This was a retrospective cohort study. All 1,513 patients with COVID-19 from the hospital isolation ward who presented to the Public Health Clinical Centre of Chengdu from January 16, 2020, to September 30, 2022, were retrospectively recruited ( Figure 1 , Table 1 ). The study was approved by the Public and Health Clinical Centre of Chengdu Ethics Committee (ethics approval number: PJ-K2020-26-01). Written informed consent was waived by the Ethics Commission of the designated hospital because this study was related to emerging infectious diseases.

The inclusion criteria were as follows: no sex limit; age ≥18 years old; COVID-19; and inpatient isolation and treatment time >1 day.

The exclusion criteria were as follows: age<18 years old and isolation and treatment time <1 day.

The criteria for COVID-19 clinical typing, disease diagnosis and cure were in accordance with the seventh Trial Version of the Novel Coronavirus Pneumonia Diagnosis and Treatment Guidance (15).

The diagnosis criteria were cases with one of the following etiological pieces of evidence: real-time fluorescence reverse transcription-polymerase chain reaction (RT-PCR) detected the positive nucleic acid of the new coronavirus and sequencing of viral genes.

The typing criteria were as follows: (1) asymptomatic infection indicated that there were no clinical symptoms and no pneumonia manifestations on imaging; (2) the light type indicated that the clinical symptoms were mild, and there were no pneumonia manifestations on imaging; (3) the common type indicated that the clinical symptoms included fever and respiratory tract, and pneumonia could be seen on imaging; (4) the severe type indicated that the patients had any of the following criteria: respiratory distress, RR≥30 times/min; in the resting state, oxygen saturation ≤ 93%; arterial blood oxygen partial pressure (PaO2)/oxygen concentration (FiO2)≤300mmHg (1mmHg=0.133kPa), living in areas with high altitude (over 1000 meters above sea level), and PaO2/FiO2 should be corrected according to the following formula: PaO2/FiO2*[atmospheric pressure(mmHg)/760]; pulmonary imaging showed that lesions with significant progress over 50% within 24–48h were managed as heavy; (5) the critical illness type criteria included one of the following conditions: respiratory failure occurs and mechanical ventilation is needed; Shock occurs; and combining other organ failure requires intensive care units (ICU) monitoring.

The cured discharge standard was as follows: the body temperature returned to normal for more than 3 days; respiratory symptoms improved significantly; lung imaging showed a significant improvement in acute exudative lesions; and two consecutive sputum, nasopharyngeal swabs and other respiratory specimens tested negative for nucleic acid (sampling time at least 24 h apart).

The relevant serological assays were performed by Enzyme-Linked Immunosorbent Assay (ELISA) in the laboratory of the hospital.

Among the 1,513 COVID-19 cases, 1102 and 411 cases were divided into the no immunity group (primary infection and no vaccination) and the immune breakthrough group (previous infection or vaccination), respectively ( Figure 1 , Table 1 ).

Among the 411 immune breakthrough cases, 73 patients with previous infection were assigned to the natural immunity group, 322 patients who had been vaccinated (no distinction was made between vaccine type and dose) were assigned to the acquired immunity group, and 16 patients who had both previous infection and been vaccinated were assigned to the mixed immunity group ( Figure 1 , Table 1 ).

Among the 1,513 COVID-19 cases, 1,471 noncritical patients (patients with asymptomatic infection, with light and with common clinical type) were assigned to the noncritical group, and 42 critical patients (patients with severe and with critical illness clinical type) were assigned to the critical group ( Figure 1 , Table 1 ).

Among the 1,513 COVID-19 cases, 1505 surviving patients were assigned to the survival group, and 8 dead patients were assigned to the death group ( Figure 1 , Table 1 ).

The disease severity included critical illness (COVID-19 patients with severe or critical illness clinical type) and noncritical illness (COVID-19 patients with asymptomatic infection, light or common clinical type). The prognosis included death and survival within four weeks after admission. The coronavirus negative conversion time was the time from onset to the first negative nucleic acid test meeting the discharge criteria.

