Introduction

SARS-CoV-2 (SARS2) infects and eventually disrupts ACE2 plus TMPRSS2 expressing cells of the lower respiratory tract, causing severe complications such as diffuse alveolar damage and bacterial or fungal bronchopneumonia [1, 2]. Intubation under general anesthesia poses a significant risk of postoperative pulmonary complications [3]. Surgical operations requiring general anesthesia have the potential to cause severe outcomes in patients with perioperative SARS2 infection.

Although some studies did not find significant associations between surgery and adverse outcomes among perioperative SARS2 PCR-positive patients [4], other studies have also shown significant associations [5]. One study found that surgery <7 weeks after recovering from COVID-19 was significantly associated with mortality [6]. The American Society of Anesthesiologists recommends waiting 4–12 weeks after a confirmed COVID-19 infection (COVID-19 and elective surgery (asahq.org)). These recommendations resulted in screening all patients with respiratory PCR testing, even for patients without symptoms, and deferring elective surgery for PCR-positive patients.

The burden of screening all patients with PCR tests and the medical consequences of deferring surgery for asymptomatic PCR-positive patients is a significant health issue. The outcomes of screening surgical patients are of medical interest, but the data are sparse.

We evaluated the outcomes of PCR-positive patients who underwent surgery under general anesthesia.

Material and methods

Study design and data collection

We conducted a single-center retrospective cohort study among surgical patients who underwent mechanical ventilation for general anesthesia, and compared perioperative SARS2 PCR-positive patients with controls. Symptomatic patients underwent deferred surgery unless they had a life-threatening surgical condition and were thus not included in the cohort. Symptomatic patients were defined using WHO clinical criteria (https://apps.who.int/iris/bitstream/handle/10665/360579/WHO-2019-nCoV-Surveillance-Case-Definition-2022.1-eng.pdf). We analyzed a propensity score-matched (1:2) subgroup.

We obtained data from the hospital's electronic records by filtering the hospitalized surgical patients from January 2021 to May 2022. We excluded patients with SARS2 multiplex PCR-positive results before 7 days and 5 days after surgery. Thus, we kept patients with PCR-positive test results between 7 days before and 5 days after surgery and patients with negative PCR test results as controls.

The outcome was all-cause in-hospital mortality.

We compared age, sex, and underlying diseases, such as diabetes, hypertension, cardiovascular disease, chronic renal disease, and malignant disease, between patients who survived and those who died. We also compared the American Society of Anesthesiologists (ASA) scores recorded by an anesthesiologist before surgery.

The ASA scores were as follows: (1) The patient was entirely healthy and fit, (2) the patient had a mild systemic disease, (3) the patient had a severe systemic disease but was not incapacitated, (4) the patient had an incapacitating illness that constantly threatens life, and (5) a moribund patient not expected to live 24 h with or without surgery [7].

This study was approved institute ethical committee with the number of 2022/0587.

Statistics

Continuous variables were presented as the median and interquartile range (IQR; 0.25–0.75) and compared using the Wilcoxon rank sum test. Count data were presented as numbers and percentages and compared using Pearson's Chi-squared test or, where required, Fisher's exact test.

Propensity scores were estimated using a logistic model with independent variables, including age, sex, ASA score, emergency surgery, diabetes, hypertension, cardiovascular disease, chronic renal disease, and malignant disease. We matched PCR-positive surgical patients in a 1:2 ratio to PCR-negative surgical patients using the "matchit" package using the nearest neighbor method with replacement. The balance between the before and after matching variables was estimated using mean differences and Kolmogorov–Smirnov tests.

We constructed an initial model by including variables that were found to be significant (p < 0.05) in the univariate comparisons. However, ASA score was included in the model as a categorical variable (score >3). Underlying disease was dichotomized as having one or more of the following: diabetes, hypertension, cardiovascular disease, chronic renal disease, and malignant disease. The final model was created using backward elimination of non-significant variables. Multicollinearity and interactions were tested.

We also constructed a nomogram from the final model using the predictive performance of variables using the rms package.

Statistical analyses were conducted using the open-source R version R-4.2.1 (R Foundation for Statistical Computing, Vienna, Austria).

