Our survey application provided rapid information during the early pandemic, but the quality of the collected data was unknown. Here, we summarize this survey data, and compare to see how it agrees with known global trends.
Demographic Distribution by Region, Gender, and Age
Our filtered dataset includes results from 10,954 individuals tested for COVID-19. Table 1 summarizes the demographic statistics of these individuals, categorized by governorate, gender, and age group. The largest proportion of individuals resided in the North governorate, ~ 70% of the population, followed by Mount Lebanon with ~ 25%. The gender distribution was balanced, with males (51%) slightly outnumbering females (47%); 2% of participants did not complete the age variable in the questionnaire. The majority of the population were within the age range of 18 to 29 years, indicating a youthful demographic, while those aged above 90 were the smallest group. The mean age was 38 years (Fig. 3).
Table 1
Demographic and Clinical Characteristics of PCR-Tested Individuals Across Governorates in Lebanon
| Total % (N) | Positive for SARS-CoV-2 % (N) |
Governate |
Akkar | 3.8% (411) | 9% (37) |
Baalbek-Hermel | 0.0% (2) | 0 |
Beirut | 0.9% (103) | 5.8% (6) |
Beqaa | 0.1% (15) | 6.7% (1) |
Mount Lebanon | 25.4% (2781) | 11.5% (319) |
El Nabatieh | 0.0% (1) | 0 |
North | 69.7% (7635) | 11.9% (907) |
South | 0.1% (6) | 0 |
Missing | 0.0% (0) | 0 |
Gender |
Female | 46.7% (5116) | 11.8% (605) |
Male | 51.5% (5640) | 11.3% (639) |
Missing | 1.8% (198) | 13.1% (26) |
Blood group |
A+ | 29.9% (3278) | 14.2% (465) |
B+ | 7.2% (784) | 12.2% (96) |
AB+ | 2.7% (299) | 13.4% (40) |
O+ | 26.5% (2908) | 12.4% (360) |
A- | 2.4% (263) | 15.2% (40) |
B- | 0.7% (72) | 15.3% (11) |
AB- | 0.4% (44) | 22.7% (10) |
O- | 3% (325) | 11.4% (37) |
Missing | 27.2% (2981) | 7% (211) |
Social distancing |
Yes | 39.9% (4370) | 9.8% (427) |
No | 59.6% (6525) | 12.9% (840) |
Missing | 0.5% (59) | 5.1% (3) |
Public Transportation |
Yes | 14% (1530) | 8.1% (124) |
No | 86% (9420) | 12.2% (1146) |
Missing | 0.0% (4) | 0.0% (0) |
Crowded area |
Yes | 32.9% (3599) | 8.8% (317) |
No | 67.1% (7351) | 13% (953) |
Missing | 0.0% (4) | 0.0% (0) |
Recent Travel |
Yes | 10.4% (1141) | 2.2% (25) |
No | 89.5% (9809) | 12.7% (1245) |
Missing | 0.9% (4) | 0.0% (0) |
Contact with Travelers |
Yes | 6% (653) | 4.3% (28) |
No | 94% (10297) | 12.1% (1242) |
Missing | 0.0% (4) | 0.0% (0) |
Contact with COVID positive patients |
Yes | 37.7% (4133) | 14.4% (595) |
No | 62.3% (6819) | 9.9% (675) |
Missing | 0.0% (2) | 0.0% (0) |
Hospital visit |
Yes | 8.6% (941) | 8.9% (84) |
No | 91.4% (10009) | 11.8% (1186) |
Missing | 0.0% (4) | 0.0% (0) |
Associations between demographic and behavioral factors with SARS-CoV-2 positivity rates
Our data illustrates a comprehensive overview of SARS-CoV-2 positivity rates in relation to demographic and behavioral factors (Table 1). The positivity rates were found to be similar between females (11.8%) and males (11.3%; Fisher’s exact test P = 0.43). We found that positivity rates significantly varied by blood group (Fisher’s exact test P = 0.0005). Individuals with AB- blood type exhibited the highest positivity rate at 22.7%, while the lowest rate was observed in those with the O- blood type at 11.7%. To assess the relationship between blood groups and COVID-19 positivity, we performed Fisher’s exact tests to analyze the association between each specific blood group and testing positive for SARS-CoV-2. The analysis considered the distribution of positive and negative cases across different blood groups, comparing each individual group to the combined data of all other groups. We observed statistically significant association between the blood group and COVID-19 PCR test positivity; individuals with blood type O + were significantly less likely to be positive for COVID-19 (odd’s ratio: 0.716; BH-corrected P = 0.012).
We also tested for associations between COVID-19 positivity and seven self-reported behavioral factors, all of which were significant (Fisher’s Exact Tests BH-corrected P < 0.006). Two key behaviors were associated in the expected directions. Notably, adherence to social distancing was negatively related to positivity, with a 9.9% positivity rate for those who reported social distancing versus 13.0% for those who did not. In addition, contact with COVID-19 positive patients was associated with a markedly higher positivity rate of 14.6%, in contrast to 10.0% for those without such contact.
