Prevention of nosocomial transmission of SARS-CoV-2 using pre-operative chest CT: a monocentric study during the outbreak

To curb the risk of nosocomial infection during the COVID-19 pandemic, most hospitals postponed non-urgent surgeries by analysing the risk–benefit balance for patients on a case-by-case basis. However, some operations could not be delayed without inducing some risk (e.g. cancer treatment) [1]. Meanwhile, asymptomatic pneumopathies were described on low-dose chest computed tomography (ld-CT) [2,3]. In agreement with the nosocomial infection control committee and occupational health services, the study hospital implemented pre-operative screening as part of the active preventive measures (APMs) during the peak of the epidemic. The aim of this study was to assess the use of APMs to diminish nosocomial transmission rates. The institutional ethics committee, following the guidelines of the Declaration of Helsinki, approved this retrospective clinical study (CIL No. 2020–63).

The study hospital employs 3142 healthcare workers and has 862 beds. During the outbreak in the Provence region of France (from 17^th March to 19^th April 2020), the hospital was part of the global governmental sanitary strategy, and opened 10 intensive care beds and 50 hospital beds (in four units) dedicated to patients with coronavirus disease 2019 (COVID-19) in addition to normal care units (e.g. nephrology, gynaecology, maternity).

The implementation of multiple APMs commenced on 23^rd March 2020 to minimize the risk of exposure to the virus. They consisted of social distancing [4], personal protective equipment [5] (e.g. handwashing, gloves, goggles, face shields and masks) and screening (e.g. temperature measurement). This article will discuss two additional APMs that were implemented systematically before surgery: all patients were subjected to a ld-CT chest scan and reverse transcriptase polymerase chain reaction (RT-PCR) testing of nasopharyngeal swabs. To perform each screening, a slot in the imaging department [6] was scheduled the day before surgery with triaging [6] on the basis on the patient's level of risk. Radiographers were trained to acquire images safely. In a standardized report, two radiologists (4 and 30 years of experience, respectively) assessed images in four grades: normal, mild, moderate or high suspicion of COVID-19. On this basis, a multi-disciplinary board determined personalized care. The authors consider that the APMs could have started being effective from 2^nd April 2020. The 10-day interval between APM implementation and potential effectiveness is due to the duration of incubation and the delay in getting RT-PCR results to the caregivers.

The effectiveness of APMs was assessed by monitoring the daily infection rate of caregivers, rather than patients, in an effort to maintain continuity of observation (due to the constant turnover of patients in the hospital). Infections among caregivers were monitored prospectively by occupational medicine. At the same time, intrahospital viral circulation was assessed daily by intensive care admissions and deaths due to COVID-19 in regional hospitals. Fisher's exact test (P<0.05) was used to compare these incidence rates before and after the implementation of APMs.

Pre-operative screening of 120 patients (69 males and 51 females), with a mean age of 58 [standard deviation (SD) 18] years, was part of the APMs. Twenty-six patients underwent full-dose CT for oncologic purposes. The mean dose of radiation used for ld-CT (N=94) was 33 (SD 6) mGy.cm. The turnaround time for CT was 66 min. One hundred and eleven CT scans were considered normal, whereas nine had a high or intermediate probability of infection (7.5%). Of these nine patients, two tested positive on RT-PCR and all were treated as infected. Surgery was postponed for three of these patients, and six underwent surgery in a dedicated operating room under optimized prevention conditions, without any adverse consequences at 2-month follow-up. The mean number of deaths increased from 8 (SD 7) before APM implementation to 17 (SD 7) after APM implementation (P<0.001), and the mean number of ICU hospitalizations increased from 168 (SD 114) before APM implementation to 413 (SD 29) after APM implementation (P<0.001). Noticeably, the mean infection rate among caregivers did not increase during this period, but instead decreased from 2.3 (SD 2.3) to 0.9 (SD 1) (P=0.02) (Figure 1 ).

Figure 1.

Incidence of cases before and after the implementation of active preventive measures (APMs). In (a), the curves represent in-hospital deaths (purple line) and intensive care admissions (/100) (orange line) in the Provence region of France. In (b), blue bars represent infected healthcare workers, and red bars represent a positive finding on low-dose computer tomography.

This supports the idea that APMs did limit nosocomial infection in caregivers [5]. Pre-operative ld-CT was part of the APMs and led to a 7.5% change in the way that patients were handled. Unfortunately, it is not possible to estimate the weight of each individual APM [5].

In agreement with Li et al. [7], the authors believe that operating theatres should be treated as high risk because the first clusters were discovered in the endocrinology and ear, nose and throat (ENT) surgical departments. Upper airway surgeries, such as ENT surgeries, are potentially highly contaminating procedures (e.g. involving tracheostomy).

To the authors' knowledge, the rate of negative PCR results and positive CT results has not been assessed to date (seven cases reported here). Very few false-negative CT findings have been reported: Ling et al. [8] reported that only 1% of RT-PCR-positive asymptomatic patients had negative CT findings.

During the COVID-19 outbreak, pre-operative screening using CT chest scans identified more cases with suspected COVID-19 than RT-PCR. These additional cases were included in the hospital's preventive measures. As ld-CT chest scans are straightforward to perform and provide rapid results, further research should be undertaken to determine their potential added benefit in the prevention of nosocomial infections.

Conflict of interest statement

None declared.

Funding sources

None.