Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1.　While this study provides valuable insights into the use of droxidopa for vasopressor withdrawal in the ICU, its major limitations lie in its single-center, retrospective, observational nature and the lack of a comparison group. This prevents a clear comparison of the efficacy of droxidopa with other treatments, limiting its external validity. We recommend that this point be more clearly described in the "Discussion" section and that specific proposals be made for future prospective and comparative studies.

2.　The primary endpoint was "cessation of IV vasopressors within 72 hours of droxidopa initiation." More specific explanation of the clinical significance and rationale for this timeframe would enhance reader understanding.

3.　While the study describes concomitant use of opioids, steroids, and sedatives, it would be beneficial to further address the potential impact of these medications on hemodynamics and vasopressor withdrawal in the discussion section.

Author Response

Please see the attachment 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Thanks to the authors for this retrospective paper on droxidopa addition to midodrine for patients on IV vasopressors.

 

Overall it is well written. Methodologically it is tough with the lack of control group. At least some patients on IV vasopressors for 17 days, are likely to have them stop of their own accord. 

 

I am interested in why patients who 

"clinically deteriorated after droxidopa initiation requiring vasopressor escalation for >12 hour" were excluded? Do you mean patients were deteriorating anyway? But it would be useful to understand downsides of initiating droxidropa. Could you describe this population and outcomes?

Can you provide norephinephrine doses (or NEE) received by patients at administration of droxidopa?

And then it might be useful to plot step change in norephinephrine pre and post droxidopa initiation? 

Another reference similar style to your paper : https://pubmed.ncbi.nlm.nih.gov/36168754/

You have the indication for admission to ICU, but not the indication for vasopressor, that could be helpful.

Median of 9 days of droxidopa, but you are reporting cessation @ 72hours, and time to cessation of 2.3 days. Can you explain the reasons for continuation out to 9 days?

 

 

 

 

 

 

Author Response

Please see the attachment 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors studied the role of droxidopa in weaning of IV vasopressor in the ICU patients. They stated that norepinephrine as the most common agent they tried to wean off. It would be more informative to the readers if they discussed in brief about the pharmacology such as t1/2 and steady state concentration of both the NE infusion and oral dose of droxidopa. Droxidopa was used for treatment of orthostatic hypotension.30/42 patients are undergoing dialysis, who are more prone to significant hemodynamic fluctuations. Could this have been a treatment bias to wean this patient of norepinephrine drip?

Author Response

Please see the attachment 

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

This single-center retrospective descriptive study evaluates the safety and effectiveness of droxidopa to facilitate IV vasopressor weaning among ICU patients, many already receiving midodrine. Of 46 included patients (38 on IV vasopressors at droxidopa start), 44–45% discontinued IV vasopressors within 72 hours (median 2.3 days). No adverse events were captured through chart review. About 29% were discharged on droxidopa. The topic is clinically relevant given the operational burden of “vasopressor-dependent but not in shock” patients and the mixed data for midodrine.

Major Comments

Study design and causal inference

The retrospective single-arm design without a comparator (e.g., continued midodrine optimization or usual care) precludes any causal claims about droxidopa effectiveness. The manuscript should frame findings as descriptive/associative and avoid language implying efficacy. Consider adding a contemporaneous control or historical matched cohort in a sensitivity analysis, or, at least, acknowledge the direction and magnitude of likely confounding by indication (clinicians may initiate droxidopa when a patient is already improving or, conversely, more refractory).

Population definition and denominators

Inclusion required “>4 consecutive doses” of droxidopa for vasopressor weaning, yet 17% (8/46) were not on IV vasopressors at droxidopa initiation. Clarify why these patients were eligible (e.g., to prevent re-initiation) and ensure the primary analysis set is clearly pre-specified.

Explain missingness: Table 1 footnotes indicate fewer patients with recorded MAP/SBP goals and midodrine dose (e.g., midodrine dose based on N=30 although 42 were on midodrine). Report the extent of missing data and how it was handled.

