectious disease epidemics, implementation of infection prevention and control (IPC) is of great importance in healthcare settings to mitigate the risk of HCWs becoming infected (2).
The World Health Organisation (WHO) has recommended that HCWs use appropriate personal protective equipment (PPE) when assessing and managing patients with suspected and confirmed COVID-19 infection (3) and the WHO guidelines have been adopted by the National Institute of Communicable Diseases (NICD) for our local setting (4).
SARS-CoV-2 can be spread during routine care via respiratory droplets produced by sneezing or coughing which are inhaled person to person, or via respiratory droplets landing on surfaces which are then transferred by contact via contaminated hands to a person's face and mucous membranes. Aerosol generating procedures (AGPs) such as nasopharyngeal swabbing, intubation, extubation, open suctioning of the intubated patient may confer transmission via aerosolization (4).
The recommended use of PPE during the study period was for the use of a surgical mask, face shield and gloves to be worn by all clinical staff during routine care of suspected or confirmed COVID-19 patients. Non-clinical staff were also advised to wear a surgical mask and gloves, but the requirement of a face shield fell away. When performing AGPs an N95 mask, face shield, gloves, and apron were advised (4). Usually PPE is discarded after a single patient or procedure, however, due to an acute shortage of PPE during the first wave of the COVID-19 outbreak worldwide, the WHO and Centre of Disease Control considered extended use and/or reuse of certain types of PPE. The NICD noted that an N95 mask could be used for up to a week (extended use) unless the respiratory integrity or leak proof seal was compromised (4).
Despite PPE recommendations, HCWs worldwide have been infected by SARS-CoV-2. In some instances, shortages of PPE and lack of provision of training for IPC have been implicated. It is also important to note the potential risk of transmission between HCWs when they are not caring for patients such as during clinical case discussions, clinical handovers and lunch breaks and potential risk of transmission unrelated to the health care environment (5).
The clinical spectrum of COVID-19 infection varies from asymptomatic or mildly symptomatic infections to severe respiratory symptoms and death (6). Several studies have described the rate of HCW infections with COVID-19 and described the characteristics of both the HCWs and their clinical course of illness. Several studies have highlighted potential PPE shortages as a risk factor for COVID-19 infection in HCWs (7–10). However, few studies have examined this in an African or South African setting.
There is a paucity of data regarding the risk of COVID-19 infection in HCW in South Africa. Thus, this study aimed to study the prevalence of COVID-19 and attendant risk factors in HCW in an Emergency Department (ED) of a large public hospital in South Africa.
METHODS
The study was conducted at Helen Joseph Hospital, an urban, public hospital in Johannesburg, South Africa. The ED of Helen Joseph Hospital attends to approximately 60000 patients annually.
This prospective, qualitative, cross- sectional, observational study was undertaken by means of a self-administered, anonymous voluntary survey distributed to all HCWs known to have contracted COVID-19 working in the ED over a 6-month period from 1st April 2020 to 30 September 2020.
All HCWs were screened (temperature, symptoms of Covid or recent Covid-19 contact) daily upon arrival to the hospital by nurses stationed at the front entrance. Screening was uniform for all HCWs, whether they were clinical or non-clinical. HCWs were required to undergo COVID-19 PCR testing if they met positive screening criteria. On return to work following a positive COVID-19 PCR test and isolation period, HCWs were requested to complete a paper based self-administered voluntary survey. (Supplementary Appendix A) No routine weekly COVID-19 PCR testing was undertaken.
Data collected included: HCW demographics; symptoms that prompted HCW to test for COVID-19; clinical course of illness; risk exposure of the HCW over the 7 days preceding their illness (non-clinical, clinical undertaking routine management only or clinical undertaking both routine management and AGPs); and whether PPE shortages or any other perceived contributing factors were experienced by the HCW in the 7 days preceding their illness. Perceived contributing factors included ED overcrowding, access block, water and hand sanitiser outages. ED overcrowding was defined by over 10 respiratory patients awaiting ED review for > 1 hour resulting in congestion and a lack of space and staffing to meet the needs of patients in a timely manner. Access block to general medical / COVID-19 wards was defined as over 10 respiratory patients awaiting admission to the general medical / COVID-19 wards for > 8 hours; access block to critical care was defined as any patient awaiting admission to critical care / ICU either intubated, on non-invasive ventilation (NIV) or high flow nasal cannula (HFNC) spending over 6 hours in the ED awaiting admission. Water outage was defined as a lack of running water in the ED (including no running water to wash hands or to flush toilets) for > 2 hours; hand sanitiser outage was defined as a lack of hand sanitiser availability at a workstation for > 2 hours.
