HBOT, hyperinflammatory burden, comorbidity burden and drug exposure in relation to repeated-measures respiratory severity in hospitalized adults with COVID-19: a post hoc secondary analysis of a randomized clinical trial

HBOT, hyperinflammatory burden, comorbidity burden and drug exposure in relation to repeated-measures respiratory severity in hospitalized adults with COVID-19: a post hoc secondary analysis of a randomized clinical trial | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article HBOT, hyperinflammatory burden, comorbidity burden and drug exposure in relation to repeated-measures respiratory severity in hospitalized adults with COVID-19: a post hoc secondary analysis of a randomized clinical trial Natalia Jermakow, Klaudia Brodaczewska, Jacek Kot, Arkadiusz Lubas, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9389851/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 8 You are reading this latest preprint version Abstract Background Respiratory failure in COVID-19 reflects both impaired gas exchange and systemic inflammation, and treatment effects may be missed when only single oxygenation metrics are assessed. Methods We performed a post hoc repeated-measures secondary analysis of 28 randomized adults from a clinical trial of standard care alone or standard care plus hyperbaric oxygen therapy (HBOT). Respiratory severity was represented by a modified respiratory SOFA (rSOFA) score derived from PaO₂/FiO₂ and advanced respiratory support, inflammatory burden by a reduced cHIS-derived score, comorbidity burden by a count of Quan/Elixhauser-style ICD-10 categories, and medication context by ATC3-derived drug axes. Four ordinal generalized estimating equation models were fitted: HBOT moderation, comorbidity moderation, drug-axes moderation, and a global main-effects model. Results In the HBOT moderation model, higher cHIS was associated with worse respiratory severity (OR 5.17, 95% CI 2.44–10.94), whereas the cHIS × HBOT interaction was negative (OR 0.20, 95% CI 0.08–0.49). In the comorbidity model, the cHIS effect remained positive (OR 2.75, 95% CI 1.18–6.41), while the Elixhauser main effect and interaction were imprecise. The ATC3 drug-axes model showed the widest coefficient spread and was interpreted cautiously. Conclusions In this exploratory repeated-measures reanalysis, the most coherent signal was a weaker association between hyperinflammatory burden and respiratory worsening in the HBOT arm than in the standard-care arm. These findings are hypothesis-generating and should not be interpreted as proof of efficacy. Trial registration: EudraCT 2020-002722-90, registered 3 May 2020. COVID-19 hyperbaric oxygen therapy respiratory failure respiratory SOFA cHIS comorbidity burden Background Coronavirus disease 2019 (COVID-19) is now understood as a systemic illness in which respiratory failure emerges from an interaction between alveolar injury, endothelial dysfunction, vascular abnormalities and dysregulated inflammation rather than from gas-exchange impairment alone [ 1 – 4 ]. Patients with apparently similar levels of hypoxemia can follow different clinical trajectories because oxygen requirement, inflammatory activity and supportive care intensity change over time. This heterogeneity is clinically important and also creates a methodological problem for treatment studies that rely on a single oxygenation index as the primary lens through which respiratory course is interpreted. The inflammatory component of COVID-19 has substantial prognostic relevance. Higher interleukin-6 concentrations and related inflammatory abnormalities are associated with severe disease, and longitudinal profiling has shown that adverse trajectories are characterized not only by high inflammatory peaks but also by persistence of maladaptive immune programs [ 5 – 7 ]. Webb and colleagues formalized this concept in the COVID-19-associated hyperinflammatory syndrome (cHIS) score, a pragmatic framework linking hyperferritinaemia, coagulopathy, cytokinaemia and hematologic dysfunction to worsening oxygen requirement, mechanical ventilation and mortality [ 8 ]. That framework is clinically attractive because it captures a multidimensional inflammatory state rather than any single laboratory abnormality. Hyperbaric oxygen therapy (HBOT) has been explored as an adjunctive treatment in COVID-19 because it increases dissolved oxygen content in plasma and may also influence inflammatory and endothelial responses [ 10 , 14 , 15 ]. Clinical evidence, however, remains mixed. Our randomized trial demonstrated feasibility and safety, lower normobaric oxygen requirements in the HBOT arm, and favorable changes in selected inflammatory and immune variables, but it did not show a statistically significant between-group difference in serial PaO₂/FiO₂ values [ 9 ]. Other randomized studies and systematic reviews likewise suggest that HBOT is biologically plausible and feasible, while still falling short of establishing a uniform treatment effect across all clinical endpoints [ 10 – 13 ]. That inconsistency makes endpoint choice critical. If treatment influences oxygen support needs or changes the coupling between inflammation and respiratory