The Omicron cases were detected among the samples with no amplification of the spike (S) gene (non-S gene target failure) by real-time reverse transcription PCR using the TaqMan SARS-CoV-2 mutation panel (Thermo Fisher Scientific, https://www.thermofisher.com) for single-nucleotide polymorphism genotyping focused on the K417N and L452R mutations. Samples positive for the K417N mutation and negative for L452R were considered compatible with Omicron. The analysis method was validated through whole-genome sequencing of 63 samples. Libraries were constructed by using Ion AmpliSeq SARS-CoV-2 Insight Research Assay and were sequenced with Ion GeneStudio S5 system (both Thermo Fisher Scientific). Next-generation sequencing data were analyzed using Torrent suite software and were assembled by IRMA (2). Lineage assignment was done by Pangolin (3) by using consensus fasta.

We obtained data on sociodemographic characteristics (age), vaccination status (nonvaccinated or fully vaccinated), and presence or absence of symptoms, as well as symptom onset date (SOD) or diagnosis date (DD) for asymptomatic cases, from the Contact Tracing Program of Cantabria (Appendix). We obtained the same information for the 1,708 coronavirus disease cases of November 2021, when the Delta variant of SARS-CoV-2 represented 100% (1,299/1,299) of samples. We identified 12,587 close contacts and 2,201 secondary cases.

In Spain, close contacts were tested as early as 3 days and as late as 9 days after the date of last contact, depending on when the patient came into the system (4). We defined SAR as the proportion of secondary cases among close contacts (those who had been at a distance of <2 m for >15 min) identified through contact tracing (contact 2 days before to 10 days after index case SOD or diagnosis). We classified each relationship by the setting where it took place (household, social, or occupational). We defined global SAR as the average of secondary cases among all relationships (5). All SARs (with 95% CIs) are presented by index case-patient’s vaccination status. We tested difference in SAR between Delta and Omicron and differences between SARs for vaccinated and unvaccinated persons by variant and contact setting by Pearson χ² test(Table 1).

Global SAR was 39% (95% CI 36.5%–42.2%) for Omicron cases and 26% (95% CI 25.3%–27.4%) for Delta, a 13-point absolute increase (9.9–16.1; p<0.0001) (Table 1). A higher SAR was also registered in social settings (30.5% for Omicron vs. 16.2% for Delta) and occupational (31% vs. 10.5%) settings but not between household close contacts (49.4% vs. 48%).

Among Delta variant cases, unvaccinated persons showed an overall increased SAR of 7.8% (95% CI 5.6%–10%; p<0.001), household SAR of 9.9% (95% CI 5.8%–14%; p<0.001), and occupational SAR of 9.5% (95% CI 0.8%–18.1%; p = 0.01) compared with vaccinated persons. In contrast, for the Omicron variant, we found no differences between vaccinated and unvaccinated persons in any of these categories.

We selected only symptomatic index cases to calculate transmission, incubation, and serial interval periods. We defined transmission period as the distribution of days from index case SOD to date of last contact with close contacts who became secondary cases. For incubation period and serial interval, we required that the secondary case-patient also be symptomatic. We defined the incubation period as the number of days between date of last contact and secondary case SOD and serial interval as the number of days between the index case SOD and the secondary case SOD (6).

Figure 1

Figure 1. Distribution of Omicron variant SARS-CoV-2 cases, Cantabria, Spain, December 2021. A) Transmission period; B) incubation period; C) serial interval. Each panel shows case density over time (top), a typical boxplot...

For the 3 periods, we report mean (SD) and median (interquartile range [IQR]). We calculated Omicron-Delta mean differences and Student t test, 95% CI, and p values (Table 2). We constructed histograms, density plots, boxplots, and cumulative distribution functions for Omicron (Figure 1) and Delta (Appendix Figure).

Figure 2

Figure 2. Transmission period distribution for Omicron and Delta variant SARS-CoV-2 cases by vaccination and symptom status, Cantabria, Spain. SOD, symptom onset date.

The transmission period of Omicron cases was shorter (mean 0.5, median 0 days) than Delta cases (mean 0.8, median 1 day) (Figure 1, panel A) and grouped around day 0 after SOD. Mean differences between both variants were significant (−0.3 days; SD −0.56 to −0.02), and IQRs remained equal (Figure 2).

Incubation period had a median of 3 days for both variants and IQR was shorter for Omicron (Figure 1, panel B). We found no mean differences in incubation period. Finally, mean serial interval was significantly shorter for Omicron (4.8 vs. 5.4 days, SD −0.6 to −0.15; p = 0.008) (Figure 1, panel C) with a median of 4 versus 5 days. We found no differences within variants between vaccine status for any of the periods.

Omicron has spread quickly worldwide since its first notification on November 11, 2021 (7). Our findings demonstrate a significant increase in SAR for Omicron cases in Cantabria, Spain, compared with Delta in a similar period and with high vaccine coverage (>80% of target population). Global SAR and social SAR increased by ≈50% (26.3% to 39.3% for global and 16.2% to 30.5% for social), but we did not find significant differences in household SAR. By the end of December, cases increased exponentially, and the Christmas holiday could have affected the number of contacts per case in the occupational and social settings

In this study, vaccinated Omicron index case-patients seemed to have the same transmission capacity as nonvaccinated persons. We did not find this increased transmission capacity for the Delta variant, where significant differences in SAR were observed in global, household, and occupational settings (Table 1) within groups.

Omicron’s increased transmissibility is consistent with the registered tendency of transmission when persons are asymptomatic or early in the symptomatic phase. SARS-CoV-2 transmission took place from day −1 to day +3 of SOD, when most secondary case contacts happen. Median day of transmission was reduced from +1 after symptom onset in Delta to day 0 (SOD) in Omicron (Table 1). Even though the incubation period did not statistically differ, serial interval was significantly decreased in Omicron (mean 4.8 vs. 5.3, median 4 vs. 5) and was again more grouped to the left (IQR 3 vs. 5). Of secondary cases, 90% had an incubation period of 6 days for Omicron and 7 days for Delta.

It has been hypothesized that Omicron’s increased SAR is derived from a concentration of contagion events in the presymptomatic or paucisymptomatic period, when infected persons might be unaware of their status and containment measures such as contact-tracing, isolation, and quick testing are not possible. Half of Omicron contagion events happened before symptom onset. This finding could imply that the effectiveness of nonpharmaceutical measures targeting symptomatic cases (such as contact tracing, quick testing, and isolation) would be substantially decreased in the absence of preventive measures such as social distancing and limiting large gatherings or social meetings.

The social and economic effects of isolation and quarantine have led to continued debate regarding appropriate and adequate quarantine periods, especially in light of possible changes in disease dynamics caused by the Omicron variant (8–13). In this study, transmission for Omicron and Delta >5 days after SOD was rare, accounting for 8/356 (2%) of secondary cases in Omicron and 79/1,642 (5%) in Delta (Figure 2). This finding could potentially contribute to the debate about quarantine and isolation periods and lessening the social and economic costs of COVID-19 control measures.

This article was preprinted at https://www.researchsquare.com/article/rs-1279005/v1.