India, like the rest of world, is struggling with the COVID-19 pandemic with Mumbai, its financial capital being the worst affected city in the country [1]. Absence of HCWs from work due to illness and quarantine stresses the already overburdened medical services even more, and their replacement isn’t easy due to limited numbers of trained personnel [2]. Moreover, HCWs may act as super-spreaders in the hospital set up, especially when asymptomatic and in the incubation period, transmitting the infection to vulnerable patients [3]. Two of the major government initiatives to reduce this rate of infection have been HCQ prophylaxis for HCWs at high risk, and the segregation of hospitals (public as well as private) into designated COVID-19 and non-COVID-19 facilities [4]. The former has been effective in reducing infection rates, while there is only empirical evidence of the effectiveness of the latter. The rates of SARS-CoV-2 infection among HCWs in different countries have ranged between 5-44% [3,5-7]. Currently, literature regarding risk factors for infection among HCWs in India is very limited. Such information can provide important insights for devising and implementing strategies to reduce the burden of COVID-19 cases among HCWs [8,9].
In this cross-sectional evaluation, an attempt has been made to ascertain demographic, co-morbidity and exposure characteristics with real-time RT-PCR positivity for SARS-CoV-2 infection among HCWs.
Materials and Methods
The present study was a cross-sectional study which was conducted on HCWs including doctors, nurses and ancillary workers, working in three large public tertiary care hospitals participated in the seroprevalence study conducted during June 2020. Ancillary workers include staff cleaners, social workers, staff in mortuary, laboratory technicians, paramedical staff, security officers and porters who have direct patient contact. The study was conducted in accordance to the tenets of the Declaration of Helsinki. The study protocol was approved by the Institutional Review Board, JJ Hospital and Grant Medical College, Mumbai, Maharashtra, India (IEC/Pharm/RP/125/Jun/2020).
All three hospitals belong to the same management group ensuring standardisation of COVID protection protocols across the hospitals. One hospital is a designated non-COVID hospital admitting only COVID-19 negative patients while the other two hospitals are designated COVID-19 hospitals admitting only COVID positive patients. During the pandemic HCWs have been working across the hospitals to cover for colleagues who have taken ill or required quarantine due to exposure to a COVID-19 patient either at work or at home.
Sample size was not formally calculated, all the HCWs in the above three hospitals were contacted and those willing to participate and gave informed consent were included in the study. A total of 801 HCWs were enrolled in this seroprevalence study for antibodies against COVID-19. In the present study the sub cohort of RT-PCR positive HCWs and the risk factors associated with the infection were examined.
All participants were self-administered a pre-designed, validated questionnaire. The english questionnaire was validated by forward-back translations into Hindi and Marathi, the lingua franca of most of the ancillary staff.
The questionnaire had questions designed to elicit demographic details, information related to co-morbidities, history of COVID-19 related symptoms, contact with confirmed COVID-19 patients, risk factors for COVID infection at home, protective precautions taken at the work place, visit to a fever clinic during last one month, prior diagnosis of COVID-19, and if positive, date of test performed.
Statistical Analysis
Frequency and percentages were calculated for categorical variables. Median and range was reported for continuous variables. The overall and risk group specific RT-PCR positivity rates were reported with 95% confidence intervals using Open Epi (Open Source Epidemiologic Statistics for Public Health). Additionally, positive RT-PCR rates were reported according to demographics, co-morbidities, work related risk/exposures and prevention practices. Difference in proportion was examined by Chi-square tests with Yates’ correction, if required. According to needs, Fishers-exact test was also used. The p-value of <0.05 using two-tailed test was considered as statistically significant.
Results
Total of 801 HCWs included 201 doctors (25.1%), 308 nurses (38.5%), and 292 ancillary staff (36.4%). Four hundred one (50.1%) study participants were working in a dedicated COVID-19 hospital, whereas 400 (49.9%) were working in a non-COVID-19 hospital. Of these, 386 (48.2%) were males, with only 8 (1%) subjects being over the age of 60 years. A total of 682 (85.1%) study participants did not have any co-morbidity requiring treatment [Table/Fig-1].
