Dear Editor,
Recently, the paper titled “Risk factors of critical & mortal COVID-19 cases: A systematic literature review and meta-analysis” was published in the Journal of Infection in April 2020. The results from Zheng et al. indicated that fever was negatively associated with the progression of COVID-19 such as severe illness and death (OR = 0.56, 95% CI [0.38–0.82], P = 0.003) and shortness of breath/dyspnea was positively associated with the progression of COVID-19 such as severe illness and death (OR = 4.16, 95% CI [3.13–5.53], P < 0.00001),1 which suggests that COVID-19 patients with fever may have a lower risk to develop to severe and critical disease outcomes and COVID-19 patients with dyspnea may have a higher risk to develop to severe and critical disease outcomes. However, Fu et al. observed that there was no statistically significant association between fever or shortness of breath and the severity of patients with COVID-19.2 To unambiguously identify the risk factors for predicting mortality in patients with COVID-19, we carry out a meta-analysis to evaluate whether fever and dyspnea (not included shortness of breath) were associated with the risk of mortality in COVID-19 patients.
This meta-analysis was carried out based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guideline. Li Shi and Ying Wang systematically searched the electronic databases, including Web of Science, Chinese National Knowledge Infrastructure (CNKI) and PubMed. These search engines were utilized to capture available literature by using the following three groups of keywords: “coronavirus 2019, 2019-nCoV, SARS-CoV-2, COVID-19″, “outcome, mortality” and “clinical”. The last search was conducted on May 4, 2020. Only articles reporting the number of COVID-19 patients with clinical symptoms of fever or dyspnea in the survival group and non-survival group were identified as eligible articles. All calculations were implemented with Stata 11.2 software. The pooled odds ratio (OR) with the corresponding 95% confidence interval (CI) was used to evaluate the risk of mortality in COVID-19 patients with fever or dyspnea. The robustness of the results was appraised by performing a sensitivity analysis. Both Begg's test and Egger's test were applied to evaluate publication bias.3 , 4
After selecting 1589 articles, 15 articles were finally obtained for this meta-analysis. As displayed in Table 1 , data on 2851 COVID-19 patients (2114 survivors and 737 non-survivors) were available in these articles. The sample size ranged from 27 to 663. Most of the articles were performed in China, with the exception of one in the UK.
Table 1.
Characteristics of the included studies.
| Author | Location | Case | Non-survival patients |
Survival patients |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n | Age, years | Male | Fever | Dyspnea | n | Age, years | Male | Fever | Dyspnea | |||
| Yang X et al. P: 32105632 | China | 52 | 32 | 64.6 ± 11.2 | 21 (65.6) | 31 (96.9) | 21 (65.6) | 20 | 51.9 ± 12.9 | 14 (70.0) | 20 (100.0) | 12 (60.0) |
| Zhou F et al. P: 32171076 | China | 191 | 54 | 69.0 (63.0–76.0) | 38 (70.4) | 51 (94.4) | NR | 137 | 52.0 (45.0–58.0) | 81 (59.1) | 129 (94.2) | NR |
| Cao J et al. P: 32239127 | China | 102 | 17 | 72 (63–81) | 13 (76.5) | 12 (70.6) | NR | 85 | 53 (47–66) | 40 (47.1) | 61 (71.8) | NR |
| Ruan Q et al. P: 32253449 | China | 150 | 68 | 67 (15–81) | 49 (72.1) | 59 (86.8) | 59 (86.8) | 82 | 50 (44–81) | 53 (64.6) | 68 (82.9) | 51 (62.2) |
| Deng Y et al. P: 32209890 | China | 225 | 109 | 69 (62–74) | 73 (67.0) | 95 (87.2) | 77 (70.6) | 116 | 40 (33–57) | 51 (44.0) | 94 (81.0) | 22 (19.0) |
| Zhang J et al. P: 32304745 | China | 663 | 25 | 67.1 (61–78) | 15 (60.0) | 19 (76.0) | 11 (44.0) | 638 | 59.1 (43–68) | 306 (48.0) | 508 (79.6) | 150 (23.5) |
| Wu C et al. P: 32167524 | China | 84 | 44 | 68.5 (59.3–75.0) | 29 (65.9) | 39 (88.6) | 29 (65.9) | 40 | 50.0 (40.3–56.8) | 31 (77.5) | 39 (97.5) | 21 (52.5) |
| Chen T et al. P: 32217556 | China | 274 | 113 | 68.0 (62.0–77.0) | 83 (73.5) | 104 (92.0) | 70 (61.9) | 161 | 51.0 (37.0–66.0) | 88 (54.7) | 145 (90.0) | 50 (31.1) |
| Wang L et al. P: 32240670 | China | 339 | 65 | 76 (70–83) | 39 (60.0) | 56 (87.5) | 38 (59.4) | 274 | 68 (64–74) | 127 (46.4) | 255 (93.4) | 100 (36.6) |
| Yuan M et al. P: 32191764 | China | 27 | 10 | 68 (63–73) | 4 (40.0) | 6 (60.0) | 10 (100.0) | 17 | 55 (35–60) | 8 (47.1) | 15 (88.2) | 1 (5.9) |
| Leung C et al. P: 32353398 | China | 154 | 89 | 75 (67–81) | 53 (59.6) | 44 (67.7)* | 25 (40.3)* | 65 | 68 (66–74) | 36 (55.4) | 58 (90.6)* | 4 (6.7)* |
| Wang D et al. P: 32354360 | China | 107 | 19 | 73.0 (64.0–81.0) | 16 (84.2) | 19 (100.0) | 15 (78.9) | 88 | 44.5 (35.0–58.8) | 41 (46.6) | 85 (96.6) | 20 (22.7) |
| Yan Y et al. P: 32345579 | China | 48 | 39 | 70.5 ± 10.1 | 30 (76.9) | 36 (92.3) | 30 (76.9) | 9 | 64.7 ± 7.3 | 3 (33.3) | 7 (77.8) | 3 (33.3) |
| Wang K et al. P: 32361723 | China | 296 | 19 | 65.6 ± 12.6 | 11 (57.9) | 10 (52.6) | NR | 277 | 46.0 ± 14.4 | 129 (46.6) | 203 (74.9)* | NR |
| 44 | 14 | 69.0 ± 13.4 | 10 (71.4) | 12 (100.0)* | NR | 30 | 48.8 ± 14.2 | 14 (46.7) | 27 (90.0) | NR | ||
| Tomlins J et al. P: 32353384 | UK | 95 | 20 | 77 (72–85) | 12 (60.0) | 12 (60.0) | NR | 75 | 74 (56–82) | 48 (64.0) | 56 (74.7) | NR |
All values are n (%), median (IQR), or mean±SD. P, PMID.