In contrast, additional behaviors, which would be expected to be associated with higher positivity rates (were all possible confounders excluded) were significantly associated with lower positivity rates. In particular, the use of public transportation was associated with an 8.2% positivity rate, whereas those not using public transportation had a higher rate of 12.3%. Being in crowded areas was associated with a lower positivity rate of 8.9% compared to a 13.1% positivity rate for those who avoided crowded areas. Similarly, recent travel was negatively related to positivity rates, with recent travelers showing a rate of 2.2% against 12.9% for those who had not traveled. In addition, the rate of positivity was higher in people who did not have contact with travelers (12.2 versus 4.3 Hospital visits were associated with a positivity rate of 9.0%, whereas those who had not recently visited a hospital showed a higher rate of 12.0%.
Age-Specific Analysis of COVID-19 Positivity Rates
In the analysis of COVID-19 positivity rates across different age groups, we observed notable variation in the positivity percentage (Fig. 4). The study encompassed a diverse population, segmented into age groups ranging from 5 to over 90 years. A total of 1,270 PCR positive cases were identified from the tested samples, with the age-specific positivity rates shedding light on the infection distribution among the age cohorts. The analysis of COVID-19 positivity rates across various age groups reveals distinct patterns. Notably, positivity rates are higher among early teenagers and the oldest adults. For instance, the 11 to 17 age group exhibited a higher infection rate compared to younger children. Meanwhile, other age groups displayed varying rates, with younger adults and middle-aged cohorts showing relatively lower percentages. Detailed figures for each age group are provided in Fig. 4.
COVID-19 Positivity Rates by Caza
We analyzed COVID-19 positivity rates across cazas, which are geographical subregions in Lebanon, and found significant variation, as shown in Fig. 5. This variability likely reflects differences in viral spread and testing efficacy across regions. Notably, El Koura and El Meten reported the highest positivity rates among cazas with sufficient testing, at 12.6% and 12.4% respectively, suggesting they are focal points of infection. Tripoli, meanwhile, exhibited the highest positivity rate at 18.5%. Akkar and El Batroun also showed considerable positivity rates of 9.1% and 10.9%, indicating moderate transmission levels.
Trends in COVID-19 Positivity Rates: An Analysis from April 2020 to March 2021
Between April 2020 and March 2021, COVID-19 positivity rates varied widely, as shown in Fig. 6. Starting from negligible levels, the rates increased gradually through spring and early summer. In the fall, a significant rise occurred, culminating in a peak in October. Although there was a slight reduction in November, January 2021 experienced the most pronounced spike, which coincided with the emergence of the Delta variant, leading to the highest positivity rate in the period. This was subsequently followed by a notable decrease, with rates falling significantly by March 2021.
Symptom Distribution Among Individuals Screened for COVID-19
Individuals presenting at the testing center exhibited a range of symptoms, leading to assessments for COVID-19. Notably, symptom presence was not a requirement for participation, allowing for the inclusion of asymptomatic individuals in the study (Fig. 7). Among those with symptoms, fever was reported in 1,273 individuals, showing a positivity rate of 23%. Headaches and sore throats, reported in 3,058 and 1,844 cases respectively, each had a positivity rate of 17%. Additionally, the loss of smell and taste, reported by 1,065 and 914 individuals respectively, correlated with a higher positivity rate of 25%.
Our data analysis demonstrates that the loss of smell and taste were more strongly correlated with positive COVID-19 test results than other common symptoms such as fever, myalgia, and cold. This correlation is illustrated by the heatmap in Fig. 6, which highlights the distribution and frequency of reported symptoms among the tested cohort, revealing distinct symptom clusters associated with positive cases.
Further analysis explored the relationship between the number of symptoms reported and COVID-19 positivity (Fig. 8). This revealed a gradual increase in positivity rates with the number of symptoms reported, particularly starting from three symptoms.
To validate the survey results against known symptom signals from this period, we investigated symptom enrichment in positive cases (Fig. 9). Significant symptom sets including fever and myalgia alongside smell and taste loss were markedly enriched in positive cases. Specific symptom combinations were evaluated for their association with positive test results, revealing significant odds ratios for combinations involving respiratory and systemic symptoms like cold, dry cough, and myalgia (Fig. 10).
Additionally, the UpSet plots (Fig. 11) unveil the frequency and intersectionality of symptoms within positive and negative cohorts, with a notable predominance of smell and taste loss among positive cases. The nuances of these symptom distributions were further explored through a Principal Coordinates Analysis (PCoA) (Fig. 12), which graphically disperses individual cases based on the similarity of their symptom profiles. Although there is a significant difference (PERMANOVA: R2 = 0.0046, P = 0.001), the effect size is negligible, suggesting that while symptomatic differences exist, they account for a minor portion of the variation in reported symptoms.