Infectious Disease · Medical Literature
Assessing the prevalence of anosmia and taste dysfunction in COVID-19 patients in Saudi Arabia: a systematic review and meta-analysis.
Abstract
Patients with coronavirus disease-2019 (COVID-19) frequently report olfactory and gustatory dysfunctions, including anosmia and ageusia. Multiple studies have assessed the prevalence of these symptoms in COVID-19 patients in Saudi Arabia. This systematic review and meta-analysis aimed to estimate the pooled prevalence of anosmia and taste dysfunction in confirmed COVID-19 patients in Saudi Arabia across inpatient, outpatient, and community settings, from March 2020 to March 2023. A systematic review and meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines. No review protocol was prospectively registered prior to this review. A comprehensive search was conducted across PubMed, ScienceDirect, CINAHL, MEDLINE, and Google Scholar (supplementary) from 1 March 2020 to 1 March 2023. Data were pooled using a random-effects model. The methodological quality of included studies was evaluated using the Joanna Briggs Institute (JBI) critical appraisal tools for prevalence studies. Of 230 records identified across all databases, 156 remained after deduplication; 24 met the eligibility criteria, encompassing 26,068 COVID-19 patients (58.2% female). The pooled prevalence of anosmia was 44% (95% CI: 37-52%) and of taste dysfunction was 46% (95% CI: 39-54%). Subgroup analyses comparing validated versus non-validated assessment instruments revealed no statistically significant difference in prevalence estimates. Anosmia and taste loss are common manifestations of COVID-19 in Saudi Arabia. The findings should be interpreted with caution given the methodological heterogeneity of included studies, reliance on self-reported measures in the majority of studies, and the absence of a prospectively registered protocol. Future studies using objective validated instruments are warranted.
Clinical relevance
What is the pooled prevalence of anosmia and taste dysfunction among confirmed COVID-19 patients in Saudi Arabia across inpatient, outpatient, and community settings from March 2020 to March 2023? __SNAPSHOT__ {"studyDesign":"Systematic review and meta-analysis (PRISMA 2020-compliant); random-effects model; no prospectively registered protocol","setting":"Saudi Arabia; inpatient (hospitalized), outpatient, and community settings; literature search across PubMed, ScienceDirect, CINAHL, MEDLINE, and Google Scholar","sampleSize":"24 studies included; encompassing 26,068 confirmed COVID-19 patients (58.2% female); 24 studies contributed to anosmia analysis, 20 studies to taste dysfunction analysis","population":"Confirmed COVID-19 patients (PCR-positive) of all ages across Saudi Arabia; predominantly female (58.2%); spanning early pandemic through Omicron-era periods (March 2020–March 2023)","intervention":"Exposure: SARS-CoV-2 infection confirmed by RT-PCR; assessment of olfactory dysfunction (anosmia) and gustatory dysfunction (taste loss/ageusia) via validated and non-validated self-report instruments","comparator":"Not applicable (single-group prevalence meta-analysis; no uninfected comparator group)","primaryOutcome":"Pooled prevalence of anosmia and pooled prevalence of taste dysfunction among confirmed COVID-19 patients","followUp":"Not reported (cross-sectional prevalence estimates; no longitudinal follow-up applied across included studies)","fundingSource":"Not reported"}
Key findings
Of 230 identified records, 156 remained after deduplication, and 24 studies met eligibility criteria. The pooled prevalence of anosmia was 44% (95% CI: 37–52%) and of taste dysfunction was 46% (95% CI: 39–54%). Substantial heterogeneity was present across all analyses (I² >87% in all subgroups), consistent with the high methodological and clinical variability across included studies. Subgroup analyses comparing validated versus non-validated assessment instruments revealed no statistically significant difference in prevalence estimates. Two pre-specified sensitivity analyses supported robustness of the primary estimates: (1) restricting to studies using validated instruments only (n=5 for anosmia) yielded a pooled anosmia prevalence of 50% (95% CI: 31–69%); (2) restricting to adequately quality-rated studies (JBI ≥6/9 criteria; n=21 anosmia, n=18 taste dysfunction) yielded pooled anosmia prevalence of 43.9% (95% CI: 35.3–52.7%) and taste dysfunction prevalence of 44.3% (95% CI: 36.0–52.8%). Publication bias assessment via Egger's test was non-significant for both anosmia (p=0.245) and taste dysfunction (p=0.705). NNT/NNH is not applicable to this prevalence meta-analysis. No adverse event data were reportable from this study design. No p-values for primary pooled estimates were reported.
Practice implications
This meta-analysis establishes that anosmia and taste dysfunction were common manifestations of COVID-19 in Saudi Arabia during the pandemic period studied, affecting approximately 44% and 46% of confirmed patients, respectively. However, the review does not establish diagnostic utility, as sensitivity, specificity, and likelihood ratios for these symptoms as diagnostic markers are not derivable from prevalence data alone. The findings apply to confirmed COVID-19 patients in Saudi Arabia from March 2020 to March 2023 and may not be generalizable to post-Omicron patient cohorts or to populations outside the Arabian Peninsula, given evidence that olfactory symptom rates declined substantially with Omicron variants globally. The majority of included studies relied on non-validated self-report instruments, which likely introduces recall and ascertainment bias. Subgroup and meta-regression analyses were exploratory, underpowered due to small numbers of studies in individual subgroups, and must not be used for formal inference. The absence of a prospectively registered protocol introduces risk of selective outcome reporting. Only the abstract was available for the initial query; the full text was retrieved and informed this summary, though clinicians should review the complete article for forest plots, individual study data, and supplementary analyses. __LIMITATIONS__ Extreme statistical heterogeneity (I² >87% in all subgroups) severely limits the interpretability of the pooled estimates and indicates substantial unexplained between-study variability in patient populations, care settings, SARS-CoV-2 variants, and symptom ascertainment methods.,The majority of included studies used non-validated, self-report instruments for olfactory and gustatory assessment; objective psychophysical testing (e.g., Sniffin' Sticks, UPSIT) was used in very few studies, likely introducing ascertainment bias.,No prospective protocol registration (PROSPERO) was completed prior to the review, raising concern for post-hoc outcome selection and analytic flexibility.,Studies spanned March 2020 to March 2023, covering multiple SARS-CoV-2 variants (original, Delta, Omicron) with markedly different olfactory symptom profiles; variant-stratified analyses were not reported.,Cross-sectional design of most included studies precludes assessment of symptom duration, recovery rates, or progression to long COVID olfactory dysfunction.,Subgroup analyses and meta-regression were underpowered given small numbers of studies per subgroup and must be considered hypothesis-generating only.,Single-country focus (Saudi Arabia) limits generalizability; no comparison to non-Saudi populations was conducted.,Funding source was not reported, limiting assessment of potential industry or institutional influence.,Study population was predominantly female (58.2%), which may not represent the broader Saudi COVID-19 case distribution and could influence prevalence estimates given sex differences in olfactory symptom reporting. __BOTTOM_LINE_CATEGORY__ confirmatory __BOTTOM_LINE__ In a PRISMA-compliant meta-analysis of 24 Saudi Arabian studies encompassing 26,068 confirmed COVID-19 patients, pooled prevalence of anosmia was 44% (95% CI: 37–52%) and taste dysfunction was 46% (95% CI: 39–54%), with sensitivity analyses broadly consistent with these estimates. Findings should be interpreted with caution given extreme heterogeneity (I²>87%), near-universal reliance on non-validated self-report instruments, absence of a prospectively registered protocol, and the multi-variant study period. These data characterize the burden of chemosensory symptoms in a defined population and setting but do not establish diagnostic utility or support changes to current clinical screening practice.