1 INTRODUCTION
The front‐line health‐care workers (HCWs) are at a significant risk of contracting severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) infection as compared to the general population.1 With the advancement of the coronavirus disease 2019 (COVID‐19) pandemic, several countries have now reported COVID‐19 in HCWs.2–6 Earlier, we reported co‐infection of vector‐borne diseases (malaria and dengue) in pregnant women with COVID‐19 in India.7 However, currently, information on co‐infection of SARS‐CoV‐2 infection with malaria or dengue in HCWs is completely lacking. This information is crucial for the countries having higher burden of malaria and dengue as well as COVID‐19. The information on virus clearance will help the hospital authorities to provide necessary assistance of self‐protection to the HCWs. The information will be useful to health policymakers for development of policies to mitigate infection in hospital setting. The objective of the study was to estimate the prevalence, understand demographic characteristics and clinical presentations among HCWs with COVID‐19 and compare the viral clearance between symptomatic and asymptomatic SARS‐CoV‐2 infection. We also aimed to study the impact of malaria co‐infection on the virus clearance in HCWs with COVID‐19.
2 METHODS
The study was approved by the Institutional Ethics Committee of TNMC and BYL Nair Ch. Hospital (NH), Mumbai. NH is a dedicated COVID‐19 hospital8 since April 18th with 1043 beds and more than seven thousand patients have been treated till October 31st. The data was captured from medical case records and the telephonic interviews of HCWs as described elsewhere.9 The information on COVID‐19 exposure was collected on telephonic interviews. All HCWs gave informed consent and agreed for the anonymous use of their medical data. The study is registered with clinical trial registry of India (CTRI/2020/09/027516). We included HCWs with reverse transcription‐plymerase chain reaction (RT‐PCR) confirmed diagnosis of COVID‐19 in the study and the data was collected from April 6th–October 31st, 2020. The disease severity was classified as mild, moderate, and severe as per the National clinical management protocol COVID‐19 of Ministry of Health and Family Welfare, Government of India.10 Virus clearance was defined as two consecutive oropharyngeal negative swabs after 48–72 h of cessation of COVID‐19 symptoms (Ref). The inclusion criteria of co‐infection with dengue was NS1 antigen positive on enzyme‐linked immunosorbent assay (ELISA) or RT‐PCR positive for less than 5 days of illness or immunoglobulin M positive on capture ELISA for greater than 5 days of illness.11 The inclusion criteria of co‐infection of malaria was positive rapid malaria antigen test or microscopic confirmation of malarial parasite on peripheral smear.11
Continuous variables were described as medians and interquartile ranges (IQRs). Categorical variables were described as frequency and percentages. The virus clearance was calculated for ≤7 days, 8–15 days and 6–40 days in a total of 467 HCWs with COVID‐19. The Mann–Whitney U test, χ2 test, and Fisher exact test were used according to variable types as appropriate. p < .05 was considered statistically significant. A multiple linear regression model was used to know the relation between clinical characteristics and time taken for viral clearance. Dummy variables were created for categorical variables. Analyses was performed using SPSS statistical software version 20.0 (IBM Corp).
3 RESULTS
3.1 Sociodemographic, epidemiological characteristics of HCWs with coronavirus disease 2019
A total of 3711 HCWs (frontline [n = 2758, 74.32%], non‐frontline [n = 953, 25.68%]) were working at NH during the study period. Of which, 491 were found to be SARS‐CoV‐2 positive on RT‐PCR. The prevalence of SARS‐CoV‐2 infection among HCWs during the first seven months of the pandemic was 13.2% (491/3711). Prevalence of SARS‐CoV‐2 infection amongst frontline HCWs was 12.9% and non‐frontline HCWs was 14.3%. The distribution of COVID‐19 HCWs according to their work profile was as follows: 132/491 (27%) physicians, 133/491 (27%) nurses, and 226/491 (46%) healthcare assistants along with other staff members. The prevalence of SARS‐CoV‐2 infection amongst physicians, nurses and healthcare assistants was 12.6%, 14% and 13.2%, respectively. There was a higher prevalence of SARS‐CoV‐2 (24.5%, 36/147) amongst security personnel in our study. Majority (97%) of the security personnel who were COVID‐19 had no direct contact with COVID‐19 patients in the hospital (Figure 1).
