The authors have declared that no competing interests exist.
To understand the clinical characteristics of COVID-19 patients with clinically diagnosed bacterial co-infection (CDBC), and therefore contributing to their early identification and prognosis estimation.
905 COVID-19 patients from 7 different centers were enrolled. The demography data, clinical manifestations, laboratory results, and treatments were collected accordingly for further analyses.
Around 9.5% of the enrolled COVID-19 patients were diagnosed with CDBC. Older patients or patients with cardiovascular comorbidities have increased CDBC probability. Increased body temperature, longer fever duration, anhelation, gastrointestinal symptoms, illness severity, intensive care unit attending, ventilation treatment, glucocorticoid therapy, longer hospitalization time are correlated to CDBC. Among laboratory results, increased white blood cell counting (mainly neutrophil), lymphocytopenia, increased procalcitonin, erythrocyte sedimentation rate, C-reaction protein, D-dimer, blood urea nitrogen, lactate dehydrogenase, brain natriuretic peptide, myoglobin, blood sugar and decreased albumin are also observed, indicating multiple system functional damage. Radiology results suggested ground glass opacity mixed with high density effusion opacities and even pleural effusion.
The aged COVID-19 patients with increased inflammatory indicators, worse lymphopenia and cardiovascular comorbidities are more likely to have clinically diagnosed bacterial co-infection. Moreover, they tend to have severer clinical manifestations and increased probability of multiple system functional damage.
The Corona Virus Disease 2019 (COVID-19), caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), has affected most countries all over the world since its first case in end of 2019. The genomic characteristics of SARS-CoV-2 was initially reported by Lu and colleagues, suggesting this coronavirus had enveloped RNA, resembling severe acute respiratory syndrome coronavirus (SARS-CoV) in both structural and homological ways [
During this COVID-19 attack, besides the primary infection of SARS-CoV-2, many other complications are emerging, contributing greatly to the mortality. Among these, co-infection plays a crucial role, threatening many COVID-19 patients’ lives [
To decrease the mortality of COVID-19 patient as much as possible, the early recognition and managements of bacterial co-infection seem rather indispensable. In the present study, 905 confirmed COVID-19 patients from 7 different cities in both Hunan and Hubei province, China, were enrolled for investigation. The cases with suspected bacterial co-infection were clinically diagnosed according to some specific diagnostic criteria (see below). The clinical data of these clinically diagnosed bacterial co-infection (CDBC) patients were collected for further retrospective analyses in order to throw light upon the characteristics of them, facilitating early recognition and managements.
The initial hospital admittance time of the enrolled patients range from January 10th to February 28th, 2020. Every participant of this study has either signed written or made oral consensus (the aims of the conducting study, the privacy protection policy, the duties and rights of enrolled patients were all told before oral consensus was made and the oral consensus was recorded by a voice recorder), as we were short of health personnel to make every participant sign a written consent during the viral outbreak. And this was permitted by the Ethics Committee of The Second Xiangya Hospital of Central South University. COVID-19 diagnoses were made according to the diagnostic criterion from the
All the 905 enrolled patients were from the separated COVID-19 units of these 7 hospitals. Computer tomography (CT) scan evaluations were made by at least 2 specialists from the radiology department. The SARS-CoV-2 nucleic acid RT-PCR test quality control was performed by specialists from the clinical laboratory department. The collected clinical data include: the demographic descriptions, main symptoms, comorbidities, regular laboratory results (e.g. blood routine examination, renal function, liver function, myocardial enzyme), main treatments, etc. To be specific, all the laboratory and radiological results, from the beginning to the end, of those CDBC patients were collected when the bacterial co-infection diagnoses were clinically made (diagnostic criteria were described below). However, the test results of those non-CDBC patients were collected with the most abnormal ones during the whole hospitalization period. For privacy reasons, the raw data of these patients are not presented.