Statistical analysis was performed using SPSS V25 (Statistical Packages for the Social Sciences) IBM ®. The profile of HCWs, reported symptoms and PPE shortages were computed in frequencies and percentages. The chi-square test of association compared PPE shortages and disease severity by HCWs’ profiles and risk exposure. The significance and strength of the relationship between the number of patients and HCWs testing positive for COVID-19 was tested using the Spearman rank correlation. Significance testing was set at the 95% confidence level.
Ethical approval for the study was obtained from the University of the Witwatersrand Human Research Ethics Committee
RESULTS
A total of 201 HCWs were based in the ED during the study period, of which 58 were doctors (30 full time and 28 part time), 84 were nurses (all full time) and 59 were non-clinical inclusive of porters (8 full time), cleaners (4 full time), ECG technicians (3 full time) and radiographers (44 part time split between the main radiology department and the ED radiology department). Forty-nine (24%) developed symptomatic confirmed COVID-19 within the six-month study period and all completed the anonymous voluntary survey.
The majority of HCWs that developed symptomatic confirmed COVID-19 were aged below 35 years (n=26; 53%) with a range from 21 years of age to 56 years of age (median 34 years of age [IQR = 27-34.5]). Most HCWs with symptomatic confirmed COVID-19 were female (n = 36; 73%).
Most symptomatic HCWs with confirmed COVID-19 infection reported no comorbidities (n=37; 76%). Of the 12 (24%) HCWs that did report comorbidities: 7 reported that they had underlying hypertension (14% of the total 49 HCWs with symptomatic confirmed COVID-19), 4 had asthma (8%) and one each (2%) had diabetes (in addition to hypertension), arthritis and hypothyroidism (in addition to asthma). Body mass index (BMI) was calculated for 39 of the 49 HCWs and ranged from 18.3 to 40.8 (median BMI of 27.9 [IQR 9.6]). Twenty-six (67%) of the HCWs with a calculated BMI had a BMI > 25. Ten respondents omitted their weight on the voluntary survey.
The profile of the study group is summarised in Table 1 and a breakdown of HCW role in the ED is highlighted in Table 2.
Symptomatic COVID-19 infections within the Emergency Department - patient profile
| Age Group | n = 49 | Co-morbidities | 12 (n = 49) | BMI | n = 39 |
|---|---|---|---|---|---|
| 20 - 34 | 29 | Diabetes | 1 | < 18.5 | 1 |
| 35+ | 23 | Hypertension | 7 | 18.5 – 24.9 | 12 |
| Sex | Asthma | 4 | 25 – 29.9 | 11 | |
| Male | 13 | Arthritis | 1 | 30 – 34.9 | 10 |
| Female | 36 | Hypothyroidism | 1 | 35 – 39.9 | 3 |
| >40 | 2 |
Symptomatic COVID-19 infections within the Emergency Department by role
| N = 49 (%) | Total number of staff of specified role in ED | |
|---|---|---|
| Nurse | 24 (49%) | 84 |
| Doctor | 17 (35%) | 58 |
| Non clinical staff | 8 (16%) | 60 |
The most common symptom experienced by HCWs at the time of testing was fatigue (n=26; 53%). Table 3 demonstrates the wide range of symptoms experienced by symptomatic HCWs at the time that they tested positive for COVID-19.
Symptoms present in HCWs with COVID-19 upon testing
| Frequency (n = 49) | Percentage | |
|---|---|---|
| Fatigue | 26 | 53% |
| Fever | 21 | 43% |
| Sore throat | 20 | 41% |
| Cough | 18 | 37% |
| Myalgia | 14 | 29% |
| Nasal congestion | 14 | 29% |
| Shortness of Breath | 12 | 24% |
| Diarrhoea | 5 | 10% |
| Anosmia / Ageusia | 4 | 8% |
| Chest pain | 4 | 8% |
| Nausea | 2 | 4% |
| Headache | 1 | 2% |
| Body pains | 1 | 2% |
| Severe backpain | 1 | 2% |
None of the HCWs that tested positive for COVID-19 were asymptomatic. The majority of symptomatic HCWs with COVID-19 were mild (n=30; 61%) with a minority either severe (NICD criteria), but not requiring hospitalisation (n=14; 29%) or severe requiring hospitalisation (n=5; 10%). The mean length of stay for HCWs that required admission was 3.1 days (SD 1.0).
