2.1 Study site
Medical data were retrospectively analyzed for COVID-19 patients who were admitted between December 2022 and February 2023 to receive COVID‐19 related treatments at a large medical center in Chin. It is a university-affiliated, tertiary care center featuring 4,864 beds, 88 inpatient wards, and 54 clinical departments. The study period was chosen to correspond to when China declared its “reopening policy”, relieving strict virus control measures that had been in place since 2019; this led to a sharp rise in the number of infections during the next two months. During the study period, all elective surgeries were cancelled; only emergency surgeries were permitted.
The medical staff were volunteered to take part in this program. There was at least one pulmonogist in each medical team. Both junior and senior medical staff from different specialties were trained through a streamlined program to provide respiratory care. Normal work shifts were modified to ensure adequate staffing in the remodeled departments while also allowing workers to take breaks. There was a standard guideline for all medical staff on how to treat the COVID-19 patients. During the study period, remodeled respiratory departments (n = 42) worked in parallel with conventional respiratory departments (n = 4) to provide care to COVID-19 patients.
2.2 Inclusion and exclusion criteria
Patients were enrolled in the study if they were diagnosed between December 2022 and February 2023 with SARS-CoV-2 infection based on a PCR test and required hospitalization because of COVID-19. Patients were excluded if they were younger than 18, not have hospitalization, not diagnosed with SARS-CoV-2 infection, or admitted and discharged on the same day.
2.3 Data collection and study outcomes
The primary outcome in this study was all-cause in-hospital mortality. Secondary outcomes were the length of hospitalization (total length of stay in any part of the hospital), rates of patients who were critically or non-critically ill, rate of complications, healthcare costs, and rates of COVID-19 related intervention and medications.
Data on demographic, clinical, laboratory, and healthcare costs were extracted from the central hospital database. Disease diagnosis, comorbidities, and complications were extracted as the International Classification of Diseases (ICD-10) codes. Extracted data were carefully checked against the original database to ensure the absence of errors. Oxygen interventions included air, non-invasive ventilation, high-flow nasal oxygen, prone positioning, and invasive mechanical ventilation with tracheal intubation. Medications included antiviral therapy (nirmatrelvi, azvudine), glucocorticoids (dexamethasone, methylprednisolone), the interleukin-6 inhibitor tocilizumab, and the Janus kinases (JAK) inhibitor (baricitinib). We analyzed healthcare costs for tests and products that attending physicians considered necessary to treat COVID-19 and comorbidities; these costs were borne by participants’ health insurance and local government. Separate costs were calculated for drugs; therapies, including supplemental oxygen and hemopurification; diagnostic laboratory tests, including microbiology, hematology, biochemistry, and blood gas analysis; diagnostic imaging, including X-ray, computed tomography (CT), and B-mode ultrasonography; consultations with specialists; and surgery, including use of the operating theater and intensive care unit, as well as the costs of the surgery itself.
2.4 Propensity score matching
Instead of analyzing all study participants, we analyzed a subset of patients treated in each kind of respiratory department based on propensity score matching for the following variables, which differed significantly between the two groups and have been associated with COVID-19 severity: sex, age, body mass index, smoking history and comorbidities such as diabetes, hypertension, chronic lung diseases or cancer [24; 25]. Patients treated in a conventional respiratory department were matched 1:3 with those treated in a remodeled department. The success of the matching procedure in balancing out baseline differences between the two patient groups was assessed using propensity score histograms.
2.5 Statistical analysis
All analyses were performed using R 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria). Data for all continuous variables were assessed for normal distribution using the Shapiro-Wilk test. Normally distributed continuous data were presented as mean and standard deviation (SD), while skewed continuous data were reported as the median and interquartile range (IQR). Categorical data were reported as n (%). Inter-group differences were assessed for significance using the Mann-Whitney U test, Student’s t test, chi-squared test, or Fisher test, as appropriate. Statistical results that were associated with two-tailed p < 0.05 were defined as significant.
Effects after propensity score matching were estimated using generalized equations in which matched group identifiers served as cluster labels. Effect sizes were reported as odds ratios (ORs) with 95% confidence intervals (CIs).
2.6 Ethics approval and participant consent
The study protocol was approved by the ethics committee of our hospital [2023(30)], and the study was performed in accordance with the Declaration of Helsinki. The Ethics Committee waived the requirement for informed consent because patients or their legal guardians provided written consent for the patient’s anonymized medical data to be analyzed and published for research purposes. At the Ethics Committee’s request, we posted an announcement on the hospital website giving participants the opportunity to opt out of the study.