Prevalence of SARS‐CoV‐2 infection among HCWs according job profile. HCW, healthcare worker; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2
The sociodemographic characteristics of HCWs with COVID‐19 are presented in Table 1. The HCWs with COVID‐19 had a median age of 32 (27–44.5) years; frontline HCWs were young (32, 27–41) compared to non‐frontline HCWs (36.5, 27–51). Majority (69%) of the HCWs with COVID‐19 were young (<40 years), including 74% frontline HCW and 95% doctors. Incidence of COVID‐19 was higher in non‐frontline male HCWs (77%) than female non frontline HCWs (23%). Majority of the HCWs (66%, 322/491) were working in wards or intensive care unit, and 12% (58/491) were working in the office setup. Around six percent of healthcare assistants were working in other clinical departments like laboratory, radio‐diagnosis and blood bank and majority of them (86%, 25/29) were categorized as non‐frontline. Majority (71%) of the HCWs were residing in the containment or sealed zone as Mumbai city and Mumbai Metropolitan Region (MMR) was epicenter of the COVID‐19 pandemic in India.
Demographic Information for 491 healthcare workers with confirmed SARS‐CoV‐2 infection
| Characteristics | All HCWs, n = 491 | Frontline HCWs, n = 355 | Non‐Frontline HCWs, n = 136 | p Value |
|---|---|---|---|---|
| Age in years | ||||
| Median (IQR) | 32 (27–44.5) | 32 (27–41) | 36.5 (27–51) | <.001 |
| 18–30 | 215 (43.8) | 164 (46.2) | 51 (37.5) | |
| 31–40 | 125 (25.5) | 99 (27.9) | 26 (19.1) | |
| 41–50 | 77 (15.7) | 53 (14.9) | 24 (17.6) | |
| 51–61 | 74 (15.1) | 39 (11) | 35 (25.7) | |
| Gender | ||||
| Male | 267 (54.4) | 163 (45.9) | 104 (76.5) | <.001 |
| Female | 224 (45.6) | 192 (54.1) | 32 (23.5) | |
| Area of residence | ||||
| Containment or sealed zone | 347 (70.7) | 252 (71) | 95 (69.9) | |
| Job category | ||||
| Physician | 132 (26.9) | 123 (34.6) | 9 (6.6) | <.001 |
| Nurse | 133 (27.1) | 130 (36.6) | 3 (2.2) | |
| Other staff | 226 (46) | 102 (28.7) | 124 (91.2) | |
| Place of work | ||||
| Fever clinic | 18 (3.7) | 17 (4.8) | 1 (0.7) | <.001 |
| Ward and ICU | 322 (65.6) | 311 (87.6) | 11 (8.1) | |
| Operation theater | 14 (2.9) | 14 (3.9) | 0 | |
| Lab or Xray or CT scan | 23 (4.7) | 3 (0.8) | 20 (14.7) | |
| Office | 58 (11.7) | 0 | 58 (42.6) | |
| Blood bank | 6 (1.2) | 1 (0.3) | 5 (3.7) | |
| Mortuary | 4 (0.8) | 2 (0.6) | 2 (1.5) | |
| Security | 36 (7.3) | 1 (0.3) | 35 (25.7) | |
| Other | 9 (1.8) | 6 (1.7) | 3 (2.2) | |
- Note: Data are n (%) or median (IQR). p Values are for the comparison between frontline and non‐frontline HCWs.
- Abbreviations: CT, computed tomography; HCW, healthcare worker; ICU, intensive care unit; IQR, interquartile range.