There were no significant differences in the severity of illness related to different age groups (20 – 34 years of age; 35+ years of age). Symptomatic COVID-19 positive female HCWs suffered worse severity of illness compared to their male counterparts. Of the total number of HCWs with severe illness (n = 19), 17 (89% 95%) were female and 2 (11%) were male giving a prevalence rate of severe COVID-19 infection amongst symptomatic COVID-19 positive HCWs of 47% in females (95% CI [31; 63]) vs 15% in males (95% CI [-4; 34]) p = 0.04. Whilst overall severity did not correlate with comorbidities, severe COVID-19 infection that required hospitalisation did have a statistically significant correlation. Five of the six HCWs that required hospital admission reported an underlying comorbidity (83%) p = <0.01. The remaining 7 HCWs with comorbidities had severe not requiring hospitalisation (n = 2, 29%) and mild (n = 5, 71%) illness respectively. Thus, symptomatic COVID-19 positive HCWs with comorbidities were noted to have a prevalence rate of 42% (95% CI [14;70]) for severe COVID-19 requiring hospitalisation, p = <0.01, which was statistically significant. Increased BMI > 25 had no statistically significant relationship with severity.
Of the HCWs that contracted COVID-19 over the study period, 8 (16%) undertook non-clinical duties in the 7 days preceding the onset of illness, 20 (41%) undertook clinical duties that did not involve aerosolising procedures and a further 20 (41%) undertook clinical duties that involved aerosolising procedures in the 7 days preceding the onset of illness. Data is missing for one HCW on the anonymous survey form. There was no statistically significant correlation between clinical duties performed and severity of illness.
Twenty-nine (62%) of HCWs reported that they had experienced a shortage of PPE in the 7 days preceding their onset of illness. Several other possible contributing factors experienced by HCW in the 7 days preceding onset of illness were noted and have been documented in Table 4 below with the majority of HCWs reporting experiencing at least one of these potential contributing factors in the 7 days preceding the onset of illness.
Reported contributing factors other than PPE shortages
| Frequency | Percentage | |
|---|---|---|
| ED overcrowding | 43 | 88% |
| Access block - general | 19 | 39% |
| Access block – critical care | 18 | 37% |
| Water outages | 6 | 12% |
| Hand sanitiser shortages | 3 | 6% |
There was a statistically significant (p=0.03) and medium strength (rho=0.60) positive correlation between the number of patients that tested positive for COVID-19 in the hospital and the number of ED HCWs that tested positive for COVID-19 on a weekly basis (Figure 1).
DISCUSSION
Health care workers in Africa are familiar with occupational exposure to many pathogens, however, SARS-CoV-2 currently poses an arguably greater threat to HCWs than any other infectious agent to date (11). Studies suggest that HCWs have a COVID-19 infection rate of up to 11% which is higher than in the general population (12) with studies from the UK and USA demonstrating that frontline HCWs were 3.4 times more likely than the general population to test positive for COVID-19, after adjusting for testing patterns (13). The current study demonstrated that 24% of the HCWs based in the ED developed symptomatic COVID-19 over the study period. Given that community transmission and the role of social interactions is likely amongst HCW's, we are unable to infer that all HCW infections are directly as a result of their exposure within the ED. Furthermore, this study only identified symptomatic COVID-19 infections and thus it is likely to be an underrepresentation of the true COVID-19 infection rate amongst HCWs.
Seventy-three percent of confirmed COVID-19 cases in our study sample were in female HCWs, which reflects the findings by the WHO, that women form 70% of the health workforce globally (14). Thus COVID-19 infections in our study mirror gender patterns present in the global health workforce and is in keeping with international data on HCWs demonstrating COVID-19 infections in predominantly young, healthy HCWs (15).
Fatigue was the most commonly experienced symptom amongst COVID-19 HCWs at the time of testing, which is different from the classical NICD suspected COVID-19 case definition of cough, sore throat, shortness of breath, anosmia, dysgeusia; with or without other symptoms (fever, weakness, myalgia or diarrhoea). (6) It is recognised that a case definition is a trade-off between sensitivity and specificity.
