The authors present some of the results of an open-label trial conducted at 2 emergency department during 2 influenza seasons in which patient with confirmed influenza were randomized to either IV peramivir vs. oral oseltamivir. The study was designed as a pilot/feasibility one. Patients were technically followed up for 28 days either outpatient or inpatient in the 2015-16 and 2016-17 influenza seasons in the US. The protocol was not provided with the submission, but it is available together with the SAP at clinicaltrials.gov.
The study is of interest, but despite using the CONSORT guidance, several key elements are missing from the report to properly appraise the findings. Several points follow.

1. The study was a pilot feasibility one, yet a lot of feasibility details are missing in the report. The protocol points to formal study visits/contact on days 3, 7, 14 and 28. Yet, the study required daily FLU-PRO questionnaires to be done daily for 14 days and for daily assessments for hospitalized patients. The protocol says it would be done over the internet, or by contact with study personnel. How was this actually done? Did patients go home with printed questionnaires? If by a combination of methods, it would be good for the readers to understand how this was implemented and how successful this was. How was privacy preserved if using remote methods and the internet?

2. The intensity of daily questionnaires, especially for outpatients, is prone to missing data, yet we are not told in the results how complete the data are, how many patients dropped out from the study before Day 28. Patients may have been too sick or delirious to complete questionnaires. How was missing data handled or analyzed in terms of efficacy and safety endpoints? None of the score graphs shows how many patients contributed to each data point.

3. The study flow diagram needs to be expanded to include post-randomization events, how many patients randomized to each arm, how many patients did not receive study assignment and why, how many patients completed treatment in each arm, how many discontinued and why, how many patients completed the study. Please see the CONSORT statement guidance for this.

4. In the study ~1/3 of patients were hospitalized and ~2/3 went home, which is an interesting mix. There is no mention on the actual antiviral exposure for both arms other than some peramivir patients went on to received oseltamivir. The discussion mentions 20% of patients on oseltamivir did not complete treatment, but this is not in the results. Did every patient on peramivir receive 1 full dose? How many patients received more than 1 dose as the protocol contemplated? How was adherence confirmed if patients were not seen physically again if they went home?

5. The protocol mentions collecting NP swabs at baseline and on day 3 and 7 for hospitalized patients. What was done with these samples? No results are provided regarding this.

6. Can the authors confirm that research coordinators, not licensed physicians, consented the patients for the study? The consent form posted in clinicaltrials.gov suggests that an investigator had to obtain consent. This is an interesting feasibility issue that should be mentioned.

7. When did the investigators or coordinators approach patients for the study in terms of the decision to admit or go home?

8. We only find out the study provided 5 days of oseltamivir for participants in the discussion. Please mention in the methods.

9. The NCT number is provided twice. Would delete from the introduction, just leave the mention in the methods.

10. How was the random sequence generated and shared between sites? Was there any block randomization to guarantee similar number of participants per arm? How was the randomization sequence concealed? This is important as the study was open label, so If the randomization sequence was knowable or predictable, patients could be selectively approached for consent.

11. How many patients in each arm used the rescue peramivir mentioned if not better by day 5?

12. None of the planned comparisons of adherence are presented in the results.

13. Were patients “encountered” or “seen”/”evaluated” in the ED?

14. The fact that 2/3 of patients were African American is important to highlight, as they have been underrepresented in other influenza randomized trials reported to date.

15. If the authors were targeting higher-risk patients, makes more sense to mention who many patients were >65-years old, or just present the median and ranges of ages of the participants, which I favor.

16. The reports moves into comparing groups in terms of IV vs. oral. I think this is inappropriate. The study is comparing 2 different drugs, one with a long half-life that is only available IV, and an oral one with a shorter half-life that is given twice daily for patients with normal renal function. Would say peramivir vs. oseltamivir throughout the text and in the abstract.

17. I would cite the harder endpoints before going into the questionnaire descriptions: how many patients completed treatment, how many patients completed the study, how many lost to follow up.

18. As these drugs are approved for patients who present within 48h of symptoms, but less than 50% were in this situation, I think it is pertinent to present these results. They are mentioned in passing in the discussion, but without specifics which should be in the results.

19. All plots for scores should have error bars and mention the number of patients assessed at each data point per arm.

20. The claims to be the first in the initial paragraph of the discussion are a bit exaggerated. There have been several influenza trials of hospitalized patients and ambulatory patients were patients were diagnosed and assessed in emergency rooms initially. What do we really learn from this study that combined both types of patients? Would think this through.

21. Zanamivir (check spelling) is mentioned in the introduction as a one-dose IV drug for influenza—This is incorrect. The drug is approved in the US in the inhaled form only. The IV formulation was approved in Europe this year, but it is given twice or once daily depending on renal function. The authors may want to compare their results to the IV zanamivir vs. oral oseltamivir trial published in 2017 in Lancet. The authors should also discuss/mention prior randomized trials that compared IV peramivir vs placebo to standard of care/oral oseltamivir.