3.2 Exposure information
The telephonic interview data for exposure information from 463 HCWs showed that in majority (55%, 253/463) of the HCWs the most likely source of SARS‐CoV‐2 infection could be within the hospital, either in COVID‐19 wards or non‐COVID‐19 location inside the dedicated COVID‐19 hospital (Table S1). The community exposure was reported in 35% (161 out of 463) of HCWs. Frontline HCWs reported more exposure from COVID‐19 wards or fever clinic (59% vs. 6%) whereas non‐frontline HCWs reported more exposure from community (50% vs. 29%). Twenty‐eight (6%) HCWs with COVID‐19, reported high risk exposure while attending to COVID‐19 patients in the wards or other hospital locations. Out of those, 50% reported that PPE was not used early in the pandemic period when the hospital was not declared as a dedicated COVID‐19 hospital and there was no protocol in place, 40% reported break in PPE while working and one reported exposure to body fluid. Twenty eight percent of HCWs reported contact with COVID‐19 patients in the community or at home while 1% reported domestic air travel during the pandemic as a source of infection. The median (IQR) duration of the work in the hospital before contracting SARS‐CoV‐2 infection was 94 (55–131) days. The median duration of work (days) before contracting the SARS‐CoV‐2 infection in non‐frontline HCWs was lesser (87 days, range 58–130) than the frontline HCWs (101 days, range 53.75–132). Majority (86%, 399/463) COVID‐19 HCWs were tested because of symptoms of COVID‐19 whereas 7% of HCWs were tested due to exposure or history of contact with COVID‐19 patients.
3.3 Co‐morbidities, clinical presentations, and outcome of COVID‐19 in HCWs
Comorbidities were reported in 22% (103 out of 463) of HCWs with COVID‐19. Of these, 30 (7%) HCWs had multiple co‐morbidities. Hypertension (11%), Diabetes Mellitus (8%) and bronchial asthma (4%) were the most common co‐morbidities reported (Table S2). Majority (87%) of the HCWs with COVID‐19 were symptomatic and 13% were asymptomatic. Most common symptoms were fever (64%, 298), dry cough (29%, 136), myalgia (27%, 124), and sore throat (24%, 110). Twenty‐four percents (111 out of 463) of HCWs with COVID‐19 were having symptoms suggestive of severe acute respiratory infection which is defined as acute respiratory infection with fever ≥38°C with cough, requiring hospitalization as per the ICMR testing guidelines.12 The hydroxychloroquine (HCQ) prophylaxis was received by 54% (261 out of 484) of HCWs who were SARS‐CoV‐2 positive. The median (IQR) duration of HCQ prophylaxis was 3 (2–5) weeks.
Majority (74%) of HCWs with COVID‐19 had mild disease, 13% (58/463) developed moderate disease and 2% (11 out of 463) developed severe disease. Twelve (3%) had ARDS, 6% needed oxygen therapy and 4% needed mechanical or noninvasive ventilator support. Frontline HCWs (2%) and non‐frontline (3%) developed severe disease with 0.4% (2 out of 491) mortality. One non‐frontline HCW in late sixties (58‐years‐old) with hypertension and diabetes became critical and was started on BiPAP mechanical ventilation. His HRCT thorax was typical for COVID‐19 pneumonia (CORADS‐6). The severity score of the disease based on visual assessment of overall area of lung involved was two (25%–50% overall lung involvement). Patient was treated with piperacillin or tazobactam, oseltamivir, low molecular weight heparin (LMWH), methylprednisolone, tocilizumab, and ivermectin but succumbed owing to a complication of SARS‐CoV‐2 infection. Another mortality was reported in 51 years old non‐frontline HCW, who died within 24 h of admission due to respiratory failure with lower respiratory tract infection with ARDS with cardiogenic shock. He was symptomatic (diarrhea, sore throat, fever and breathlessness) since 2 days. He was a known case of diabetes mellitus, hypertension and IHD (CABG 2 years back). He also had past history of right popliteal artery thrombosis and gas gangrene in left groin. He received piperacillin or tazobactam, remdesivir, LMWH, frusemide, and noradrenaline support.
