Prevalence of skin manifestations in patients with COVID-19: a systematic review and meta-analysis

Abstract

Background: COVID-19 presents extrapulmonary manifestations that can aid in the diagnosis. Skin manifestations have been reported but their characteristics are not yet clear. Health professionals need information about its prevalence and main characteristics.

Methods: This systematic review followed the PRISMA criteria. The protocol was registered in the PROSPERO (number CRD42020193173). Seven electronic databases and the gray literature were searched independently by two researchers. Observational analytical studies that presented data on the prevalence of skin manifestations in patients aged 19 or older with COVID-19 were included. Prevalence estimates were synthesized through a meta-analysis using random-effects models. Association meta-analysis and comparisons were performed for individual characteristics.

Results: We included 31 studies with 10,934 patients, of which 10,121 tested positive for COVID-19. The general prevalence of skin manifestations was 29% (95% CI: 17.0-43.0; I²: 99%), the most in Africa, with a mean duration between 7 and 9 days and the most frequently affecting feet+hands (75%) and the trunk (71%). Patients with mild/moderate COVID-19 had more of chilblain-like+pernio-like lesions (97%) and inflammatory lesions (86%) than patients with severe or critical COVID-19. Manifestations of vascular origin were only in elderly patients and were significant with the severity of COVID-19 (p = 0).

Conclusion: The global prevalence of skin manifestations is similar to other signs and symptoms of COVID-19. Skin assessment should be considered when investigating and diagnosing COVID-19 in adult and elderly patients.Systematic review registration: PROSPERO, identifier CRD42020193173, https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42020193173.

Keywords: COVID-19; adult; aged; meta-analysis; skin manifestation; systematic review.

Figure 1

Flow diagram of search and selection process. Adapted from Page et al. (18).

Figure 2

General prevalence of skin manifestations in adults and elderly patients with a positive laboratory test for COVID-19 and by type of study (p < 0.05 for all prevalence data). 95% CI, 95% confidence interval; I², inconsistency index.

Figure 3

Prevalence of skin manifestations in adults and elderly patients with a positive laboratory test for COVID-19 by geographic region (p < 0.05 for all prevalence data). 95% CI, 95% confidence interval; I², inconsistency index. *The prevalence of skin manifestations for each geographic region was calculated considering the total number of patients positive for COVID-19 per continent (n value shown in the figure); prevalences that did not present an I²-value were calculated from only one study.

Figure 4

Prevalence of COVID-19 skin manifestations by type. (A) General prevalence of skin manifestations (p < 0.05 for all prevalence data). (B) General prevalence of skin manifestations by body area and type of injury (n = 383). 95% CI, 95% confidence interval; I², inconsistency index; NS, Not Specified. *Head = head and face; Trunk = neck, chest, back, abdomen, and/or hips; Hands + Feet = palms, soles, hands, feet, and/or fingers; Entire Body = disseminated lesions (more than 2 body segments); Extremities = arms, legs, hands, and feet. (Body image taken from free-access Vecteezy resources).

Figure 5

Prevalence of inflammatory and vascular skin manifestations by morphological type. (A) Inflammatory manifestations (rash) (n = 758). (B) Vascular manifestations (n = 131). 95% CI, 95% confidence interval; I², inconsistency index; NS, not specified.

Figure 6

Forest plot of the association analysis of the occurrence of skin manifestations among patients with a positive diagnostic test about (A) sex (female or male) and (B) severity of COVID-19 (mild/moderate or severe/critical). Chi², chi-square test; I², inconsistency index; M-H, Mantel–Haenszel Test; OR, odds ratio; Q, Cochran’s Q test; Z, Z-test.

Figure 7

Proportion and comparison between the severity of COVID-19 and the type of skin manifestation presented. *The p-value was based on the chi-square test (Chi²).