The chronic effects of SARS-CoV-2 infection, including postacute sequelae of SARS-CoV-2 infection (PASC) symptoms and unexplained physical weakness that persists long-term, continue to affect a significant portion of COVID-19 survivors. In this research, we developed a prediction model utilizing clinical data from patients admitted to the First Affiliated Hospital of Jinzhou Medical University for COVID-19 treatment, following China's "Category B and B Management" implementation and full societal reopening. LASSO regression analysis identified crucial factors associated with long COVID, culminating in the selection and external validation of four clinical variables—the CCI, CT score, oxygen requirement, and lymphocyte percentage—using clinical data from Panjin Central Hospital.
The CCI, which combines age and comorbidities across all age groups and systemic conditions, was first proposed by Charlson in 1987[10] to predict long-term mortality and has since been widely adopted for forecasting long-term prognosis and survival. Given its efficacy, simplicity, and applicability and considering that the occurrence of long COVID is also part of the prognosis of patients with COVID-19, the CCI was introduced in this study for the first time. Statistical findings revealed that a higher CCI was correlated with an increased likelihood of long COVID.
While most published studies have focused solely on either age or comorbidities, our review identified fourteen study sets that considered age, with twelve confirming the predictive value of age. This research encompasses data from China, Italy, the United States, the United Kingdom, and other countries[4, 5, 11–14], covering most designated medical centers for treating the novel coronavirus.
With respect to comorbidities, determining whether long COVID symptoms originate from preexisting conditions, new symptoms caused by COVID-19, or a combination of both is challenging. Consequently, long COVID definitions emphasize duration over distinguishing between these factors. A 2022 JAMA article [12] highlighted that an increased number of comorbidities elevates the risk of long COVID. Specifically, a study from Italy noted the importance of an allergic constitution among various comorbidities. Another study in the BMJ from the United States noted that individuals with preexisting interstitial lung disease are more likely to develop severe COVID-19.
The ongoing impact of coronavirus on T-cell function leads to immune escape, allowing the virus to persist at low concentrations for extended periods. Data from past coronavirus outbreaks, such as SARS and MERS [15, 16], show that some patients continue to suffer from lung damage after the infection has resolved. The novel coronavirus, characterized by its low virulence and viral load, does not typically cause systemic symptoms but can persist in certain organs, leading to specific clinical symptoms. A recent Nature study [17] demonstrated that most patients struggle to completely eliminate the virus during the acute phase, with approximately 0.1% − 0.5% of patients harboring the virus for at least 60 days. The continued presence of the virus weakens the immune system's ability to clear it, particularly in individuals with a history of medical conditions, thus increasing the likelihood of long-term effects from the novel coronavirus.
Current research often overlooks the simultaneous consideration of age and comorbidities. A nomogram developed by the National Infectious Disease Center team at Fudan University in Shanghai, China [11], included five risk factors. Being over 75 years of age, having chronic kidney disease, and suffering from chronic lung disease were identified as the most significant risk factors. This highlights the particular importance of age and comorbidities in predicting the risk of severe COVID-19 in the Chinese population.
The CT score, initially utilized to assess the prognosis of the interstitial pneumonia index [18], has been adopted for its quantitative analysis, high accuracy, and ability to provide rapid diagnostics. Many research teams have adapted and revised it [19] as a tool for real-time assessment of the severity of lung involvement in COVID-19 patients. Compared with other indicators of disease progression, the majority of research teams [20] recognize the CT score as the most accurate index reflecting disease progression, with more severe COVID-19 cases being more likely to develop into long COVID [13].
This study highlights the CT score as the optimal indicator of disease progression. For the first time, the CT score was applied to COVID-19 patients and included in the final prediction model through LASSO regression analysis. In the longitudinal follow-up of patients with COVID-19, persistent imaging changes are often observed, which are closely correlated with the peak CT score during the acute phase [21]. Current research primarily describes these persistent imaging changes, with pulmonary fibrosis-like changes being the most frequently observed. A follow-up study of the initial set of COVID-19 patients from Wuhan Jinyintan Hospital[21] statistically identified a CT score exceeding 18 during the acute phase of COVID-19 as an independent risk factor for developing pulmonary fibrosis-like changes during follow-up. These persistent imaging changes correlate with long-standing clinical symptoms, and long COVID can be considered an umbrella term encompassing these clinical manifestations.
According to the novel coronavirus infection diagnosis and treatment plan (10th version), severe and critical patients are more focused on supportive treatments than ordinary and mild patients are. Respiratory system support therapy, specifically oxygen therapy, is a primary treatment strategy. In a multicenter study including 2,433 cases [13], 71.6% of hospitalized patients received oxygen therapy, and 0.9% underwent mechanical ventilation. These studies have established a connection between oxygen therapy, mechanical ventilation, and the persistence of fatigue, which is one of the most common long COVID symptoms [4]. Our research focused on long COVID among hospitalized patients with acute SARS-CoV-2 infection, indicating that symptomatic oxygen support is provided on the basis of the severity of the patient's condition. Our findings suggest that a lower oxygen requirement increases the likelihood of developing long COVID. This conclusion may be attributed to two factors. First, under conditions of insufficient blood oxygen saturation, oxygen is preferentially supplied to vital organs such as the brain, heart, and lungs. Other organs may remain hypoxic, leading to long-term tissue decompensation, chronic cellular damage, and the persistence of symptoms and discomfort after discharge, which are recognized as clinical manifestations of long COVID. Second, for patients with severe conditions, timely high-flow oxygen support is often provided. A significant proportion of these patients might die before discharge or during follow-up; thus, they were excluded from this study, introducing bias. Therefore, the incidence of long COVID appears to be lower among recipients of high-flow oxygen therapy. On the other hand, prolonged oxygen therapy can cause lung damage. Several post-COVID-19 follow-up studies have revealed [21] interstitial lung lesions in the CT scans of patients with long COVID. It remains unclear whether this damage is attributable to COVID-19 itself or to the side effects of oxygen therapy received during hospitalization. This ambiguity highlights the need for further research to fully understand the relationship between oxygen therapy and long COVID. The association between oxygen requirements and long COVID essentially reflects a mismatch between the oxygen demand of the body and the administration of oxygen therapy, which is highly subjective. This study, which is based on clinical data from two medical centers, has limitations, and the relationship between oxygen therapy and long COVID warrants additional investigation.
The percentage of lymphocytes in the blood is a quick indicator used to identify the source of infection, with lymphocyte percentages < 20% indicating viral infection. A lower lymphocyte percentage is associated with a greater likelihood of viral infection. Research has indicated that adults infected with COVID-19 exhibit lymphopenia, increased platelet counts, and elevated lactate dehydrogenase levels, with a decrease in lymphocyte count in severe cases indicating more severe disease and a greater risk of mortality [22]. In this study, the lymphocyte percentage was included as an indicator of COVID-19 severity. This metric, derived from a routine blood test available to all patients, holds significant potential for widespread application.
This research utilized data from the first major outbreak to develop a prediction model for long COVID syndrome in patients with severe COVID-19, offering a scientific basis for early identification with clinical relevance. Nonetheless, this study has several limitations: (1) This was a multicenter retrospective study, and further large-scale multicenter prospective studies are needed for validation. (2) An assessment of lung involvement in long COVID severity is lacking, including further subgroup analysis predictions. (3) It does not include data from intensive care units, particularly from mechanically ventilated patients.