3.4 Co‐infection of malaria or dengue in HCWs with COVID‐19
There were 31 cases (6.3%) of co‐infection with vector borne diseases; malaria (n = 27) or dengue (n = 5). Nearly 50% (15 out of 31) of HCWs with co‐infection were residing in the hospital premises of which 60% (9 out of 15) were resident doctors. SARS‐CoV‐2 co‐infection with malaria or dengue in HCWs was higher in months of July (n = 15), June and August (5 each) as compared to May (n = 1), September (n = 3) and October (n = 2). SARS‐CoV‐2 infection in HCWs was higher in months of May (108, 22%) and July (115, 23.4%) compared to other months (Figure 2). Ninety percent (28 out of 31) of HCWs with COVID‐19 and co‐infections of malaria or dengue were symptomatic and only three were asymptomatic.
Monthly distribution of HCWs with COVID‐19 and co‐infection with malaria or dengue. COVID‐19, coronavius disease 2019; HCW, healthcare worker
3.5 Virus clearance
There was a variation in duration for virus clearance from the respiratory tract (Table 2). Significant differences were observed in the days for viral clearance among the HCWs who were symptomatic (11 days [IQR 8–15 days] [n = 381]) and asymptomatic (7 days [IQR 5–11 days] [n = 86)] (p < .001). Similarly, we observed significant differences among the HCWs who presented with co‐infection with malaria [7 days (IQR 6–8 days)] and without co‐infection with malaria (11 days [IQR 7–15 days]) (p < .001). We further chose duration of viral clearance as dependent variable and the variables that were significant in the baseline analysis (symptomatic and presence of co‐infection with malaria) as independent variables for multiple linear regression model. The model explained 6.6% variance for duration of viral clearance by the independent variables. The results of multiple linear regression model shows that presence of symptoms have a significant and positive association (ß = 0.240) with duration of viral clearance, while presence of co‐infection (ß = −0.030) has negative association but was not significant. Co‐infection with malaria had faster recovery (mean 7.7 days) compared to those who did not have co‐infection with malaria (mean 11.5 days) (p < .005).
SARS‐CoV‐2 virus clearance in healthcare workers in India (n = 467)
| Days of viral clearance | All, n = 467 (%) | Asymptomatic, n = 58 (%) | Symptomatic, n = 409 (%) | Coinfection with malaria, n = 27 (%) | Coinfection with dengue, n = 4 (%) |
|---|---|---|---|---|---|
| Mean ± SD | 11.31 ± 5.75 | 8.07 ± 4.48 | 11.77 ± 5.77 | 7.70 ± 3.22 | 14.00 ± 3.91 |
| 95% CI | 10.78–11.83 | 6.89–9.25 | 11.20–12.33 | 6.43–8.98 | 7.77–20.23 |
| Median (IQR) | 11 (7–14) | 6 (5–11) | 11 (7–15) | 7 (6–8) | 14.5 (12–16.5) |
| Minimum | 2 | 2 | 2 | 4 | 9 |
| Maximum | 40 | 24 | 40 | 19 | 18 |
| p Valueb | <.005 | <.005 | .180 | ||
| Viral clearance, days | |||||
| ≤7 | 140 (30) | 34 (58.6) | 106 (25.9) | 16 (59.3) | 0 (0) |
| 8–15 | 240 (51.4) | 20 (34.5) | 220 (53.8) | 10 (37.0) | 2 (50) |
| 16–40 | 87 (18.6) | 4 (6.9) | 83 (20.3) | 1 (3.7) | 2 (50) |
| p Valuec | <.005 | .003 | |||
| Multiple regression model | |||||
| Clinical characteristics | Regression coefficient (lower, upper 95% CI) | ||||
| Presence of co‐infectione | −0.030 (−3.282, 1.808) | ||||
| Symptomatic HCWsf | 0.240 (2.032 5.097)d | ||||
- Abbreviations: CI, confidence interval; HCW, healthcare workers; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2.