Baseline demographics of the study participants.
Parameter | n (%) |
---|
Profile of healthcare workers n (%) |
Doctors | 201 (25.1%) |
Nurses | 308 (28.5%) |
Support staff | 292 (36.4%) |
Gender n (%) |
Male | 386 (48.2%) |
Female | 415 (51.8%) |
Age group n (%) |
20-40 years | 413 (51.6%) |
41-60 years | 380 (47.4%) |
>61 years | 8 (1%) |
Co-morbidities n (%) |
Atleast one co-morbidity | 103 (12.9%) |
Two or more co-morbidities | 16 (2%) |
Diabetes | 38 (4.7%) |
Asthma | 35 (4.4%) |
Previous diagnosis of cancer | 5 (0.6%) |
Receiving immunosuppressive treatment | 11 (1.4%) |
Cardiac disease | 48 (6%) |
A total of 62 (7.7%) study participants had tested positive with RT-PCR test for COVID in the past [Table/Fig-2]. Of these, the highest rate of infection was found in doctors, followed by nurses and ancillary staff. The rate of infection was significantly higher in non-COVID-19 hospitals as compared to COVID-19 (p=0.032).
RT-PCR positivity rate as per demographic details and co-morbidities.
Parameter | RT-PCR Negative | RT-PCR positive | % positive | 95% CI | p-value (Chi-square/Fisher’s Test) |
---|
COVID-19 result (n=801) | 739 | 62 | 7.74 | 6.07-9.81 | |
Profile | 0.39 |
Doctors (n=201) | 182 | 19 | 9.45 | 6.06-14.36 |
Nurses n (n=308) | 283 | 25 | 8.12 | 5.51-11.75 |
Support staff (n=292) | 274 | 18 | 6.16 | 3.87-9.58 |
Hospital | 0.032* |
COVID hospital (n=401) | 366 | 35 | 8.72 | 6.95-12.76 |
Non-COVID-hospital (n=400) | 346 | 54 | 13.5 | 4.02-8.08 |
Gender | 0.97 |
Male (n=386) | 356 | 30 | 7.77 | 5.46-10.91 |
Female (n=415) | 383 | 32 | 7.71 | 5.48-10.71 |
Age group | 0.17 |
20-40 years (n=413) | 380 | 33 | 7.99 | 5.71-11.04 |
41-60 years (n=380) | 353 | 27 | 7.11 | 4.89-10.17 |
>61 years (n=8) | 6 | 2 | 25.00 | 6.3-59.91 |
Immunocompromised (Cancer/Immunosuppressants) | 0.98 |
No (n=105) | 97 | 8 | 7.62 | 3.7-14.52 |
Yes (n=14) | 13 | 1 | 7.14 | 0-33.54 |
Asthma | 0.048* |
Yes (n=35) | 32 | 3 | 5.71 | 62-19.57 |
No (n=766) | 760 | 6 | 0.78 |
Cardiac problem | 0.83 |
Yes (n=48) | 44 | 4 | 8.33 | 2.76-20.08 |
No (n=753) | 695 | 58 | 7.70 |
Diabetes | 0.92 |
Yes (n=38) | 35 | 3 | 7.89 | |
No (n=763) | 704 | 59 | 7.73 |
*p-value significant
There was significant difference in rate of RT-PCR positivity in those with asthma (95% CI 62-19.57; p=0.048) as compared to those without asthma. For other risk factors i.e., immunocompromised status because of cancer/immunosuppressant drugs, cardiovascular morbidity/diabetes there was no significant difference in RT-PCR positivity rate [Table/Fig-2].
