Covid paper reviews Contents

I reviewed Covid papers for a few months, after I joined an academic endeavor that I successfully leveraged as a means of reliably getting myself to read approximately 1 Covid preprint a week. I’m storing my Covid paper/preprint reviews on this page in reverse chronological order.

DISCLAIMER: I really do not know anything about these topics. I am not formally trained in literature review, Covid research, or anything else, and the extent of my informal training is reading a bunch of Slate Star Codex literature reviews. This is not medical advice; please do not treat it as authoritative. I am kind of putting this out there so that, as per Cunningham’s law, people will give me feedback on my evaluations, so please give me feedback.

01 December: “Severe COVID-19 induces molecular signatures of aging in the human brain”

Main claims:

• Comparison of postmortem tissue samples from 12 severe Covid patients and 12 matched controls reveals that Covid patients’ brains show the same upregulation and downregulation of genes associated with aging: “For all frontal cortex aging datasets, genes upregulated in aging are upregulated in severe COVID-19; likewise, genes downregulated in aging are also downregulated in severe COVID-19”.
• Immune activity was also upregulated, as it is in aging.
• This is especially important for young people: “Altogether, our analyses suggest that the brain aging effects of COVID-19 are far more pronounced in younger patients than in older patients.”

Strength of evidence (highlight): Strong / Reliable / Potentially informative / Not informative / Misleading

Why is it important to review this preprint?

• Cognition is important; Covid’s effects on it should influence policy (e.g., fewer restrictions might be favored because lockdowns tend to reduce quality of life, but if more people then get Covid and develop cognitive issues, those policies might not actually be advisable, because the costs of widespread cognitive issues might outweigh the benefits of a lack of lockdowns).
• Implications for other pathogens: “Although our study does not examine the specificity of COVID-19-induced transcriptomic changes in the brain, the implications of our findings may readily extend to related pathologies. For instance, prior clinical trials have shown that cognitive impairment is observed in 55% of survivors of [...] (SARS) 12 months after discharge [31]. Such behavioral observations suggest that similar molecular effects in the brain may be observed not only in severe COVID-19 but also in other conditions characterized by increased peripheral and central inflammation, severe hypoxic insults, and microvascular brain pathologies”.

Recommendations for questions/issues for reviewers to address?

• To what extent do molecular signatures of aging correspond to actual perceptible cognitive decline? (Cf. aducanumab beta-amyloid plaques)
• Use larger sample sizes / investigate more experimental conditions; do other illnesses, such as severe flu, produce the same issues, or is Covid unique here?
• Possible selection or other bias because it’s a matched control study (e.g., young people unusually susceptible to severe Covid also have other issues, compared to matched controls, who might have died from other things, like car accidents)
• The controls were collected before Covid; to what extent could Covid policies (e.g., lockdowns causing social isolation / psychiatric illness) be affecting brain health?
• We don’t know what Covid cognitive decline looks like in the long run and if people recover over time (both because Covid is so new and because the people in the study died).

What kind of expert should review this preprint?

• Someone with expertise in these areas, such as the genetics of aging

17 November 2021: “Global prevalence of post-acute sequelae of COVID-19 (PASC) or long COVID: A meta-analysis and systematic review”

Main claims:

• Meta-analysis of 40 long Covid studies finds high prevalence of long Covid: “The empirical findings suggest a global PASC prevalence of approximately 43%. Based on a WHO estimate of 237 million worldwide COVID-19 infections, this global pooled PASC estimate indicates that around 100 million individuals currently experience or have previously experienced long-term health-related consequences of COVID-19.”
• Hospitalization and being female correlate with increased risk: “Individuals who were hospitalized during acute COVID-19 infection had higher PASC prevalence at 57%. Female adults had both higher prevalence and risk of having PASC than male adults (49% vs 37%).”
• “The five most prevalent symptoms were the following, with corresponding estimated pooled symptom-specific prevalence: fatigue at 0.23 (95% CI: 0.13, 0.38), dyspnea at 0.13 (95% CI: 0.09, 0.19), insomnia at 0.13 (95% CI: 0.06, 0.28), joint pain at 0.13 (95% CI: 0.05, 0.29), and memory problems at 0.13 (95% CI: 0.10, 0.18).”

Strength of evidence (highlight): Strong / Reliable / Potentially informative / Not informative / Misleading

• Why is it important to review this preprint?
• Long Covid is an important topic; for many vaccinated young people, it’s a bigger concern than hospitalization or death, so it should be factored into public health decisions.
• It’s #1 on Covid-19 Primer for papers released today (335 social media mentions), and #6 for past 7 days.

Recommendations for what questions/issues reviewers can address?