- a Data was not available for 22 patients and 2 deaths were excluded from this analysis.
- b Mann–Whitney test was applied at the significant level of p < .05.
- c χ2 test was applied at the significant level of p < .05.
- d Statistically significant p < .05.
- e Co‐infection of malaria (n = 27).
- f Symptomatic excluding co‐infection (n = 381).
4 DISCUSSION
In the present study, we observed 13% prevalence of SARS‐CoV‐2 infection, 25% among security guards, 6% co‐infection of malaria or dengue and 1% mortality among HCWs. Co‐infection with malaria or dengue was highest in the start of monsoon season in the month of July, which coincides with the monsoon in India and peak of COVID‐19 pandemic in the Mumbai city where the study was conducted. Fever and cough were the most common symptoms reported in our study. During the initial seven months of pandemic, City of Mumbai along with MMR was catered by NH, the first dedicated COVID‐19 hospital in India.8 Despite the multitude of caseload of COVID‐19, prevalence of SARS‐CoV‐2 in HCWs was comparatively low in our study than other reported studies in UK (18%)13 and Italy (18%).14 Study from Spain (11%)5 reported similar prevalence whereas study from Netherlands15 reported less (1%) prevalence than our study. These differences in prevalence of SARS‐CoV‐2 infection in HCWs could be due to differences in selection criteria, study design and duration of study period. In our study, there was no difference in proportion of doctors, nurses and other HCWs infected with SARS‐CoV‐2 which strongly support the complacent PPE protocols at our dedicated COVID‐19 hospital. In our study, symptomatic HCWs with COVID‐19 were reported at a higher proportion (doctors, nurses [87%], other HCWs [88%]) which is in contradiction with other studies as majority of the SARS‐CoV‐2 infection in HCWs was asymptomatic2, 3 or mild to moderate.4–6 In our study, the mortality associated with COVID‐19 infected HCWs was low unlike China that reported comparatively higher mortality in HCWs with COVID‐19 infection.16
During ongoing pandemic, the entire focus was on the frontline HCWs whereas staff with no direct contact with patients were neglected globally, especially the security personnel. Working in high risk locations, such as hospital premises, the security personnel were involved in directing the patients, relatives and managing the entry and exit of the patients, visitors and hospital staff. Majority of the security personals had a working place outside the hospital buildings within the hospital campus. It was impossible for them to wear PPEs in extreme heat and humidity at their work location. However, basic prevention strategies like hand hygiene, wearing N‐95 masks, head shield or eye goggles were used by them.
Longer duration for virus clearance was observed in symptomatic as compared to asymptomatic HCWs in our study. We found a significant association between time taken for viral clearance and the presence of co‐infection with malaria and symptoms. These findings suggest that presence of symptoms prolonged the viral clearance. However, our model showed a very low variance suggesting that there might be other factors influencing the time taken for viral clearance. In Italian study, the mean duration for viral clearance in symptomatic HCWs and asymptomatic HCWs was 34.2 and 22 days, respectively17 which was higher than viral clearance time reported in our study. Surprisingly, co‐infection with malaria reported faster recovery and faster virus clearance. This could possibly be explained by the hypothesis of cross reactivity of Glycosylphosphatidylinositol (GPI) antibodies (immunoglobulin G [IgG]) to plasmodium specific antigens with SARS‐CoV‐2 antibodies. Possibly, this might be playing a role in early remission among COVID‐19 patients with malaria.18, 19 Thus, malarial infection may confer protection from SARS‐CoV‐2 Infection. Higher proportion of anti‐malarial antibodies are reported in residents of malaria‐endemic areas. Therefore, future studies should be conducted to check if there is cross‐reactivity of these antibodies, along with GPI antibodies and COVID‐19 for demonstrating a cause‐effect relationship.
