Glioblastoma is the least survivable form of brain cancer, with a 5-year relative survival rate of 6.9% in the US as of data published in 2023 [1]. This is up from 4.7% on data published in 2012 [12].

The Central Brain Tumor Registry of the United States (CBTRUS) publishes annual statistical reports detailing (among other things) relative survival rates for diagnoses made in a period of years prior to publication.

Since 2012, CBTRUS's published figures for relative survival rates of glioblastoma were (with 95% confidence intervals in parentheses):

• 2023 6.9% (6.8-7.1) ([1], table 10)

• 2022 6.9% (6.7-7.1) ([2], table 11)

• 2021 6.8% (6.7–7.0) ([3], table 24)

• 2020 7.2% (7.0-7.4) ([4], table 23)

• 2019 6.8% (6.7–7.0) ([5], table 21)

• 2018 5.6% (5.3–5.8) ([6], table 20)

• 2017 5.5% (5.2–5.8) ([7], table 21)

• 2016 5.5% (5.2–5.8) ([8], table 23)

• 2015 5.1% (4.8–5.4) ([9], table 23)

• 2014 5.0% (4.8–5.4) ([10], table 22)

• 2013 4.7% (4.4–5.0) ([11], table 19)

• 2012 4.7% (4.4–5.0) ([12], table 21)

Here in graphical form:

This market will resolve YES if CBTRUS's 2035 report shows a relative survival rate across all ages and diagnoses in the given time period, for glioblastoma in the United States, of greater than 15.0%. This considers only the point estimate and not the confidence interval. It will resolve NO if the 2035 report shows a relative survival rate of 15.0% or less.

If CBTRUS's reports cease to be published, this market may use a different source, or resolve N/A at my judgement. Any alternative source much have a closely-matching methodology and dataset. If methodological, data source, or definitional differences change non-trivially (in my judgement) such that the comparisons to 2023 CBTRUS data is not meaningful, the market will resolve N/A.

If CBTRUS doesn't publish a report in 2035, but it seems reasonably likely they will publish again in the future, then resolution will be delayed pending a future publication, on my judgement. If their next report's survival statistics are based on diagnoses from the range of years expected to have been included in the 2035 report, then data from that publication will be used for resolution. If data from a different date range is used, then resolution will be based on a linear interpolation between the data from the last report published before 2035, and the first published after 2035. The relevant dates for interpolation will be the midpoint of the range of dates of diagnoses included in the report, not the publication date.