Treatment and Outcomes in Patients With COVID-19 and Competing Acute Surgical Diseases in Moscow

Overview

The COVID-19 pandemic has posed a significant challenge to surgical practice. This study aims to investigate the structure of hospitalizations, treatment outcomes, and surgical tactics in patients with COVID-19 and competing different acute surgical diseases (ASD). The research is based on real-world data (RWD) from a large multidisciplinary hospital repurposed as a COVID-19 center. The study analyze the mutual impact of these two acute conditions on patient outcomes, develop prognostic models for in-hospital mortality, and evaluate changes in surgical approaches during the pandemic.

This is a single-center, ambidirectional observational cohort study based on routinely collected clinical data from the Moscow Multidisciplinary Clinical Center "Kommunarka," Moscow, Russia. The study evaluates the organization and outcomes of care for patients with acute surgical conditions during two distinct periods of hospital activity: the COVID-19 pandemic period and the post-pandemic period. The study was designed to address the methodological challenges of analyzing patients with two potentially competing acute conditions: COVID-19 and an acute surgical disease. The clinical spectrum includes acute abdominal diseases, thrombotic and hemorrhagic conditions, soft-tissue infections, thoracic complications, and other conditions requiring surgical assessment or treatment. No diagnostic, surgical, interventional, or medical treatment was assigned or modified by the study protocol. All clinical decisions were made by the treating teams as part of routine care. Data were obtained from hospital information systems and routine electronic medical records. Source data included demographic characteristics, diagnoses, operative and interventional procedures, selected clinical severity indicators, comorbid conditions, hospitalization dates, discharge status, and other variables required for the planned analyses. The principal unit of analysis was a hospitalization episode rather than an individual data row. Repeated technical records and duplicated entries relating to the same hospitalization were identified and resolved before analysis. A multistage data-management process was used to transform poorly structured real-world data into clinically interpretable analytical cohorts. The initial screening procedure used diagnostic codes, relevant text fields, procedure information, and predefined clinical terms to identify potentially eligible hospitalization episodes. More specific analytical datasets were subsequently created using stricter diagnostic definitions and clinical review. Certain clinically complex disease groups required manual review of medical records, reassessment of the final diagnosis, clarification of the surgical condition, or verification using imaging and procedural information. Quality-control procedures included assessment of duplicate records, repeated hospitalizations, inconsistent diagnostic classification, incomplete variables, implausible values, and discrepancies between diagnosis and procedure fields. The process also included review of hospitalizations that may have been incorrectly included or excluded by the initial screening algorithm. Changes to cohort composition, classification rules, and derived variables were documented through versioned datasets and change records. A cohort or dataset passport was used to describe the origin of the data, unit of analysis, cohort-formation rules, level of clinical validation, data limitations, and intended analytical use. Missing data were not automatically interpreted as absence of the corresponding clinical condition. The amount and pattern of missingness were assessed for relevant variables. Statistical analyses were conducted using the available data for each predefined analytical task. Sensitivity analyses were used where appropriate to assess whether the main findings were affected by alternative cohort definitions, stricter phenotyping rules, or differences in data completeness. The pandemic and post-pandemic cohorts were analyzed separately and comparatively. The inter-period analysis was intended to assess whether the structure of acute surgical hospitalizations, treatment patterns, and clinical outcomes differed between the specialized pandemic hospital setting and subsequent routine hospital activity. Both broad and more specific cohort definitions were used to evaluate the robustness of the observed differences. The study also included the development of an organizational and methodological model for conducting research based on real-world data within a medical organization. The model describes the sequence of research activities, distribution of professional roles, clinical validation procedures, data-quality controls, standard operating procedures, dataset documentation, and change management. This component represents a methodological and organizational output of the study and does not constitute an evaluation of the clinical effectiveness of an implemented institutional program. Because this is a nonrandomized observational study, associations between patient characteristics, treatment strategies, historical periods, and outcomes are interpreted cautiously. Surgical intervention may reflect disease severity, clinical selection, and treatment necessity and is not interpreted as an independently assigned exposure or a proven cause of outcome.