Primary outcome and analytic choices

The 72-hour endpoint is pragmatic, but potentially arbitrary. Provide clinical rationale and perform additional time-to-event analyses (Kaplan–Meier curves for vasopressor discontinuation; median time with IQR) and a 7-day competing-risk analysis (death/deterioration as a competing event).

Present a multivariable model (e.g., logistic regression for 72-hour discontinuation, or Cox model for time to cessation) adjusting for key confounders: age, ICU type, illness severity (APACHE II/SOFA), vasopressor burden (norepinephrine-equivalents), CRRT, mechanical ventilation, steroid use, and midodrine dose.

Exposure characterization

Dosing strategy is heterogeneous. Provide a dosing/titration algorithm (initial dose, increments, max dose, criteria for escalation) and report dose intensity at specified time points (e.g., day 1, day 2, at weaning). Consider dose-response analyses (e.g., max TTD vs. success).

Reconcile pharmacology language: droxidopa is a prodrug converted to norepinephrine (which acts at alpha- and beta-adrenergic receptors); droxidopa itself is not an alpha/beta agonist. Edit for pharmacologic precision.

Concomitant therapies and BP targets

Concomitant midodrine (91%), corticosteroids (57%), opioids (60%), sedatives (19%) and variable BP goals (often MAP ≥60) are strong, time-varying confounders. Provide:

• A table of concomitant hemodynamically active medications at baseline and within 72 h (start/stop/intensification).
• Analyses stratified by midodrine dose (e.g., ≥80 mg/day), steroid exposure, BP target (MAP 60 vs. 65+), dialysis modality (iHD vs. CRRT), and ICU type.

Report vasopressor dose trajectories (preferably norepinephrine-equivalents) before and after droxidopa initiation to contextualize “weaning.”

Safety ascertainment

“No hypotension, hypertension, arrhythmias” identified via progress notes likely underestimates events. Strengthen safety capture by querying structured vitals (e.g., SBP/MAP distributions, proportion of readings <55 or >140), telemetry logs for arrhythmias, heart rate trends, and need for rescue therapy. Provide clear operational definitions for adverse events and report them systematically.

Since 29% were discharged on droxidopa, include post-discharge outcomes (30-day readmissions, ED visits for hypertension/arrhythmia), if available. If not available, acknowledge as a limitation.

Selection and survivorship bias

Excluding patients who clinically deteriorated after droxidopa (>12 h escalation or comfort care within 48 h) may bias toward favorable outcomes. Provide a flow diagram with explicit counts and characteristics of excluded patients and discuss direction of bias.

Illness severity and generalizability

Provide baseline illness severity (SOFA/APACHE II), vasopressor dose at index, and lactate where available to contextualize acuity and compare with prior studies.

Clarify which surgical/transplant populations dominate and whether findings apply beyond a high-acuity tertiary cohort with long ICU LOS (median 28 days).

Economic considerations

The pricing comparison (midodrine vs droxidopa) should normalize by defined daily dose and local acquisition costs. If cost is discussed, consider a simple budget impact estimate (drug costs vs. potential ICU day savings), with conservative assumptions, or limit to qualitative comments.

Reporting standards

Add a STROBE checklist; ensure the abstract states design (“retrospective descriptive single-center study”) and primary/secondary endpoints explicitly.

Provide a pre-specified analysis plan (even if retrospective) and distinguish exploratory from confirmatory analyses.

Context within Literature

The discussion references two small retrospective studies; please expand the comparison with standardised endpoints (72 h and 7 d), patient acuity, and concurrent midodrine/steroid use. Emphasise the small overall evidence base (<100 patients total across studies) and align conclusions accordingly.

Minor Comments

Terminology & clarity

Standardize “72 hours” (several places write “72”) and “TTD” on first use.

Replace “alpha/beta agonist” description of droxidopa with accurate pharmacology (prodrug to NE; NE is the agonist).

Define “refractory to midodrine” prospectively (e.g., failure to meet MAP goal despite ≥X mg/day and Y hours).

Clarify if droxidopa started in addition to or instead of midodrine, and when substitution occurred.