The data were collected from a subset of patients treated at Chengdu Public Health Clinical Medical Center from January 2020 to November 2022. All data of 1,513 cases, including clinical data, laboratory data and demographic data, were collected to establish databases. Researchers strictly controlled the accuracy, completeness and authenticity of all data.

SPSS 26.0 (SPSS, Chicago, IL, USA) and GraphPad Prism 8 (GraphPad, CA, USA) were used for statistical analyses. Measurement data with a normal distribution are presented as the mean and standard deviation, and measurement data with a nonnormal distribution are presented as the median and interquartile range (IQR). The categorical data are expressed as a percentage or proportion. Data with a normal distribution and homogeneity of variance between multiple groups were compared using one-way or two-way ANOVA, and further comparison between two groups was performed using the least significant difference (LSD) t test. Data with a normal distribution and homogeneity of variance between two groups were compared using the independent samples t test. Enumeration data are presented as percentages or proportions, and data between two or multiple groups were compared using a chi-square test. Analysis of influencing factors of disease severity and prognosis was performed using binary logistic regression analysis. Receiver operating characteristic (ROC) analysis was used to assess lymphocytes and subsets to distinguish non critical from severe COVID-19 patients. P<0.05 was considered statistically significant.

Patients and the public were not involved in the development of the research questions or in the design of the study. Patients received verbal and written information about the study; however, they were not involved in the recruitment of subjects or the conduct of the study. Additionally, the burden of the intervention was assessed by the investigators. The participants were assessed for eligibility, and data collection was performed. Dissemination of the general results (without personally identifying data) will occur on demand. The Ethics Committee of the Public Health Clinical Centre of Chengdu approved this study (ethics approval number: PJ-K2020-26-01). Written informed consent was waived by the Ethics Commission of the designated hospital because this study is related to emerging infectious diseases.

A total of 1,513 patients with COVID-19 were included in this study. Their demographic and clinical characteristics are listed in Table 1 . The median age of all patients included in the study was 35 years, and males accounted for the majority (71.0%). The median coronavirus negative conversion time was 11.0 days, and the duration of hospitalization was 15.0 days.

In addition, 966 (63.40%) patients had comorbidities, 374 (25.0%) patients had one comorbidity, 233 (15.6%) patients had two comorbidities, 340 (22.8%) patients had three or more comorbidities, and 547 (36.6%) patients had no comorbidities. Among them, 42 (2.8%) patients had critical illness, 1,471 (97.2%) patients had noncritical illness, 1505 patients survived, and only 8 (0.5%) patients died.

Imported cases accounted for 85.0% of the total, while domestically transmitted cases made up the remaining 15.0%. Regarding immune status, 411 (27.2%) patients had preexisting immunity by previous infection or vaccination, and 1102 (72.8%) patients had primary infection. Among vaccinated patients, 13 (0.9%) patients accepted one dose, 235 (15.5%) patients accepted two doses, 84 (5.6%) patients accepted three doses and 6 (0.4%) patients accepted four doses.

In the immune breakthrough group, the proportion of domestically transmitted cases was significantly higher than that in the no immunity group ( Table 2 ) (p<0.05), while the critical illness rate was significantly lower than that in the no immunity group ( Table 2 ) (0.24% vs. 3.72%, p<0.05). The mortality rate was slightly lower than that in the no immunity group, although the difference was not statistically significant ( Table 2 ) (p=0.083).

The age ( Figure 2A ) was slightly younger than that in the no immunity group (p<0.05), and there was no significant difference in BMI ( Figure 2B ) or duration of hospitalization ( Figure 2C ) between the two groups (all p>0.05). However, in the immune breakthrough group, the coronavirus negative conversion time ( Figure 2D ) was significantly shorter than that in the no immunity group (p<0.001).