There was significant difference in RT-PCR positivity rates in symptomatic patients versus asymptomatic people (p<0.001). Symptoms like loss of taste/smell, acute febrile illness, acute respiratory illness, non-specific illness were associated with higher rates of RT-PCR positive rates than those without these symptoms (p<0.005, highly significant statistically). Other less common symptoms such as acute gastric/enteric illness/redness of eyes and skin rash were not associated with significant rates of RT-PCR positivity (p>0.05) [Table/Fig-3].
RT-PCR positivity rate based on symptoms.
Parameter | RT-PCR negative | RT-PCR positive | % positive | 95% CI | p-value (Chi-square/Fisher’s-test) |
---|
Symptomatic |
Yes (n=167) | 33 | 134 | 80.24 | 73.51-85.61 | 0.001* |
No (n=634) | 605 | 29 | 4.57 | 3.18-6.51 |
Loss of taste/smell |
Yes (n=10) | 7 | 3 | 30.00 | 10.33-60.77 | 0.03377* |
No (n=791) | 736 | 55 | 6.95 |
Acute febrile illness |
Yes (n=28) | 15 | 13 | 46.43 | 29.53-64.19 | 0.001* |
No (n=773) | 724 | 49 | 6.34 | 4.81-8.29 |
Acute respiratory illness |
Yes (n=97) | 79 | 18 | 18.56 | 4.67-8.3 | 0.001* |
No (n=704) | 660 | 44 | 6.25 | 11.98-27.52 |
Non-specific illness |
Yes (n=90) | 76 | 14 | 15.56 | 9.37-24.56 | 0.008* |
No (n=711) | 663 | 48 | 6.75 | 5.11-8.55 |
Acute gastric/enteric illness |
Yes (n=8) | 7 | 1 | 12.50 | 0.11-49.21 | 0.5986 |
No (n=793) | 732 | 61 | 7.69 | 6.02-9.76 |
Redness of eyes |
Yes (n=12) | 11 | 1 | 8.33 | 0 | 0.8587 |
No (n=789) | 728 | 61 | 7.73 | 6.05-8.54 |
Skin rash |
Yes (n=9) | 9 | 0 | 0 | 0 | 0.4825 |
No (n=792) | 730 | 62 | 7.83 | 6.14-9.92 |
*p-value significant
RT-PCR positivity rate was significantly higher in those who visited fever clinic, having positive household member and directly exposed to COVID-19 patient (p<0.05 for all; [Table/Fig-4]). There was no difference in the rate of RT-PCR positivity in HCWs having neighbours positive for COVID-19, using shared toilet, living in hotspot/containment zone or working in the tertiary care hospitals (p>0.05) [Table/Fig-4].
RT-PCR positivity rate based on exposure.
Parameter | RT-PCR Negative | RT-PCR positive | % positive | 95% CI | p-value (Chi-square/Fisher’s test) |
---|
Visited fever clinic |
Yes (n=132) | 109 | 23 | 17.42 | 11.84-24.85 | 0.001* |
No (n=669) | 630 | 39 | 5.83 | 4.27-7.88 |
Household member positive |
Yes (n=74) | 57 | 17 | 22.97 | 14.78-33.83 | 0.001* |
No (n=727) | 682 | 45 | 6.19 |
Neighbourhood positive |
Yes (n=397) | 364 | 33 | 8.31 | 5.95-11.47 | 0.55 |
No (n=404) | 375 | 29 | 7.18 | 5.01-10.15 |
Quarantined |
Yes (n=176) | 135 | 41 | 23.30 | 17.64-30.1 | 0.001* |
No (n=625) | 604 | 21 | 3.36 | 2.18-5.11 |
Shared toilet |
Yes (n=314) | 292 | 22 | 7.01 | 4.62-10.43 | 0.53 |
No (n=487) | 447 | 40 | 8.21 | 6.06-11.01 |
Living in hotspot/containment zone |
Yes (n=531) | 485 | 46 | 8.66 | 6.53-10.76 | 0.17 |
No (n=270) | 254 | 16 | 5.93 | 3.61-6.56 |
Directly exposed to COVID-19 patient |
Yes (n=343) | 304 | 39 | 11.37 | 8.4-15.199.46-7.44 | 0.003* |
Maybe (n=283) | 266 | 17 | 6.01 |
No (n=175) | 169 | 6 | 3.43 |
Worked in COVID hospital |
Yes (n=603) | 542 | 61 | 8.46 | 6.47-10.97 | 0.185 |
No (n=198) | 187 | 11 | 5.56 | 3.02-9.77 |
*p-value significant
Use of protective measures like mask use outside home (irrespective of type of mask), use of PPE at work, social distancing outside home and persons in room were not associated with significant difference in positive rates for RT-PCR for COVID-19. Hydroxychloroquine (HCQ) use was associated with significantly lesser rates of RT-PCR positivity than those who did not use it (p<0.05) [Table/Fig-5].