data missing for patients; NR, not reported.
We found that dyspnea was significantly associated with higher mortality in COVID-19 patients on the basis of 11 studies with 2091 cases (OR = 4.34, 95% CI [2.68–7.05], P < 0.001; I2 = 69.2%, P < 0.001, random-effects model) (Fig. 1 A). However, we did not observe a significant association between fever and the risk of mortality in patients with COVID-19 on the basis of 15 studies with 2818 cases (OR = 0.74, 95% CI [0.50–1.09], P = 0.127; I2 = 38.0%, P = 0.062, random-effects model) (Fig. 1B). As presented in sensitivity analysis, none of the individual studies significantly effected the overall OR, which proved the robustness of our results (Figs. 1C and D). No evidence of publication bias was provided by Begg's test (dyspnea: P = 0.350 and fever: P = 0.964, respectively) and Egger's test (dyspnea: P = 0.294 and fever: P = 0.854, respectively).
Fig. 1.
The pooled odds ratio (OR) with the corresponding 95% confidence interval (CI) on the relationship between dyspnea (A) and fever (B) and the risk of mortality in COVID-19 patients. Sensitivity analysis for evaluating the relationship between dyspnea (C) and fever (D) and the risk of mortality in COVID-19 patients.
To our knowledge, the most common clinical symptoms were fever, cough, fatigue and dyspnea in COVID-19 patients.5., 6., 7. Zheng et al. demonstrated that the proportion of fever was significantly lower in critical/mortal group compared with the non-critical group,1 which suggests that fever may protect COVID-19 patients from developing to severe and critical disease outcomes. Fu et al. reported that the prevalence of fever in critical group was slightly higher than that in the non-severe group (80.8%, 95% CI [41.1–100.0]) vs. (71.2%, 95% CI [23.8–99.9]), but the difference was not statistically significant.2 Our present study showed that fever was not significantly associated with the risk of mortality in COVID-19 patients. In addition, our study suggested that dyspnea was positively associated with the risk of mortality in COVID-19 patients. Taken together, dyspnea, rather than fever, is recommended as an indicator of poor outcome in COVID-19 patients, further well-designed studies with larger sample sizes are needed to validate the findings of our current study.
Contributors
LS, YDW, and HYY designed the study. LS and YW screened the literature and extracted the data. LS performed the meta-analysis and wrote the manuscript. HYY, YDW and GCD provided guidance and reviewed the manuscript. All authors have read and agreed the final manuscript.
Declaration of Competing Interest
All authors report that they have no potential conflict of interest.
Funding
This study was supported by the National Natural Science Foundation of China (grant number 81973105).
References
- 1.Zheng Z., Peng F., Xu B., Zhao J., Liu H., Peng J. Risk factors of critical & mortal COVID-19 cases: a systematic literature review and meta-analysis. J Infect. 2020 doi: 10.1016/j.jinf.2020.04.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Fu L., Wang B., Yuan T., Chen X., Ao Y., Fitzpatrick T. Clinical characteristics of coronavirus disease 2019 (COVID-19) in China: a systematic review and meta-analysis. J Infect. 2020 doi: 10.1016/j.jinf.2020.03.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Begg C.B., Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50(4):1088–1101. [PubMed] [Google Scholar]
- 4.Egger M., Davey Smith G., Schneider M., Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629–634. doi: 10.1136/bmj.315.7109.629. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Rodriguez-Morales A.J., Cardona-Ospina J.A., Gutierrez-Ocampo E., Villamizar-Pena R., Holguin-Rivera Y., Escalera-Antezana J.P. Clinical, laboratory and imaging features of COVID-19: a systematic review and meta-analysis. Travel Med Infect Dis. 2020 doi: 10.1016/j.tmaid.2020.101623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sun P., Qie S., Liu Z., Ren J., Li K., Xi J. Clinical characteristics of hospitalized patients with SARS-CoV-2 infection: a single arm meta-analysis. J Med Virol. 2020 doi: 10.1002/jmv.25735. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hu Y., Sun J., Dai Z., Deng H., Li X., Huang Q. Prevalence and severity of corona virus disease 2019 (COVID-19): a systematic review and meta-analysis. J Clin Virol. 2020;127 doi: 10.1016/j.jcv.2020.104371. [DOI] [PMC free article] [PubMed] [Google Scholar]