Fig 1.
PRISMA flow diagram of study selection.
Table 1.
Pooled prevalence of fever in COVID-19 patients from different regions.
Fig 2.
Prevalence of chills in adult COVID-19 patients.
Fig 3.
Risks of fever prevalence in (A) severe or critical vs non-severe, (B) non-survived vs survived (recovered or discharged) and (C) ICU vs non-ICU adult COVID-19 patients.
Table 2.
Pooled prevalence and characteristics of fever in different subgroups of COVID-19 patients.
Fig 4.
Prevalence of (A) low (37.3–38.0°C), (B) medium (38.0–39.0°C) and (C) high-grade (>39.0°C) fever in adult COVID-19 patients.
Fig 5.
Risks of (A) low-grade fever (37·3–38·0°C) vs medium-grade fever (38·1–39·0°C), (B) high-grade fever (>39·0°C) vs low-grade fever (37·3–38·0°C), and (C) high-grade fever (>39·0°C) vs medium-grade fever (38·1–39·0°C) in adult COVID-19 patients.
Table 3.
Risk of different grades of fever in adult COVID-19 patients.
Fig 6.
Funnel plots on (A) adult and (B) paediatric COVID-19 studies.
Fig 7.
Galbraith plot identified eight outlier studies as potential sources of heterogeneity.
Table 4.
Sensitivity analyses.