The letter refers to the 1976 vaccine and not the current ones. This is very old news but certainly something to consider with the new vaccines. The 25 deaths were in 52 million people immunized (24% of the US herd). There were many additional cases of less serious GBS.
Comparing 1976 apples to apples: If 1/3 of the 52MM people had caught H1N1 and there was a 0.001 case fatality rate (the mildest of the projected CFRs for the current strain of swine flu and the one I currently think is the most probable) then 17,160 of those 52MM would have died from the flu. You have to consider that no vaccine confers 100% immunity (far from it actually) so not all of those 17,160 would have been save by the vaccine.
If the current 0.0035 (0.35%) fatality rate currently occurring in Argentina is applied then 60,600 of the 52MM people vaccinated would have otherwise died from the flu. If 2.4% died (the 1918 CFR) then 429,000 of the vaccinated 52MM would have died from flu.
Whether it's a good risk depends on the tox profile of the vaccine (currently unknown) and the eventual CFR (obviously also unknown). A crap shoot in other words.
From the current WHO release on H1N1 vaccines:
Special safety concerns
Influenza vaccines have been used for more than 60 years and have an established record of safety in all age groups. While some serious adverse events have been reported, these have been rare.
Nonetheless, special safety issues will inevitably arise during a pandemic when vaccine is administered on a massive scale. For example, adverse events too rare to show up even in a large clinical trial may become apparent when very large numbers of people receive a pandemic vaccine.
Some adverse events will be coincidental – that is, associated in time with vaccine administration, yet not directly caused by the vaccine. Genuine adverse events directly caused by the vaccine may also occur, but cannot be predicted in advance. Given the safety record of seasonal vaccines, such events are expected to be rare.
Time constraints mean that clinical data at the time when pandemic vaccines are first administered will inevitably be limited. Further testing of safety and effectiveness will need to take place after administration of the vaccine has begun.
For these reasons, WHO advises all countries administering pandemic vaccines to conduct intensive monitoring for safety and efficacy, and many countries have plans in place for doing so. On the positive side, mass vaccination campaigns can generate significant safety data within a few weeks.
International sharing of data from such post-marketing surveillance will be vital in guiding risk-benefit assessments and determining whether changes in vaccination policies are needed. WHO has developed standardized protocols for data collection and reporting in real-time, and will communicate findings to the international community via its web site.