Friedberg argued that, given adequate beds/ICU/ventilators and use of emerging treatments like remdesivir, the COVID infection fatality rate (IFR) in well‑resourced settings (e.g., the U.S.) could be pushed down to ~0.2%, similar to “Rest of China” outside Wuhan, rather than ~4% as seen in Wuhan.

1. His China benchmark (~0.2%) was already too low. Early analyses of China’s first wave show that outside Hubei, the crude case fatality rate (CFR) was about 0.4%–0.9%, not 0.2%. An ecological study found fatality rates of 0.49–0.88% in the “Rest of China,” while detailed CDC modeling estimated a crude CFR of 0.35–0.43% for China excluding Hubei. (pubmed.ncbi.nlm.nih.gov) IFR (which is lower than CFR because it includes undiagnosed infections) has been modeled for China overall at roughly 0.6%, again well above 0.2%. (arxiv.org) So his reference point itself understated the true fatality risk.

2. Empirical IFR in high‑income countries was many times higher than 0.2% in the pre‑vaccine era, even with modern hospital care. An Imperial College analysis in October 2020 estimated an overall IFR of 1.15% (95% PI 0.78–1.79) in high‑income countries, versus 0.23% in low‑income countries with much younger populations. (imperial.ac.uk) A global meta‑analysis of early IFR studies (through mid‑2020) found a pooled IFR of 0.68% (0.53–0.82%) across populations. (pubmed.ncbi.nlm.nih.gov) US‑specific work using county‑level and seroprevalence data estimated IFRs around 0.8–1.0% for the first wave (e.g., ~0.86% for New York City, with similar national estimates at or below 1%). (pmc.ncbi.nlm.nih.gov) These are 3–5× higher than his 0.2% target.

3. Even over the full pandemic, with vaccines and improved care, population‑level IFR in the U.S. stayed well above 0.2%. As of 2025, CDC‑linked summaries and excess‑mortality analyses estimate roughly 1.2–1.23 million COVID‑19 deaths in the United States and more than 313 million infections by November 2022, with modeling suggesting that by late 2022 about 94%+ of Americans had been infected at least once and 97% had some immunological exposure. (en.wikipedia.org) Taking ~1.2 million deaths over ~313–>330 million infections implies a cumulative IFR on the order of 0.3–0.4%, roughly double Friedberg’s 0.2% claim, even after accounting for later waves with lower severity and widespread immunity.

4. The treatment‑based mechanism he relied on (beds/ventilators + remdesivir) did not deliver such a large IFR reduction. While the ACTT‑1 trial showed remdesivir shortened time to recovery with at best a modest, statistically uncertain mortality effect, (acc.org) the much larger WHO Solidarity trial found little or no impact of remdesivir on 28‑day mortality. (who.int) Major mortality reductions instead came later from corticosteroids, anticoagulation, non‑invasive oxygen strategies, and especially vaccination and prior infection, none of which were the main levers Friedberg emphasized. Even with these advances, age‑structured IFRs in older adults remained far above 0.2%.

5. His conditional (“assuming enough beds/ICU/ventilators”) caveat does not rescue the forecast. Many high‑income settings had periods where hospitals were strained but not completely overwhelmed, yet their inferred IFRs from serology and deaths still clustered around 0.5–1%, consistent with the Imperial and meta‑analytic estimates above and inconsistent with a generalized 0.2% IFR for all infections in a population with a rich‑country age structure. (imperial.ac.uk)

Given that:

• (a) the “Rest of China” fatality rates were significantly higher than 0.2%,
• (b) high‑income and U.S. IFR estimates in 2020 were around 0.7–1.2%, and
• (c) even with vaccines and improved care, the cumulative US IFR has stayed around 0.3–0.4%, not 0.2%,

Friedberg’s prediction that well‑resourced settings could generally drive COVID’s infection fatality rate down to ~0.2% using hospital capacity and treatments like remdesivir did not come true.