The WHO estimates that approximately 80% of patients with COVID-19 will have asymptomatic or mild infection with 10 – 15 % of patients developing severe disease and a further 5% becoming critically ill (16). In the current study, 61% had mild infection. It was noted that there was a statistically significant increase in prevalence rate of severe COVID-19 amongst symptomatic COVID-19 positive female HCWs as compared to symptomatic COVID-19 positive male HCWs. However, this should be interpreted with caution because of the small total sample size of 49 HCWs and wide confidence intervals. Evidence from China, Europe and the USA have shown a consistently higher risk of severe disease in patients with underlying health conditions (17). We found that 83% of HCWs with symptomatic COVID-19 with severe disease requiring hospitalisation had at least one comorbidity but the confidence intervals were wide. Although obesity is a recognised risk factor for poor outcome in COVID-19 (18), increased BMI in our study did not correlate with a higher risk of severe disease. Our data is confounded by 20.4% of the study participants omitting their weight on the anonymous voluntary survey.
Of interest is that in the current study HCWs in contact with patients with COVID-19 had higher rates of COVID-19 infection than HCWs with non-clinical duties. A relationship between the risk exposure and infecting dose and the risk of disease severity, has been demonstrated for several viral infections, including influenza and SARS (19,20). Thus we had anticipated that rates of infection and disease severity might be higher amongst HCWs involved in aerosolising procedures, but this was not borne out in our results.
Sixty percent of HCWs reported that they experienced a shortage of PPE in the 7 days preceding onset of illness. However, many HCWs reported that they experienced a shortage of PPE that they were not required to wear for the clinical duties that they performed as based on NICD recommendations (4). This is particularly noticeable with regards to the perceived shortage of N95 masks by non-clinical staff and clinical staff undertaking routine clinical duties other than aerosolising procedures. Despite extensive IPC training on appropriate PPE usage, many HCWs felt that they should have been issued with PPE that was not designated for their level of risk exposure. Hospital PPE stock levels were reviewed daily and reported by means of a multidisciplinary meeting involving IPC and the ED amongst other relevant stakeholders. At no stage during the study period were actual PPE shortages experienced. Our study does demonstrate that the perceived shortage of PPE by HCWs is significant, the implications of which are complex and IPC training may need to be reviewed.
A study undertaken in both the UK and USA has demonstrated that, compared to HCWs who reported adequate PPE, front line HCWs that reported the necessity to reuse PPE had an increased risk of a positive COVID-19 test (13). The reuse or extended use of PPE was considered standard practice during our study period. Hence it is difficult to ascertain what role, if any, the reuse or extended use of PPE had on rates on infection amongst HCWs in our study as we did not have a comparative study group that did not reuse PPE.
Several other possible contributory factors were noted by HCWs in the 7 days preceding their onset of illness. Overcrowding and access block were defined in our study as the level at which ED functioning would be impeded, exceeding the physical and staffing capacity of the department (21). If the ED experiences overcrowding and access block, patients and staff cannot be protected (22). A significant number of respondents (88%) in the current study reported overcrowding and 37% reported access block as contributory factors to their risk of developing COVID-19. However, the study was not designed to corroborate these findings but future studies need to examine this factor more critically.
Risks of infection in HCWs is often tied to hand hygiene, amongst other factors (23). Water supplies for hand washing, however, may be limited or unavailable in some parts of Africa. Our study was undertaken in an urban hospital in Johannesburg which suffered a severe water outage in July 2020 lasting for 7 days (7 -13 July 2020 inclusive). During this time HCWs were unable to wash their hands and had to rely on the use of hand sanitiser. It is noted that the following week the highest volume of HCW COVID-19 infections were recorded. Whilst we cannot confirm on this being causative it could be considered a potential contributory factor. Hand sanitiser shortages were rarely experienced by HCWs during the study period.
LIMITATIONS
This study is a single centre study, with a small sample size which limits its generalisability. As the survey was distributed to HCWs on their return from isolation post COVID-19, the time delay between initial diagnosis and survey completion may have resulted in recall bias. The proportion of positive tests reported in this study possibly does not reflect the true prevalence of infections amongst HCWs as asymptomatic infections are not reflected in our data.
CONCLUSIONS
This study demonstrated a high rate of symptomatic COVID-19 infection amongst HCWs in the ED as compared to the global general population. Most infections occurred in young, healthy, female HCWs and were predominantly mild in nature in keeping with international data on COVID-19 HCW infections. However, severe disease requiring hospitalisation had a significant association with HCWs with underlying comorbidities. Overcrowding and access block were considered contributory factors by many HCWs, as has been the experience globally. Those HCWs undertaking clinical duties were more likely than those not undertaking clinical duties to contract symptomatic COVID-19. As there was a significant perceived rather than actual shortage of PPE, further IPC training and HCW engagement may be required.