In addition, in the no immunity group, CD3+ counts, CD3+CD4+ counts, CD3+CD8+ counts and lymphocyte counts at admission were higher than those in the immune breakthrough group ( Figures 3A–C, E ) (all p<0.05), but all of them then decreased to the lowest level at the first week and were always significantly lower than those in the immune breakthrough group from the first week to the fourth week ( Figures 3A–C, E ) (all p<0.01). The ratio of CD3+CD4+ to CD3+CD8+ cells between the two groups from onset to the fourth week after onset was always similar to each other ( Figure 3D ) (p>0.01). Moreover, in the no immunity group, total antibody levels and specific IgG levels from onset to the fourth week were always significantly lower than those in the immune breakthrough group ( Figures 3F, H ) (all p<0.001). In the no immunity group, the specific IgM levels were always lower than those in the immune breakthrough group from the first week to the fourth week, but a significant difference was found only at the first week after onset ( Figure 3G ).

Metabolic diseases such as fatty liver, diabetes, hypertension, hyperlipidemia, etc. were the most common in both groups, with no difference (p all <0.05) ( Table 3 ). The incidence of cardiovascular disease was higher in the no immunity group (p<0.05) ( Table 3 ). Besides, the incidence of hypoxemia (p<0.001) and cardiovascular disease (p<0.05) was significantly higher in the no immunity group than in the immune breakthrough group ( Table 3 ).

Compared to the natural immunity subgroup and the acquired immunity subgroup, age was slightly younger (p<0.05), and the duration of hospitalization was obviously shorter than that in the mixed immunity subgroup (p<0.01) ( Table 4 ). The coronavirus negative conversion time in the mixed immunity subgroup was also slightly shorter than that in the other two groups, but the difference was not statistically significant ( Table 4 ) (p=0.057).

Moreover, in the acquired immunity subgroup, CD3+ counts, CD3+CD4+ counts, CD3+CD8+ counts and lymphocyte counts at admission were the highest and then decreased to the lowest level at the first month among the three groups. All of them in the other two groups showed an increase from admission to the first month after admission, and those in the mixed immunity group reached their highest levels ( Figures 4A–D ). In the mixed immunity subgroup, the total antibody levels and specific IgG levels were the highest at admission and then decreased slightly over the following month ( Figures 4E, F ). Those in the natural immunity subgroup were the lowest at admission, with a sharp decrease during the following month ( Figures 4E, F ). Only those in the mixed immunity subgroup rose from second place on admission to first place at one month ( Figures 4E, F ).

For critical cases, the age was significantly older, the number of comorbidities and the imported cases were higher (p all=0.000). Most T-lymphocyte subsets (except CD3+CD8+% and CD19+%) and all antibody levels (p all<0.05) were significantly higher in critical cases, whereas the level of CD56+% (p=0.024) was higher in non-critical cases ( Table 5 ). Similarly, all the death cases were imported cases, with an average age of 77.5 years (p=0.000), a slightly higher proportion of females (p=0.034), an increased BMI index (p=0.013), and all combined with more than 3 diseases (p=0.000), without a history of previous infection. CD3+, CD3+CD4+, CD3+CD8+, LY, LY%, CD19+ and all antibody levels in non-critical cases were higher (p all <0.01) ( Table 5 ).

The influencing factors for disease severity by multiple linear regression analysis were age, lymphocyte percentage, source of cases and CD3+CD8+ counts ( Table 6 ). Moreover, the influencing factors for prognosis were age, sex, and source of cases ( Table 6 ).

According to the ROC analysis, the total antibody, IgM and IgG levels showed good utility for predicting critical COVID-19 patients ( Table 7 ). The areas under the curve of total antibody, IgM and IgG for disease severity were 0.854, 0.904 and 0.794, respectively ( Table 7 , Figure 5 ). The sensitivities were 98.10%, 76.40%, 96.10%, while the specificities were 75.00%, 100.00%, 75.00% ( Table 7 ).