RT-PCR positivity rate based on protective measures.
Parameter | RT-PCR Negative | RT-PCR positive | % positive | 95% CI | p-value (Chi-square/Fisher’s-test) |
---|
Mask use outside home |
>75% (n=577) | 532 | 45 | 7.80 | 5.86-10.29 | 0.74 |
50-75% (n=163) | 152 | 11 | 6.75 | 3.68-11.8 |
<50% (n=61) | 55 | 6 | 9.84 | 4.24-20.19 |
Mask type |
N95 (n=559) | 515 | 44 | 7.87 | 5.89-10.42 | 0.52 |
Surgical (n=189) | 173 | 16 | 8.47 | 5.19-13.39 |
Cloth (n=53) | 51 | 2 | 3.77 | 0.3-13.48 |
PPE use frequency at work |
Always (n=130) | 117 | 13 | 10 | 5.81-16.48 | 0.35 |
On direct contact (n=449) | 413 | 36 | 8.02 | 5.82-10.92 |
Never (n=222) | 209 | 13 | 5.86 | 3.36-9.84 |
Six feet distancing outside home |
>75% times (n=292) | 269 | 23 | 7.88 | 5.25-11.59 | 0.93 |
50-75% times (n=312) | 287 | 25 | 8.01 | 5.44-11.6 |
<50% times (n=197) | 183 | 14 | 7.11 | 4.19-11.66 |
Persons in room |
<5 (n=639) | 587 | 52 | 8.14 | 6.24-10.53 | 0.41 |
>5 (n=162) | 152 | 10 | 6.17 | 3.25-11.12 |
Hydroxychloroquine use |
Yes (n=488) | 300 | 13 | 4.15 | 2.37-7.04 | 0.003* |
No (n=313) | 439 | 49 | 10.04 | 7.66-13.05 |
*p-value significant
Discussion
In this cross-sectional study, authors had compared positive rates of COVID-19 based on the RT-PCR test among HCWs in COVID-19 designated hospitals and non-COVID-19 hospitals Mumbai, Maharashtra, India. Overall rate of infection diagnosed with RT-PCR was 7.74%. In a study from the United Kingdom, the rate of positivity among 1533 symptomatic HCWs was 18% [3]. Another study from London reported 44% out of 200 HCWs to have SARS-CoV-2 infection as identified by either serology or RT-PCR [5] whereas a study from Belgium reported that overall infection rate of 12.6% [6]. In another study from Netherland, 5% HCWs out of 1796 were positive for SARS-CoV-2 [7].
The percentage of infection among HCWs in COVID hospitals in present study was significantly less than those with non-COVID hospitals (8.72% vs 13.5%; p=0.032). These observations from present study, provides an important message that HCWs in non-COVID-19 designated hospitals also need to take adequate precautions and cannot afford to be complacent towards the infection.