COVID-19 diagnosis was made by SARS-CoV-2 RNA RT-qPCR test with nasopharyngeal swabs before hospitalization of each participant. The RNA detection kits were provided by Sansure Biotech (Changsha, China) and being manipulated by specialized clinical laboratory technicians according to the manufacturer’s protocol.
The clinical diagnosis criteria of bacterial co-infection are as follow (patients meet all of these criteria shall be considered CDBC): a, newly increased WBC counting (>9.5*10^9/L), with a majority of neutrophil; b, newly increased airway purulent secretion; c, increased serum procalcitonin; d, typical peripheral ground glass opacity mixed with increased density effusion opacities; e, effective empiric anti-bacterial therapy.
COVID-19 patients who could be considered discharged when meeting all the discharging criteria from the 7th version of the
Mechanical ventilation indications: a, failed high-flow nasal cannula (HFNC) or non-invasive positive pressure ventilation (NIPPV) therapy when 300>PaO2/FiO2≥150; b, PaO2/FiO2≤150 with failed short term NIPPV.
The continuous variables were denoted as median (interquartile range, IQR) and comparisons were performed using the Mann-Whitney test. The categorical variables were denoted as n (%) and compared by using the Chi-square test or Fisher’s exact Chi-square test. The association between potential risk factors and outcomes was estimated by using logistic regression. Kaplan–Meier methods were used for survival curve plotting and examined by log-rank test. The association between clinically diagnosed bacterial co-infection and all-cause mortality was examined by using multivariate Cox regression model. All analyses were performed using R software (The R Foundation,
All the enrolled 905 confirmed COVID-19 patients were from 7 different centers (Yueyang, Shiyan, Shaoyang, Zhuzhou, Huaihua, Huanggang and Loudi). 86 out of 905 patients (9.5%) were CDBC according to our diagnostic criteria described above. Among the basic demographic features, CDBC patients tend to be older and the rest were found no differences (
All patients | COVID-19 | P value | ||
---|---|---|---|---|
Without CDBC | With CDBC | |||
905 | 819 | 86 | ||
47 (35–57) | 46 (34–56) | 62 (49–75) | <0.001 | |
0.007 | ||||
463 (51%) | 431 (53%) | 32 (37%) | NA | |
442 (49%) | 388 (47%) | 54 (63%) | NA | |
23 (21–25) | 23 (21–25) | 23 (21–26) | 0.88 | |
672 (74%) | 595 (73%) | 77 (90%) | <0.001 | |
714 (79%) | 649 (79%) | 65 (76%) | 0.43 | |
424 (47%) | 376 (46%) | 48 (56%) | 0.08 | |
16 (2%) | 12 (1%) | 4 (5%) | 0.06 | |
134 (15%) | 92 (11%) | 42 (49%) | <0.001 | |
425 (47%) | 378 (46%) | 47 (55%) | 0.13 | |
142 (16%) | 129 (16%) | 13 (15%) | 0.88 | |
48 (5%) | 44 (5%) | 4 (5%) | >0.99 | |
117 (13%) | 93 (11%) | 24 (28%) | <0.001 | |
81 (75–90) | 81 (75–90) | 86 (76–95) | 0.06 | |
20 (19–20) | 20 (19–20) | 20 (20–23) | <0.001 | |