• These estimates are much higher than previous estimates of long Covid prevalence
• The paper had less data from the regions of Africa and Australia, about children, and addressing race/ethnicity
• The paper doesn’t discuss whether studies addressed preexisting/psychosomatic/etc. symptoms of long Covid (e.g., how many people had fatigue or insomnia anyway, or despite not having Covid, or falsely attributed their issues to Covid)

What kind of expert should review this preprint?

• Someone who, unlike me, knows about statistics
• Someone who can evaluate the studies used in this meta-analysis

03 November 2021: “Comparison of the immunogenicity of BNT162b2 and CoronaVac COVID-19 Vaccines in Hong Kong”

Main claims:

• n>700 study of recipients of Pfizer and CoronaVac finds that Pfizer has more immunogenicity
• “One month after the second dose of vaccine, BNT162b2 elicited significantly higher PRNT50, PRNT90, sVNT, spike receptor binding, spike N terminal domain binding, spike S2 domain binding, spike FcR binding and antibody avidity levels than CoronaVac. [...] CoronaVac induce[s] higher CD4+ and CD8+ T cell responses to the structural protein than BNT162b2.”
• “Allowing for an expected seven-fold waning of antibody titres over six months for those receiving CoronaVac, only 16.3% would meet the 50% protection threshold versus 79.6% of BNT162b2 vaccinees.”

Strength of evidence (highlight): Strong / Reliable / Potentially informative / Not informative / Misleading

Why is it important to review this preprint?

• Knowledge about vaccine efficacy will help us set policy around which vaccines to administer and potentially consider effective for vaccine passport purposes
• It’ll also help people decide which vaccine they’d like to get out of the options available to them
• Authors suggest booster doses might be needed for “older CoronaVac recipients”

Recommendations for what questions/issues reviewers can address?

• Address specific measures of immunogenicity
• Check if there are particular aspects of Hong Kong vaccine rollout that might influence uptake and cause differences between different groups (e.g., maybe more vulnerable people tend to get a specific vaccine because of government messaging) (“The choice of vaccine was not randomized and there might be a selection bias in those opting for each vaccine.”)
• “Our study only focused on investigating the immunogenicity at 1 month after two doses of vaccination.”
• Data on Pfizer antibody waning not as incorporated into paper, so comparison can be improved

What kind of expert should review this preprint?

• Someone who can evaluate the biological and medical underpinnings of the immunogenicity assessment

27 October 2021: “REcovery and SURvival of patients with moderate to severe acute REspiratory distress syndrome (ARDS) due to COVID-19: a multicentre, single-arm, Phase IV Itolizumab Trial: RESURRECT”

Main claims:

• Itolizumab is a monoclonal antibody which has received an EUA for use in India
• This single-arm clinical trial (n=300) across multiple hospitals in India finds that itolizumab seems to reduce mortality and shorten recovery time in patients who are hospitalized and require oxygen therapy
• Study also says that itolizumab is generally unproblematic in terms of safety

Strength of evidence (highlight): Strong / Reliable / Potentially informative / Not informative / Misleading

Why is it important to review this preprint?

• It’s important to know which Covid treatments are effective, since Covid hospitalizations continue

Recommendations for what questions/issues reviewers can address?

• Potentially small sample size with no control group (authors compare their results to a meta-analysis of patients under similar conditions and find a much lower mortality rate than in the control groups of those studies (n<12000, 27%)
• Possible conflict of interest since the study was funded/conducted by the company that developed the treatment

What kind of expert should review this preprint?

• People with expertise in medical treatment of Covid-19, monoclonal antibodies, etc.

13 October 2021: “Inactivated poliovirus vaccine induces antibodies that inhibit RNA synthesis of SARS-CoV-2: An open-label, pre-post vaccine clinical trial”

Main claims:

• Researchers inoculated adult volunteers with one dose of the IPV vaccine and measured their anti-RdRp antibodies before and after inoculation
• Many RNA viruses, including the polio virus and SARS-CoV-2, have the RdRp protein, which enables replication of RNA; if RdRp is inhibited, viral replication is also inhibited
• 54 samples were tested against SARS-CoV-2; almost 95% of them exhibited downregulated activity, indicating replication was inhibited

Strength of evidence (highlight): Strong / Reliable / Potentially informative / Not informative / Misleading

Why is it important to review this preprint?

• Suggests a new public health intervention to prevent/treat Covid / reduce the severity of illness
• The polio vaccine, unlike Covid vaccines, has already been approved and in some cases has more trust among those hesitant regarding the Covid vaccines; it’s also potentially easier to transport and administer

Recommendations for what questions/issues reviewers can address?

• Small sample size (<300 people), especially with samples tested against SARS-CoV-2 (54 samples)
• There are some missing links between “has antibodies 28 days after inoculation” and “has strong immunity to Covid later on”; perhaps we should have studies comparing people who got IPV to those who didn’t and evaluating whether people who got IPV were less likely to test positive for Covid or be hospitalized / die

What kind of expert should review this preprint?

