Postoperative Complications and Mortality With In-Hospital COVID-19 Omicron Infection After Surgery

Overview

Postoperative complications and mortality in patients with COVID-19 Omicron infection who have undergone specialized thoracic surgery are scarce. Subsequently, the patient cohort was divided into two groups for comparative analysis: Group 1 (G1), which comprised patients who acquired nosocomial omicron infection after surgery, and Group 2 (G2), which comprised patients who remained uninfected with omicron during their hospitalization period. Propensity score matching (PSM) analysis was conducted using the PSMATCH function in SPSS 27 to assess the incidence of perioperative complications and mortality rates between both groups.

Data of patients who underwent thoracic surgery at the Tangdu Hospital of Air Force Medical University and the Second Affiliated Hospital of Lanzhou University between December 1, 2022, and January 15, 2023, were collected. The data encompassed various clinical parameters, including age, sex, medical history, smoking history, disease classification, the extent of surgical resection, time of postoperative omicron diagnosis, duration of symptoms, COVID-19 classification (The symptoms of novel COVID-19 infection were classified as mild, medium, severe and critical according to the diagnosis and treatment of novel COVID-19 infection version 10), postoperative complications, and 30-day postoperative outcomes. Subsequently, the patient cohort was divided into two groups for comparative analysis: Group 1 (G1), which comprised patients who acquired nosocomial omicron infection after surgery, and Group 2 (G2), which comprised patients who remained uninfected with omicron during their hospitalization period. Hospital-acquired omicron infection was defined as having a positive reverse transcription polymerase chain reaction (RT-PCR) test for omicron from a nasopharyngeal or throat swab. Before admission, all the patients tested negative for omicron within 24 hours. In cases where patients presented with symptoms such as fever, daily RT-PCR testing was performed. A single positive result was considered sufficient evidence of infection. According to the Clavien-Dindo classification, surgical complications are categorized as follows: Grade I encompasses medical interventions that deviate from the normal postoperative course but do not require drugs, surgery, endoscopy, or radiological intervention. Acceptable medical interventions may include the administration of antiemetics, antipyretics, analgesics, diuretics, electrolyte management, and physical therapy. This grade also includes superficial surgical site infections. Grade II involves complications that necessitate pharmacological treatment in addition to Grade I interventions, such as blood transfusions or total parenteral nutrition. Grade III involves complications that require surgical, endoscopic, or radiological interventions. Grade IV consists of life-threatening complications that require intermediate care or treatment in the intensive care unit. This grade includes central nervous system complications, such as cerebral hemorrhage, ischemic stroke, and subarachnoid hemorrhage, but excludes transient ischemic attacks. Finally, Grade V represents patient death. The thoracic surgeries are classified as follows. Pulmonary surgery refers to the removal of lung tissue, irrespective of the extent of resection. Digestive system surgery encompasses procedures targeting diseases of the esophagus or gastroesophageal junction. The other surgery includes surgeries that do not involve the removal of lung tissue, addressing conditions affecting the chest wall, mediastinum, and other related areas. Propensity score matching (PSM) analysis was conducted using the PSMATCH function in SPSS 27 to assess the incidence of perioperative complications and mortality rates between both groups. Considering these factors (age, sex, medical history, smoking history, disease classification, the extent of surgical resection), propensity matching was performed using a one-to-one nearest-neighbor matching algorithm with a match caliper standard deviation of 0.02. Each G1 patient was then paired with one G2 patient. Standardized mean differences were calculated to evaluate the balance between the matched groups, with differences exceeding 0.1 indicating potential imbalance. Categorical variables were described as frequencies and proportions and analyzed using the chi-square test to determine any statistical significance (P \< 0.05). To confirm the factors contributing to the occurrence of complications, a logistic regression model was constructed, which incorporated confounding variables and examined their association with the occurrence of complications.