98 (96–99) | 98 (97–99) | 95 (93–96) | <0.001 | |
125 (120–132) | 124 (119–130) | 128 (120–140) | 0.02 | |
80 (72–84) | 80 (72–84) | 80 (73–87) | 0.33 | |
37.8 (37.0–38.5) | 37.7 (37.0–38.5) | 38.5 (37.6–39.0) | <0.001 | |
473 (52%) | 447 (55%) | 26 (30%) | <0.001 | |
375 (41%) | 335 (41%) | 40 (47%) | ||
57 (6%) | 37 (5%) | 20 (23%) | ||
8 (5–12) | 8 (5–12) | 10 (6–14) | 0.03 |
We analyzed several common chronic disease types among the enrolled patients. Hypertension and coronary heart disease are more prevalent among CDBC patients, along with diabetes. However, chronic lung diseases, liver diseases, kidney diseases and tumor were not different between CDBC and no-CDBC patients (
All patients | COVID-19 | P value | ||
---|---|---|---|---|
Without CDBC | With CDBC | |||
141 (16%) | 105 (13%) | 36 (42%) | <0.001 | |
84 (9%) | 71 (9%) | 13 (15%) | 0.05 | |
37 (4%) | 29 (4%) | 8 (9%) | 0.01 | |
27 (3%) | 22 (3%) | 5 (6%) | 0.17 | |
8 (1%) | 8 (1%) | 0 (0%) | >0.99 | |
8 (1%) | 6 (1%) | 2 (2%) | 0.17 | |
11 (1%) | 9 (1%) | 2 (2%) | 0.28 |
Generally speaking, COVID-19 patients have normal WBC counting (20% of all the enrolled COVID-19 patients had increased WBC count). However, according to our findings, once WBC of COVID-19 patients increased notably, with a majority of neutrophils, they were highly suspected to have bacterial co-infection (43% of the CDBC had increased WBC count while non-CDBC only 3%). Moreover, lymphocytes count could even be further decreased (
All patients | COVID-19 | P value | ||
---|---|---|---|---|
Without CDBC | With CDBC | |||
5.0 (3.9–6.7) | 4.9 (3.8–6.2) | 10.5 (6.0–12.6) | <0.001 | |
179 (20%) | 21 (3%) | 37 (43%) | <0.001 | |
668 (74%) | 628 (77%) | 40 (47%) | 0.007 | |
58 (6%) | 170 (21%) | 9 (10%) | ||
3.6 (2.4–8.6) | 3.3 (2.3–6.6) | 10.7 (7.9–14.6) | <0.001 | |
1.2 (0.8–1.6) | 1.2 (0.8–1.6) | 0.6 (0.4–1.0) | <0.001 | |
166 (18%) | 135 (16%) | 31 (36%) | <0.001 | |
595 (66%) | 545 (67%) | 50 (58%) | 0.43 | |
144 (16%) | 139 (17%) | 5 (6%) | 0.08 | |
190 (160–240) | 130 (119–144) | 131 (118–147) | 0.96 | |
190 (160–240) | 201 (154–257) | 180 (131–258) | 0.2 |
Procalcitonin (PCT) of CDBC patients were found notably increased, supporting its widely accepted diagnostic role for bacterial infection. Erythrocyte sedimentation rate (ESR) and C-reaction protein (CRP), common inflammatory indicators as they are, were both increased in CDBC patients. In blood clotting indexes, prothrombin time (PT) and D-dimer were found increased compared with non-CDBC patients, along with aspartate amino transferase (AST) and bilirubin in liver function tests. Moreover, blood urea nitrogen (BUN), lactate dehydrogenase (LDH), brain natriuretic peptide (BNP), Myohemoglobin and random blood sugar (RBS) were all increased in CDBC patients (
All patients | COVID-19 | P value | ||
---|---|---|---|---|
