Thread by @ZdenekVrozina

A useful Germany-based study looked at what happened after seasonal COVID vaccination moved into routine care. It used real-world insurance data to compare vaccinated and unvaccinated adults during the 2023/24 autumn-winter period🧵

https://twitter.com/cidrap/status/2046294864917098853

The main pattern was consistent. Vaccinated people had fewer severe outcomes. COVID-related hospitalization was lower. Respiratory and cardiovascular hospitalizations were also lower. New long COVID diagnoses were lower too.

The follow-up period ran for up to 121 days, essentially covering 1 December 2023 to 31 March 2024. So the paper is really about short-term seasonal impact, not long-term durability over many more months.

This was not a randomized trial. It was a retrospective observational cohort study using claims data from AOK PLUS in Saxony and Thuringia.

The dataset covered outpatient care, hospital care, diagnoses, vaccination records, sick leave, and demographic information from September 2022 to March 2024. So this was a large real-world analysis, not a small clinic-based sample.

Adults 18+ were included if they had insurance coverage during the inclusion period, and during the prior 12 months. That baseline year was used to assess health status before comparison.

Vaccinated meant having a recorded SARS-CoV-2 vaccination between 1 September and 30 November 2023. Unvaccinated meant no such record in the same period.

So important methodological nuance. In this study, unvaccinated did not necessarily mean never vaccinated. Some people in the unvaccinated group may have been vaccinated earlier, which could blur the contrast between groups.

That washout period is important. It was meant to avoid mixing in people vaccinated earlier with older formulations and to focus the analysis on autumn vaccination after routine coding had been established in Germany.

A major issue in the raw data. Vaccinated people were much older and sicker than unvaccinated people. Average age was about 72 in the vaccinated group versus about 44 in the unvaccinated group. A direct comparison would have been badly misleading.

To address that, they used propensity score matching based mainly on age, sex, comorbidity burden, and prior influenza vaccination. After matching, about 73,066 people remained in each group.

The follow-up period ran for up to 121 days, essentially covering 1 December 2023 to 31 March 2024.

The study looked at several outcomes. Recorded SARS-CoV-2 infection, new long COVID diagnoses, respiratory infections, COVID related hospitalization, respiratory hospitalization, cardiovascular hospitalization, all-cause mortality, healthcare use, costs, and sick leave.

Vaccinated people had fewer severe outcomes. COVID related hospitalization was lower. Respiratory and cardiovascular hospitalizations were also lower. New long COVID diagnoses were lower too.

Some of the key estimates were - COVID hospitalization RR about 0.41, long COVID RR about 0.43, respiratory infections RR about 0.91, and all-cause mortality HR about 0.76. In other words, most outcomes favored the vaccinated group.

Recorded SARS2 infections were lower in vaccinated people too, but in the main analysis that difference did not reach conventional statistical significance. So the strongest signal was more about severe outcomes than about infection itself.

The economic results were also notable. Vaccinated people had fewer hospital days and lower inpatient costs.
There were also fewer sick-leave days related to COVID-19 and respiratory infections.

This is still an observational study. Matching helps, but it only adjusts for measured variables. It cannot fully remove residual confounding.

The authors themselves mention healthy vaccinee bias. That point becomes especially important for broad outcomes like all-cause mortality or cardiovascular hospitalization.

Four months is useful for seasonal impact, but it is not enough to answer every question about longer-term protection or delayed outcomes.

So what does this study actually support?
It supports the view that, in this real-world German analysis, seasonal COVID vaccination was associated with fewer severe outcomes, lower healthcare use, and lower economic burden.

This is a meaningful and fairly well-executed observational study.

Schmetz al al., Clinical and economic benefits of seasonal COVID-19 vaccination in Germany: results from the ROUTINE-COV19 Study, September 2022 to March 2024. eurosurveillance.org/content/10.280…