The rate of infection was numerically higher among doctors than nurses and support staff however this did not reach the statistical significance. Authors did not segregate the number of doctors with infection based on their profile of work i.e., involved in endotracheal intubation, intensive care or regular outpatient examination. This sub-group analysis might provide more insights into the high risk work among doctors. A case control study from India has reported higher risk of infection in doctors performing endotracheal intubation [4].
Symptoms are important for screening and predicting risk of COVID-19 among HCWs. Studies have reported loss of smell or taste, fever, and myalgia as the strongest predictors for positive results for COVID-19 [10,11]. This has been corroborated in present study also presence of symptoms was associated with significantly higher rates of RT-PCR positive rates as compared to those without symptoms. Authors observed significantly higher rates of positive RT-PCR among those with loss of taste/smell, fever and respiratory symptoms than without these symptoms. It should be remembered that some HCWs may not have symptoms, but still they are infected with SARS-CoV-2 [11].
Based on the analysis of those having co-morbidities with asthma were associated with higher risk of infection due to SARS-CoV-2. Cancer or use of immunosuppressant medicines was not associated with increased risk of COVID-19 among HCWs. Similarly, diabetes and cardiac problems were also not associated with increased risk of infection with SARS-CoV-2. It is known that patients with COVID-19 having hypertension or diabetes mellitus are at higher risk of more severe disease course and progression of the disease [12]. A study from China reported that laboratory confirmed cases of COVID-19 with co-morbidity have poorer outcomes as compared to those without co-morbidities. Similarly, in the same study, higher number of co-morbidities correlated with poorer outcomes [13].
A case-control study among HCWs in India reported significantly lower risk among those having taken four or more maintenance doses of HCQ. Use of PPE was also associated with a reduced risk of infection due to SARS-CoV-2 [4]. In this study, also it was found that the risk of infection was lower in those consuming HCQ. In present study, mask use outside home, PPE use frequency at work, six feet distancing outside home and number of persons in a room were not associated with significant difference in the increased risk of RT-PCR positive rates.
Interestingly, highest infection rates were seen with surgical masks (8.47%) followed by N95 masks (7.87%), and the lowest infection rates with cloth masks (3.77%). While this was not found to be statistically significant because of the small sample size, this is a very interesting finding. A possible explanation could be the fact that those working in close contact with COVID-19 patients invariably wear N95 masks as a part of disease protocol, while those working in supporting functions often wear cloth masks. The finding of the highest infection rates with surgical masks may be attributed to the hierarchy of distribution of PPE, consistent with reports of PPE shortage across the globe: where N95 masks are prioritised for doctors and nurses, with the latter, along with ancillary staff often having to make-do with surgical masks. A correlation with the role of hospital hierarchy in PPE distribution in each mask category was not possible due to the small sample size. These findings are contradictory to available evidence about the superiority of N95 masks in limiting the spread of COVID-19, and authors reiterate the importance of protective measures at work, in home and outside work to limit the spread of virus.
A small study (n=4) suggests that although RT-PCR is a useful test for diagnosis of COVID-19, some recovered patients may still be carriers of the virus [14].
Limitation(s)
This study had several limitations. While the participation was multicentric, it was limited geographically to Mumbai, and there is no representative of the prevalence rates across other healthcare facilities. Also, since the study cohort was largely voluntary, it may not be considered representative of the entire facility.
Additionally, the veracity of questionnaire-based information is always susceptible to recall bias. Also, asymptomatic COVID-19 infections are not accounted for, which may constitute a significant percentage of COVID-19 infections, since RT-PCR is mandated only for those with symptoms. Considering these limitations, observations of present study should be extrapolated with caution to general population.
Conclusion(s)
Infection rate with SARS-CoV-2 among HCWs in three public hospitals in Mumbai was found to be 7.7%. Presence of symptoms, especially, loss of taste/smell, fever and respiratory symptoms are associated with high positive rates. HCQ prophylaxis was associated with reduced rate of COVID-19 infection among HCWs.
*p-value significant*p-value significant*p-value significant*p-value significant