Without CDBC | With CDBC | |||
0.05 (0.05–0.10) | 0.05 (0.04–0.10) | 0.29 (0.10–0.57) | <0.001 | |
886 (98%) | 811 (99%) | 75 (87%) | <0.001 | |
19 (2%) | 8 (1%) | 11 (13%) | 0.007 | |
29.0 (13.0–48.0) | 28.0 (12.0–45.8) | 46.3 (28.0–82.0) | 0.002 | |
7.2 (5.0–23.8) | 6.0 (5.0–16.2) | 59.2 (21.0–110.1) | <0.001 | |
12.0 (10.8–12.9) | 11.8 (10.8–12.8) | 13.4 (12.4–14.6) | <0.001 | |
31.4 (27.3–35.7) | 31.6 (27.4–35.8) | 29.4 (25.6–34.8) | 0.22 | |
0.3 (0.2–0.6) | 0.3 (0.2–0.5) | 1.8 (0.6–4.1) | <0.001 | |
23 (16–41) | 22 (15–40) | 28 (21–55) | 0.007 | |
24 (19–34) | 22 (18–32) | 36 (25–56) | <0.001 | |
60.0 (41.5–96.5) | 58.0 (40.5–93.5) | 73.5 (46.0–159.4) | 0.09 | |
11.0 (4.0–38.5) | 10.5 (2.6–35.4) | 14.0 (8.0–51.5) | 0.09 | |
12.7 (8.7–20.8) | 12.3 (8.7–19.5) | 24.3 (10.6–33.8) | <0.001 | |
4.3 (3.9–4.6) | 4.3 (3.9–4.6) | 4.3 (3.8–5.2) | 0.38 | |
139.3 (137.5–141.2) | 139.2 (137.7–141.0) | 139.8 (136.4–143.8) | 0.18 | |
12.7 (8.7–20.8) | 3.9 (3.2–4.8) | 7.8 (4.4–16.4) | <0.001 | |
67.3 (55.1–81.4) | 66.0 (55.0–80.2) | 74.8 (62.6–113.2) | 0.002 | |
190 (160–240) | 183 (160–226) | 337 (241–510) | <0.001 | |
0.01 (0.01–0.03) | 0.01 (0.01–0.03) | 0.03 (0.02–0.03) | 0.001 | |
148.0 (25.0–420.5) | 99.0 (18.0–251.9) | 427.0 (156.7–706.0) | <0.001 | |
39.2 (33.3–43.6) | 40.3 (34.9–44.0) | 30.3 (28.5–33.6) | <0.001--- | |
35.0 (30.0–62.8) | 31.0 (30.0–51.0) | 119.9 (67.7–379.6) | <0.001 | |
5.4 (4.9–6.7) | 5.3 (4.9–6.0) | 8.3 (6.5–10.6) | <0.001 |
Every enrolled patient had at least 2 CT scan results during hospitalization. CDBC patients always have ground-glass opacities (GGO) mixed with high density effusion opacities (HDEO), and pleural effusion could even be found in some CDBC cases (
The red curve indicated non-CDBC patients and the blue-green curve indicated CDBC patients. The survival curves were plotted based on the day of hospital admittance (day 0). We further performed Cox regression analysis to investigate the association between CDBC and risk of in-hospital mortality with adjustment of confounding factors, including gender, age and pre-existing comorbidities. Compared with non-CDBC patients, those with CDBC were subjected to a higher risk of mortality with a hazard ratio (HR) 8.21 (95% CI: 4.46 to 15.10), that is, patients with CDBC have 721% more death risk than the non-CDBC.
All patients | COVID-19 | P value | ||
---|---|---|---|---|
Without CDBC | With CDBC | |||
523 (58%) | 495 (60%) | 28 (33%) | <0.001 | |
6 (1%) | 4 (0%) | 2 (2%) | 0.1 | |
30 (3%) | 24 (3%) | 6 (7%) | 0.046 | |
36 (4%) | 27 (3%) | 9 (10%) | 0.001 | |
53 (6%) | 19 (2%) | 34 (40%) | <0.001 | |
66 (7%) | 19 (2%) | 47 (55%) | <0.001 | |
15 (10–20) | 14 (10–20) | 18 (13–24) | <0.001 | |
57 (6%) | 18 (2%) | 39 (45%) | <0.001 | |
224 (25%) | 161 (20%) | 63 (73%) | <0.001 |
Ground-glass opacity (GGO); High density effusion opacities (HDEO); Pleural effusion (PE).