• An expert in biological/medical topics such as vaccinology

06 October 2021: “Impact of long-COVID on health-related quality of life in Japanese COVID-19 patients”

Main claims:

• Self-reporting from >400 Japanese Covid survivors found statistically significant quality-of-life differences correlated with the persistence of Covid symptoms
• “201 of 457 (44.0%) participants reported at least one symptom after four weeks have passed since their symptom onset due to COVID-19. The most common symptom of long-COVID was general fatigue. 58 of 457 (12.7%) participants have had general fatigue longer than four weeks. The second most common symptom was alopecia. 55 of 457 (12.0%) participants have experienced hair loss worse than usual.” (“As for patients who required supplementary oxygen support, 32 out of 70 (45.7%) presented any symptoms longer than four weeks.”)

Strength of evidence (highlight): Strong / Reliable / Potentially informative / Not informative / Misleading

Why is it important to review this preprint?

• Long Covid is a salient topic and we need more data on it
• Prevention of Covid is potentially more important compared to other respiratory viruses if long Covid imposes a substantial burden and/or if vaccination doesn't help with long Covid

Recommendations for what questions/issues reviewers can address?

• Doesn’t have a control group of people who never had Covid (i.e., no sense of baseline figures for these symptoms)
• Unadjusted confounders (e.g., people with long Covid might be both more likely to have poor health [in ways not addressed by adjustment for confounders in the paper] beforehand and to have adverse consequences from Covid)
• I don’t understand the statistics used in this paper, so someone more familiar with statistics should review it to make sure it makes sense
• This paper is most valuable in combination with a corpus of other long Covid papers
• More data needed on new variants’ tendencies regarding long Covid

29 September 2021: “Global burden of Covid-19 restrictions: National, regional, and global estimates”

Main claims:

• Based on survey data, the quality of life lost to Covid restrictions is larger than commonly assessed to be; in QALYs, it's potentially 38x the years of life lost due to deaths from the virus
• Survey-takers rated some restrictions (e.g., masks) as mostly unproblematic, but were willing to pay substantial portions of their salaries to avert school closures: “across all countries, subjects were willing to give up 22% (95% CI 0.18-0.27) of their annual salary to avoid school closures and willing to give up 21.8% (95% CI 0.17-0.26) to avoid closures of restaurants, bars and clubs. Lowest WTP was observed for removing travel restrictions (7%, 95% CI 0.02-0.12) and wearing masks in public (2% (95% CI -0.02, 0.07).”

Strength of evidence (highlight): Strong / Reliable / Potentially informative / Not informative / Misleading

Why is it important to review this preprint?

• Covid restrictions aren’t costless, and we should consider tradeoffs when making policy (e.g., where children are not as vulnerable to Covid and school closures impose a heavy burden on children and adults, schools should have an in-person option)

Recommendations for what questions/issues reviewers can address?

• Despite a robust sample size, the paper draws on survey-takers reached through Mechanical Turk who are mainly from France, India, Italy, the UK, and the US
• The survey sample did not include people under 18, though they might be uniquely affected by, e.g., school closures
• Findings should ideally be replicated with more varied groups of people across more countries, and more specific targeting can identify better tradeoffs for specific groups of people
• Covid measures aren’t necessarily consistent at the national or regional level, so measures of Covid restriction stringency might not reflect reality
• I drew some of these issues from the ACX post on lockdown effectiveness

What kind of expert should review this preprint?

• An expert in QALY-related assessment and/or survey methods

22 September 2021: “Validation of a saliva-based test for the molecular diagnosis of SARS-CoV-2 infection”

Main claims:

• Saliva-based tests satisfactorily effective compared to PCR tests
• Saliva-based tests produce more false positives than false negatives
• Other advantages of saliva tests (e.g., easier to collect / less invasive; denaturing solution makes it safer for medical personnel to handle; saliva-collected viral RNA remains stable for up to 48 hours)

Strength of evidence (highlight): Strong / Reliable / Potentially informative / Not informative / Misleading

Why is it important to review this preprint?

• We need to make accurate Covid tests widely available; the US in particular has a problem where tests are rarely available, often sold out, and quite expensive, in comparison to countries like Germany--partly because the FDA hasn’t approved many tests
• More options for tests will encourage more people to get tested so we can track and contain the spread of Covid

Recommendations for what questions/issues reviewers can address?

• Small sample size of confirmed Covid-positive patients (n=156)
• Characteristics of new variants (e.g., to what extent viral load is present in saliva) might change/improve efficacy of saliva-based test

What kind of expert should review this preprint?

• Potentially a statistics expert but the statistics look pretty clear to me
• People with expertise in PCR/testing

Rapid reviews: Covid-19 reviews of this preprint