Results

During the study period, 17,569 patients underwent surgery under general anesthesia. The characteristics of the cohort are summarized in Table 1. Briefly, of the 194 patients who died in hospital, 11 (5.7%, p < 0.001) were SARS2 PCR positive. ASA scores were higher among patients who died, and emergency surgery was more frequent (n (%), 128 (0.66); p < 0.001). Patients who died were older than those who survived (median (IQR), 68 years (50, 78); <0.001) and had more underlying diseases. A total of 217 patients were PCR positive; among these, 52 had stayed in the COVID-19 wards and received oxygen support of varying intensity. Of the 52 patients, 21 were treated in COVID-19 wards after surgery (median (days), IQR: 6 (2, 10)), while 31 were treated in COVID-19 wards before surgery (median (days), IQR: 6 (2, 9)), and among these, five and three patients died, respectively (p = 0.321).

Table 1 Characteristics of the cohort
Full size table

Due to the imbalance between PCR-positive and PCR-negative patients (PCR-positive n (%); 217 (0.1%)), we conducted comparisons on a propensity score-matched subgroup. We generated propensity scores and matched SARS2 PCR-positive patients at a ratio of 1:2 to the controls. Figure 1 displays the before and after balance between matching variables compared with the mean differences and Kolmogorov–Smirnov statistics [8]. The variables were balanced after matching. We first applied a full model and then backward-eliminated variables to obtain sufficient discriminative power and shrinkage values. Table 2 presents the point estimates for the full and final models. In the final model, the point estimate for PCR-positives was not significant (log(OR) (95%CIs), p; 0.86 (− 0.13, 1.9), 0.086). However, an ASA score >3, older age, and surgical emergency were associated with all-cause in-hospital mortality. The discriminative power of the final model was 0.85 (C-index = 0.85), and the shrinkage factor was 0.94 ([model x2 − (d.f. − 1)]/model x2) [9, 10].

Fig. 1
figure 1

Covariate balance before and after adjusting. The dotted lines represent threshold values for balance used in absolute mean differences and Kolmogorov–Smirnow statistics

Full size image
Table 2 Point estimates from the final logistic model
Full size table

Figure 2 shows the nomogram constructed from the predictive performance of the variables in the final model. Older age and ASA score >3 significantly impacted the outcome, while the SARS2 PCR test was only marginally associated with mortality.

Fig. 2
figure 2

Nomogram constructed from the predictions of the final logistic model

Full size image

Discussion

This study failed to show a significant difference between PCR-positive and PCR-negative surgical patients regarding all-cause in-hospital mortality. In other words, this study does not support the preoperative screening of asymptomatic patients using the SARS2 PCR test. In our hospital, all symptomatic patients defer surgery, unless a life-threatening emergency occurs. We studied perioperatively PCR-positive asymptomatic patients and compared them to PCR-negative controls. One study reported a 20.5% mortality rate among postoperative patients who developed COVID-19 [11]. Another study reported the significance of the timing between PCR-positive testing and surgery [6]. Some studies failed to identify significant outcome differences [4, 12]. However, neither of these studies compared perioperative asymptomatic PCR-positive patients to PCR-negative controls and thus did not explicitly address the outcomes of screening all asymptomatic preoperative patients.

We found that age and an ASA score >3 were significant risk variables associated with mortality. However, emergency surgery and a positive SARS2 PCR test only marginally affected adverse outcomes. One study found that age and being SARS2 PCR positive had a significant prognostic value in cardiac surgery [13]. However, in this study, most patients were symptomatic, and a history of heart failure was more common among PCR-positive patients. Preoperative risk stratification for asymptomatic SARS2 PCR-positive patients is of scientific interest [14].

There are some limitations to our study. First, this was a retrospective study; therefore, we could not extract the detailed baseline characteristics of the patients. Second, this was a single-center study, which limits the generalization of the findings. Third, this study did not analyze all postoperative pulmonary complications. Lastly, since we have no gold standard test to identify true positives, we must consider that false positive PCR test results might exist and negatively affect statistical inferences.

Finally, our study found that surgery among asymptomatic PCR-positive patients was not associated with increased mortality. Further studies are required to provide more robust results.