To evaluate the risk factors of CDBC, a logistic regression analysis was conducted. Decreased albumin account and glucocorticoid treatment could be the best predictors for CDBC, with the sensitivities 82% and 73% respectively. Moreover, advanced age, high WBC account, lymphopenia, PCT and CRP could all be CDBC predictors with reasonably high sensitivities. The cardiovascular comorbidities also could be considered predictors, however, the sensitivity is relatively low (
Variables | OR (95% CI) | AUC(95% CI) | P-value | Sensitivity |
---|---|---|---|---|
1.06 (1.04, 1.07) | 0.73 (0.67, 0.79) | <0.001 | 0.58 | |
1.58 (1.43, 1.75) | 0.81 (0.74, 0.87) | <0.001 | 0.63 | |
0.10 (0.04, 0.24) | 0.77 (0.69, 0.86) | <0.001 | 0.66 | |
5.11 (1.98, 13.19) | 0.83 (0.76, 0.91) | <0.001 | 0.66 | |
1.03 (1.02, 1.05) | 0.83 (0.75, 0.91) | <0.001 | 0.68 | |
2.24 (0.82, 6.06) | 0.51 (0.49, 0.54) | 0.11 | 0.06 | |
1.88 (0.99, 3.55) | 0.53 (0.49, 0.57) | 0.053 | 0.15 | |
2.14 (0.46, 10.08) | 0.51 (0.49, 0.52) | 0.33 | 0.02 | |
2.79 (1.23, 6.32) | 0.53 (0.50, 0.56) | 0.014 | 0.09 | |
0.79 (0.74, 0.85) | 0.84 (0.78, 0.90) | <0.001 | 0.82 | |
11.19 (6.74, 18.60) | 0.77 (0.72, 0.82) | <0.001 | 0.73 | |
27.53 (14.70, 51.57) | 0.69 (0.63, 0.74) | <0.001 | 0.4 | |
50.74 (27.23, 94.54) | 0.76 (0.71, 0.81) | <0.001 | 0.55 |
Up to date, few studies have reported the clinical characteristics of COVID-19 patients with bacterial co-infection, however, co-infection of influenza virus [
Viruses are one of the most common respiratory pathogens, and bacterial co-infection was not rare during primary respiratory viral infection, including influenza of many types (e.g. H1N1, H7N9, etc) [
According to our results, COVID-19 patients typically present with a normal white blood cell count, however, a suddenly increased WBC count during hospitalization shall alert the occurrence of bacterial co-infection. Lymphopenia, which could be resulted from a severe immune system disorder, is common in COVID-19 patients [
Some limitation of the present study merit consideration. Initially, it is a real-world retrospective study, and the data collected were limited. A prospective analysis could be of more confidence and less bias. Secondly, the timing of CDBC onset was not distinguished, that is, CDBC could be community, hospital acquired or ventilation associated. And this might make the results less specific to HAP. Moreover, due to the lack of understanding of this highly contagious novel virus, the pathogens of CDBC have not been identified universally for safety reasons at the beginning of the global pandemic. But currently, microbial cultivation of respiratory secretions were recommended to make studies more informative and guiding the antibiotic therapy accordingly. Furthermore, a collection of dynamic laboratory and radiological findings could provide more confidence in the results.
To summarize, 86 out of 905 (9.5%) confirmed COVID-19 patients were CDBC according to our diagnosis criteria (see methods part). CDBC patients tend to be more aged, and having more severe clinical manifestations with increased WBC account and worse lymphopenia. Inflammatory indicators are increased, along with increased biochemical laboratory results indicating multi-system function damage tendency. And GGO mixed with HDEO could be the feature of CDBC radiological results. Moreover, advanced age, high WBC account, lymphopenia, PCT, CRP, cardiovascular comorbidities, the utilities of glucocorticoid and ventilation are all risk factors of CDBC. To better identify and manage CDBC in early stage, more prospective studies regarding CDBC are extremely necessary.