We acknowledge several limitations in this study. Primarily this study is a single‐centered. Being a retrospective analysis, there was no standard control group for comparison or random sampling which could have resulted in selection bias. The distinct SARS‐CoV‐2 strain on genome sequencing in HCWs was not confirmed. To keep our interview brief, details for some exposures were curtailed. For instance, we excluded the details about the type of PPE used and donning or doffing. However consistent PPE use was determined with the help of interviews. Although our primary conclusions for prevalence and viral clearance were compelling, clinical details like severity of symptoms and treatment could not be evaluated for 28 HCWs as they were hospitalized at other hospitals; although we could capture basic details from central database for calculating prevalence after permission from ethics committee and administration. Recall bias was present for the exposure information as there was an interval of a few months between the occurrence of the event and interview for some HCWs. Although our findings support the hypothesis of cross reactivity of GPI antibodies IgG to Plasmodium specific antigens with SARS‐CoV‐2 antibodies and its role in early remission, we could not confirm this hypothesis with serological tests because of financial constraints.
5 CONCLUSION
During initial seven months of pandemic, we observed 13% prevalence of SARS‐CoV‐2 infection in HCWs (25% amongst security guards) and more than 5% percent co‐infection with plasmodium vivax. At present, universal testing of HCWs should be recommended for optimization of staffing levels as HCWs are the most priceless resource for all the nations. Family members and close contacts of HCWs should also be tested. Higher duration for virus clearance among symptomatic HCWs, two mortalities and high co‐infection rate with vector borne diseases, indicate the importance of preventive measures. In a hospital setting like ours, we observed nearly 50% of HCWs with co‐infections, who were staying in the hospital premises. Importantly, higher prevalence among security guards is quite worrisome. There is a need to effectively manage the coinfection of malaria and dengue with SARS‐CoV‐2 and implement standard protocols for prevention of vector‐borne diseases, especially in the hospitals and endemic areas.
ACKNOWLEDGMENTS
The authors acknowledge Dr Ramesh Bharmal, Dean TNMC; Dr. Mohan Joshi, former Dean TNMC and Dr. Sarika Patil, Nodal Officer, BYL Nair Charitable Hospital, Mumbai, BYL Nair Charitable Hospital, Mumbai for assistance in implementation of the study. Ms Aishwarya Bhurke is acknowledged for statistical assistance. RG is an awardee of the DBT Wellcome India alliance clinical and public health intermediate fellowship (Grant no. IA/CPHI/18/1/503933).
CONFLICT OF INTERESTS
The authors declare that there are no conflict of interests.
TRIAL REGISTRATION
Study is registered with Clinical Trial Registry of India (Registration no: CTRI/2020/09/027516).
AUTHOR CONTRIBUTIONS
Niraj N. Mahajan had full access to all of the data in the study and take responsibility for the Integrity of the data and the accuracy of the data analysis. Concept and design: Niraj N. Mahajan and Rahul K. Gajbhiye. Acquisition of data: Shubhada Bahirat, Pradip D Lokhande, Apeksha Mathe, and Vartika Srivastava. Drafting of the manuscript: Rahul K. Gajbhiye, Niraj N. Mahajan, Kshitija N. Mahajan, and Neeta Warty. Critical revision of the manuscript for important intellectual content: Niraj N. Mahajan, Rahul K. Gajbhiye, Neeta Warty, and Kshitija N. Mahajan. Statistical analysis: Niraj N. Mahajan, Rahul K. Gajbhiye, Periyasamy Kuppusamy, and Kshitija N. Mahajan. Administrative and technical or material support: Niraj N. Mahajan, Shailesh C. Mohite, Shubhada Bahirat, and Neeta Warty.
ETHICS STATEMENT
The study was approved by the Ethics Committees of TNMC (No. ECARP/2020/78 dated August 13th, 2020).
Data available on request due to privacy/ethical restrictions. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
REFERENCES
