We report from this Covid-19 outbreak investigation in HCWs from a low-risk area of the hospital that the outbreak resulted from possible suboptimal screening of asymptomatic patients with Covid-19. The outbreak was further triggered due to lack of physical distancing and sharing meals together among HCWs. We also noticed that both the physical and behavioral risk factors mentioned above were more prevalent among nurses as compared to resident physicians.

A large meta-analysis on nosocomial Covid-19 confirmed and near-outbreaks consisting of 22 studies from United States, Europe, and Asia [1], focused more on nursing homes rather than HCWs. The attack rate from this meta-analysis concerning HCWs ranged from 2.5% to 17%. However, there are only few reports of Covid-19 outbreak in a low-risk area of the hospital [8-10] with attack rate being reported as low as 2.5% in the study from the United States [8] and 5.2% from Germany [9].

An Italian study showed that with proper following of the infection protocols, the prevalence of SARS-CoV-2 is just 0.43% among all HCWs working at a healthcare institution [11]. Our study showed an attack rate of 12.05% which falls within the figures reported in the studies mentioned above. However, our findings are specifically from a low-risk area in a hospital. This might imply that these numbers are slightly higher because they are from a low-risk area. The reasons behind this relatively high attack rate are inappropriate use of PPE in the early days of the pandemic due to lack of awareness thus leading to variation in practice. Secondly, HCWs working in low-risk areas might have been illusioned that they might not contract Covid-19 as they are in low-risk areas. Thirdly, the standard infection control practices in the low-risk area were of course less stringent compared to the high-risk red zone. Furthermore, some patients with minimal suspicion of Covid-19 or those who were either asymptomatic or screened preoperatively can be the source of outbreak in low-risk green zone. Finally, patients may have tested negative for the SARS-CoV-2 PCR initially as they might have been in the incubation period. This was also seen in a study from Korea where the index case behind the outbreak initially tested negative on admission but later developed symptoms and turned out to be positive for Covid-19 [12].

The chain of transmission of Covid-19 in our ward likely started from a patient who turned out to be positive for Covid-19 and was probably missed in the initial screening. Once HCWs were infected from the patients, further propagation of Covid-19 in the ward was more likely due to transmission between HCWs rather than due to interaction between HCWs and patients. However, during this period a similar high infectivity rate was not seen in the patients admitted in the low-risk ward. Literature review shows that the healthcare community has also been pondering on the fact that the risk of Covid-19 transmission to HCWs is more during peer-to-peer interaction rather than during direct patient care [1,13]. Unfortunately, like our study, most of the other studies were also unable to establish the sequence of infection in patients and hence were unable to identify the index case [14].

Our study showed that most of the individuals affected by this outbreak were nurses. This may be because the total number of nurses working in the medicine ward was far greater than that of physicians. Moreover, limited eating areas for nurses could be the reason for nurses’ inability to maintain physical distancing while taking meals. Studies done globally show a similar trend [9,14], except a study from Wuhan which was dominated by physicians as the source of infection was a multidisciplinary meeting in that study [15]. No significant differences were observed in the practices of the infected male and female HCWs. Performing AGPs was the only activity where the male population did not properly follow the PPE protocols. However, since the number of patients turning out positive for Covid-19 was small, the significance of this difference remains unclear.

The reasons behind nosocomial outbreaks in low-risk areas can be complex. Contributing factors can include suboptimal hand hygiene or other infection control practices by HCWs; inadequate laboratory results or inappropriate triaging; incorrectly engineered heating, ventilation, and air-conditioning systems; or even misinterpretation of existing infection control guidelines [16]. The common factor in our study which predisposed our HCWs to be classified as a high-risk contact was lack of wearing a gown while interacting with either patients or other HCWs and lack of wearing a face mask by those patients and HCWs who were symptomatic for Covid-19. Whether it involves facemask compliance of patients or the compliance of PPE by the HCWs, the overall compliance of PPE and other infection control practices in a healthcare setup has remained an issue globally [17]. Literature also shows that major risk factors for other nosocomial outbreaks are lack of compliance to PPE especially masks, and proper hand hygiene [18]. Studies from Massachusetts and North Carolina showed that chances of Covid-19 transmission among HCWs were greater from an unmasked exposure to another HCWs rather than an exposure to a known infected patient [18,19].

Even with the advent of vaccines, the primary prevention technique for the Covid-19 pandemic still revolves around physical distancing [20]. There is no doubt that physical distancing also has an important part to play in a healthcare setup [21] It is ideal to have six feet between HCWs when rounding on patients or performing tasks on computers. Our investigation showed that lack of physical distancing was one of the major behavioral risk-defining activities among HCWs, with the other being sharing meals. Mealtimes are regarded as a restful and stress-relieving period and sharing of meals is often practiced all around the world. Though sharing of meals and eating out in restaurants has been seen to cause an increase in Covid-19 transmission [22], only one study has commented on the lack of physical distancing while eating unmasked as a risk factor for Covid-19 transmission among HCWs [18].

Studies from around the world have also shown that universal masking policy results in a significant decrease in nosocomial transmission of Covid-19 [18]. Apart from this, there should be an arrangement of a dedicated, well-ventilated area where HCWs can take a break from daily work activities and can have their meals [18]. This restructuring should be extended to reorganization in the ward structure to help in the implementation of physical distancing [21]. Early identification of potential patients using diversified testing criteria and having a low threshold for testing to avoid missing less suspicious cases may also avoid invariable exposures to HCWs [1] While studies have shown that serial testing of patients and HCWs every 7-10 days may identify outbreaks early [23], this may be difficult and costly in a resource-constrained setting.

Once the outbreak investigation concluded, a set of sweeping reforms were introduced to minimize the occurrences of such outbreaks in the future: (i) PPE protocols were revised in low-risk medicine wards and it was made mandatory to wear gown and face shield/goggles in addition to gloves and face mask in all interactions with patients, while face mask and face shield/goggles were made mandatory to be worn all over the hospital; (ii) Use of an N95 mask was made mandatory whenever an HCW went into a high-dependency unit; (iii) Number of HCWs in the unit at a certain time was reduced by restructuring the roster and keeping the minimum number on duty as needed to maintain quality care. This was also done to minimize the number of exposures in case any HCW became symptomatic and tested positive for Covid-19; (iv) HCWs were advised to strategically plan their rounds and reduce unnecessary physical examinations. Computers were placed six feet apart and HCWs were discouraged from taking long breaks together. In addition, taking meals within the vicinity of the ward or in larger groups anywhere in hospital was disallowed to maximize the compliance to physical distancing; (v) Universal testing for SARS Covid antigen was done on all patients being admitted form ER and better screening mechanisms for symptoms at entry points of all hospital wards was mandated; (vi) Educational and awareness sessions were conducted for HCWs for correct practice of PPE and on importance of physical distancing.

Our study is one of the first to address a nosocomial Covid-19 outbreak from Asia / developing countries. The study also helped in initiating specific reforms and policies for the safety of the HCWs working in a developing country in a resource-constrained setting. It reports both physical and behavioral risk factors for contracting infections. However, despite these strengths, the study does have some limitations. Firstly, data is lacking on the number of patients who were Covid-19 infected during this period from the two wards and were discharged prior to testing. Secondly, the data regarding the symptoms of the HCWs and their activities is limited by recall bias and the fear of being caught out due to not following the hospital infection control policies in place. Finally, there may be a few HCWs and administrative staff from other wards who may have had close contacts with HCWs of our ward leading to missed exposures.