Two dozes of the Pfizer/BioNTech COVID-19 vaccine provided just 33% protection against infection during a surge in infections in South Africa that is dominated by the new Omicron variant, according to preliminary analyses announced today (December 14) at a press briefing by Discovery Health (South Africa’s largest private health insurer) and the South African Medical Research Council. This compares with 80% protection before Omicron’s emergence.

“The omicron-driven fourth wave has a significantly steeper trajectory of new infections relative to prior waves. National data show an exponential increase in both new infections and test positivity rates during the first three weeks of this wave, indicating a highly transmissible variant with rapid community spread of infection,” said Discovery Health chief executive Ryan Noach, according to the Associated Press.

The study—based on more than 211,000 positive COVID-19 tests, about 78,000 of which were attributed to infections with Omicron—did not factor in whether patients had had a prior SARS-CoV-2 infection, which would give them a boost in immunity, notes NPR’s Goats and Soda. Thus, for people who are fully vaccinated but have never been infected with the coronavirus, the vaccination may lower their risk of infection by less than 30%.

The results jibe with another study on vaccine effectiveness against Omicron, from a pair of researchers at the University of California, Santa Cruz (UCSC). Reporting in a preprint posted on medRxiv on Sunday, Billy Gardner and Marm Kilpatrick used computational models to predict that two doses of Pfizer/BioNTech’s vaccine or Moderna’s similar mRNA vaccine provide only about 30% protection against symptomatic infection—and that’s soon after vaccination, before protection starts waning. After more than four months, protection is “essentially eliminated,” Kilpatrick tells Reuters. The analysis suggested that booster shots could restore protection to about 48%, “which is similar to the protection of individuals with waned immunity against the Delta variant (43%),” Kilpatrick says.

On a more positive note, the newly released data from South Africa and the preprint from the UCSC team, neither of which have undergone peer review, indicate that vaccines still offer fairly robust protection against severe disease, though it did take a hit as well. In the announced results from South Africa, researchers found that protection against severe disease offered by the Pfizer/BioNTech vaccine dropped from 93% during the country’s Delta surge to about 705 in recent weeks, according to Goats and Soda. Gardner and Kilpatrick found that mRNA vaccination initially provides about 86% protection against severe disease caused by Omicron, with that number waning over time but able to be restored to 91% with a booster.

Early data indicate that the Omicron variant may cause less severe disease than variants that came before it, regardless of vaccination status. While COVID-19 case numbers have skyrocketed in South Africa over the past couple of weeks—with Omicron now accounting for 90%, according to Noach—deaths have not kept pace. COVID-19–related hospitalizations are 29% lower now than they were during a surge in South Africa last year, even when accounting for differences in vaccine coverage, AP reports of the preliminary analyses. And in one study of Johnson & Johnson’s COVID-19 vaccine in South Africa, investigators haven’t recorded a single death from Omicron, South African Medical Research Council President Glenda Gray said in a news conference reported by Reuters.

“Although we have had a lot of breakthrough infections there has been very little hospital admission in comparison to the Delta period. And as of today we have had no one who has died from Omicron from the J&J study, so that’s the good news, it shows again that the vaccine is effective against severe disease and death.”

Data coming out of the UK also support the idea that Omicron is more easily transmissible, AP reports, with research there suggesting that COVID-19 vaccines from Pfizer/BioNTech and AstraZeneca are less effective at preventing symptomatic infections caused by Omicron than by previous variants.