Tables & figures

Table 1: ensure consistent denominators; explain missing MAP/SBP goals; consider adding vasopressor dose (NE-equivalents), SOFA/APACHE II, lactate, and dialysis indication.

Table 2: specify whether “Initial TTD” refers to day-1 total and whether the median max TTD reflects peak within ICU.

Table 3: align “45% within 72 h” with text “44%”; ensure consistent rounding.

Provide Figure 1 (CONSORT-style diagram) with explicit reasons for exclusions and counts at each step.

Add a Kaplan–Meier curve for time to vasopressor discontinuation (overall and stratified by CRRT, midodrine dose, BP goal).

Editing & consistency

Check author names for consistency.

Fix typographical issues: broken CC BY URL, spacing (e.g., “tar-34 gets,” “maybe” vs “may be”).

Use consistent ICU unit names (CVICU, SICU, NeuroICU, etc.) and capitalise drug names.

Methods detail

Describe the institutional data repository query, codes/keywords used, and validation approach.

Specify how MAP/SBP goals were obtained (order sets vs notes) and whether they changed after droxidopa.

Detail timing anchors (T0 = first droxidopa dose) and data windows (e.g., −24 to +72 h) for medication and vital sign extraction.

Safety

Report HR changes (median ΔHR at 6–12–24 h), new arrhythmia incidence from telemetry, and need for PRN antihypertensives.

State whether nocturnal supine hypertension screening was performed (relevant to droxidopa).

Ethics/data

Retrospective consent waiver: explicitly state that consent was waived by the IRB.

Data availability: specify what data can be shared (de-identified dataset, codebook) and under what conditions.

Limitations (to highlight more explicitly)

Single-center, retrospective, no control group; high risk of confounding by indication and selection bias.

Heterogeneous co-interventions (midodrine, steroids, sedation), variable BP targets, and incomplete capture of vasopressor dose intensity.

Safety capture via notes likely under-detects events; no structured vitals/telemetry analysis.

Long ICU/hospital LOS suggests a highly selected cohort; limited generalisability.

Suggested Additions/Analyses (prioritised)

Add Kaplan–Meier curves (overall and stratified) and a multivariable Cox/logistic model.

Report vasopressor dose trajectories (NE-equivalents) and BP/HR time series (−24 to +72 h).

Stratify outcomes by: midodrine dose/intensity, BP goal, CRRT vs iHD/no dialysis, ICU type, and steroid exposure.

Tighten definitions (refractory midodrine; adverse events) and reconcile denominators/missingness.

Refine safety capture using structured data (vitals/telemetry) and report any post-discharge events for those discharged on droxidopa.

Author Response

Please see the attachment 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for the revisions and detailed responses to the reviewer’s comments.
The authors have appropriately addressed all the concerns raised in the previous review, including clarification of the study limitations, rationale for the 72-hour primary endpoint, and discussion of the effects of concomitant medications.
The revisions are satisfactory, and the manuscript has been adequately improved.
No further changes are required from my perspective.

Author Response

Thank you for your suggestions 

Reviewer 2 Report

Comments and Suggestions for Authors

Thanks for the revisions.

 

A few comments:

 

1. . I am not entirely clear on your answer to my previous query about cessation of droxidopa in patients who deteriorated. " clinically deteriorated after droxidopa initiation requiring vasopressor escalation for >12 hours" - how do you know this was not a complication of droxidopa initiation? Might be, might not be. 
2. I would presume Stanford Health Care has an EMR. Therefore you should be able to get further data on the patients, on doses of vasopressors. For me this is really important data you need to describe. The addition you made of "“Vasopressor(s) were discontinued within 72 hours of droxidopa initiation in 58% (15/26) of patients who were on one vasopressor at the time of droxidopa initiation compared to 27% (3/11) of patients who were on two or three vasopressors" - is not really a comparison between treatment groups, just different groups of severity. 
3. Abstract - unnecessary additional "a" . "Cessation of IV vasopressor without re-initiation within 72 hours of droxidopa initiation was the primary outcomea."

Author Response

Please see attachment 

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

Thank you for the revised version of your submission. Now it is acceptable. Best wishes.

Author Response

Thank you for your suggestions