deterioration, then reliance on raw PaO₂/FiO₂ alone may dilute clinically relevant information. A patient whose oxygenation is maintained with lower respiratory support may be improving even when the PaO₂/FiO₂ ratio itself does not show a dramatic between-group divergence. This issue is especially relevant in COVID-19, where respiratory severity is closely intertwined with supportive strategies such as high-flow nasal oxygen therapy [ 3 , 4 , 18 – 20 ]. The present study was designed as a post hoc secondary analysis of that randomized trial. Instead of asking only whether HBOT improved a single oxygenation metric, we examined whether HBOT was associated with the time-varying distribution of a modified respiratory SOFA-type severity score derived from PaO₂/FiO₂ and advanced respiratory support, and whether this relationship varied with hyperinflammatory burden over repeated assessments. Because the original dataset was small, we also summarized comorbidities and medication exposures into clinically named binary axes derived from ICD-10 diagnoses and ATC3 drug classes rather than entering many sparse variables separately [ 23 – 26 ]. Our hypothesis was deliberately conservative. We did not assume that HBOT would exert a uniform main effect across the entire cohort. Rather, we asked whether the signal in this dataset was better captured as attenuation of the association between hyperinflammatory burden and respiratory worsening over time. To address that question, we compared four repeated-measures ordinal GEE model families: an HBOT moderation model, a disease model, a drug-exposure model, and a broader global model without interaction terms. Methods Study design and patients This manuscript reports a post hoc secondary analysis of a previously conducted prospective randomized clinical trial comparing standard care alone with standard care plus HBOT in hospitalized adults with COVID-19 pneumonia. The primary clinical outcomes of the parent randomized trial were reported previously [ 9 ]. The trial was registered in EudraCT (2020-002722-90), approved by the local ethics committee, and conducted after written informed consent. From 1 March 2021 to 3 February 2022, 30 adults hospitalized with RT-PCR-confirmed COVID-19 pneumonia at the Military Institute of Medicine – National Research Institute in Warsaw were randomly assigned to HBOT plus standard care or standard care alone. Two patients were excluded after randomization because they did not meet eligibility criteria, leaving 28 participants (14 per group) for the present analysis. HBOT consisted of five daily sessions at 2.5 ATA for 75 minutes. Standard of care included anticoagulation and corticosteroids in all patients. No serious HBOT-related adverse events were recorded. Three deaths occurred in the control arm and none in the HBOT arm; these were attributed to COVID-19 progression and occurred after day 10. The participant flow of the parent trial is shown in a CONSORT flow diagram [see Additional file 1]. Outcome and inflammatory predictor Respiratory severity was quantified using a modified respiratory Sequential Organ Failure Assessment score (rSOFA) derived from the Horowitz index (PaO₂/FiO₂) together with advanced respiratory support [ 18 – 20 ]. High-flow nasal oxygen therapy and invasive mechanical ventilation were treated as advanced respiratory support because both reflect clinically meaningful escalation of respiratory care in contemporary COVID-19 management [ 19 , 20 ]. As implemented in the analysis code, rSOFA was assigned as 0 when PaO₂/FiO₂ exceeded 400, 1 for values of 301–400, 2 for values ≤ 300 without advanced respiratory support, 3 for advanced respiratory support with PaO₂/FiO₂ ≤200, and 4 for advanced respiratory support with PaO₂/FiO₂ ≤100. The resulting ordinal outcome therefore ranged from 0 to 4, with higher categories indicating worse respiratory severity. Inflammatory burden was summarized using a reduced cHIS-derived score based on ferritin ≥ 700 µg/L, D-dimer ≥ 1.5 µg/mL, LDH ≥ 400 U/L, NLR ≥ 10, and IL-6 ≥ 15 pg/mL or CRP ≥ 15 mg/dL [ 8 , 21 , 22 ]. This reduced score preserves the principal inflammatory domains of the original cHIS framework while omitting the fever component, which was not consistently available in the trial dataset [ 8 ]. Comorbidity burden and medication axes Comorbidity burden was operationalized in code as a patient-level count of Quan/Elixhauser-style categories derived from ICD-10-coded secondary diagnoses [ 24 , 25 ]. ICD-10 subcodes were normalized and matched by prefix/range rules to Elixhauser-style categories in the analysis script; in this cohort, the mapped burden categories contributing to the final count were hypertension, chronic pulmonary disease, pulmonary-circulation disorders, arrhythmia, and obesity. The burden covariate used in modeling was the count of present categories per patient. Uncomplicated diabetes was not double-counted when complicated diabetes was present, and non-metastatic solid tumor was not double-counted when metastatic cancer was present. ATC3-derived treatment data were collapsed into four binary exposure axes defined directly in code by class prefixes: cardiovascular drugs (B01A, C01, C02, C03, C07, C08, C09, C10), respiratory drugs (R03, R05, R06), immunomodulatory drugs (H02A, L04A), and antibiotics (J01) [ 26 ]. Vaccination data were merged descriptively from the trial vaccine file but were not included in the ATC3-only drug models. Statistical analysis For the present analysis, day 1 was defined as the baseline pre-HBOT assessment, whereas day 10 represented follow-up after completion of five HBOT sessions. Intermediate time points reflected repeated in-hospital assessments during treatment. Missing data were supplemented using the last observation carried forward (LOCF) approach. Data processing and modeling were performed in Python using pandas, NumPy, statsmodels, matplotlib, and seaborn. Population-averaged ordinal generalized estimating equation models were fitted with the statsmodels OrdinalGEE implementation, patient-level clustering, and an independence working-correlation structure [ 16 , 17 ]. Four models were compared: an HBOT moderation model including cHIS, HBOT and their interaction; a comorbidity moderation model including Elixhauser burden and a cHIS-by-burden interaction; a drug-axes moderation model including ATC3-derived cardiovascular, respiratory, immunomodulatory, and antibiotic axes together with cHIS-by-drug interactions; and a global model including HBOT, Elixhauser burden, and ATC3-derived drug axes as main effects only. Effects are reported as odds ratios with 95% confidence intervals. Because this manuscript reports a post hoc secondary analysis of an already completed randomized trial, no new sample size calculation was performed for the present analysis; all effect estimates were interpreted as exploratory and hypothesis-generating rather than confirmatory. Results Of the 30 randomized patients, 28 were analyzed (14 HBOT, 14 control). Baseline characteristics were similar between groups (Table 1 ), including age (52.8 ± 13.5 vs. 56.1 ± 14.0 years; P = 0.52), sex (14.3% vs. 21.4% women; P > 0.99), NEWS (median 2 in both groups; P = 0.42), SpO2 (96.8 ± 2.8% vs. 93.4 ± 9.3%; P = 0.27), CRP (3.8 ± 4.4 vs. 5.4 ± 4.3; P = 0.27), and PCT (0.09 ± 0.07 vs. 0.18 ± 0.21; P = 0.22). Confirmed COVID-19 vaccination was documented in 2 HBOT-treated patients and 1 control patient (P = 1.000). Three deaths occurred in the control group and none in the HBOT group (21.4% vs. 0%; P = 0.217), all were attributed to COVID-19 progression or its complications and occurred after day 10. Table 1 Baseline characteristics of randomized participants by treatment group. Legend: Values are shown as mean ± SD unless otherwise indicated. NEWS is presented as median (IQR). Characteristic Control HBOT Age, mean ± SD 56.1 ± 14.0 52.8 ± 13.5 Female, n (%) 3 (21.4) 2 (14.3) Deaths, n (%) 3 (21.4) 0 Vaccinated, n (%) 1 (7.1) 2 (14.3) NEWS, median (IQR) 2 (1) 2 (0) Baseline SpO₂ 93.4 ± 9.3 96.8 ± 2.8 Baseline CRP 5.4 ± 4.3 3.8 ± 4.4 Baseline PCT 0.18 ± 0.21 0.09 ± 0.07 For repeated-measures modeling, 28 patients contributed 168 repeated observations (baseline, day 2–5, follow-up on day 10). A descriptive summary of the patient-level and observation-level variables entered into the final models is provided in Table 2 . Observation-level rSOFA was concentrated in category 2 (140/168, 83.3%), with fewer observations in categories 0, 1, 3, and 4. Observation-level cHIS values spanned the full 0–5 range, most commonly categories 1–3. At the patient level, Elixhauser burden was 0 in 14 patients, 1 in 11 patients, and ≥ 2 in 3 patients; ATC3-derived cardiovascular, respiratory, immunomodulatory, and antibiotic axes were present in 28, 18, 26, and 21 patients, respectively. Table 2 Summary of variables included in the repeated-measures model dataset. rSOFA and cHIS counts are observation-level counts across the 168-row repeated-measures dataset. Treatment allocation, Elixhauser burden, ATC3-derived drug axes, and vaccination are patient-level counts across 28 randomized participants. Domain Variable Samples n (%) Treatment allocation HBOT 14 (50.0%) Control 14 (50.0%) Observed rSOFA category rSOFA 0 5 (3.0%) rSOFA 1 13 (7.7%) rSOFA 2 140 (83.3%) rSOFA 3 6 (3.6%) rSOFA 4 4 (2.4%) Observed cHIS score cHIS 0 28 (16.7%) cHIS 1 37 (22.0%) cHIS 2 53 (31.5%) cHIS 3 28 (16.7%) cHIS 4 14 (8.3%) cHIS 5 8 (4.8%) Patient-level comorbidity burden Elixhauser count 0 14 (50.0%) Elixhauser count 1 11 (39.3%) Elixhauser count ≥ 2 3 (10.7%) Patient-level ATC3 drug axes Cardiovascular drugs 28 (100.0%) Respiratory drugs 18 (64.3%) Immunomodulatory drugs 26 (92.9%) Antibiotics 21 (75.0%) Descriptive context Vaccinated 3 (10.7%) The clearest signal was observed in the HBOT moderation model (Fig. 1 a,e). Higher cHIS was associated with higher odds of worse respiratory severity (OR 5.17, 95% CI 2.44–10.94), whereas the cHIS × HBOT interaction was negative (OR 0.20, 95% CI 0.08–0.49). The HBOT main effect alone was not clearly different from the null (OR 1.89, 95% CI 0.25–14.14). The negative interaction term indicates a shallower association between hyperinflammatory burden and worsening respiratory severity in the HBOT arm than in the standard-care arm. The comorbidity moderation model (Fig. 1 b,f) replaced the earlier diagnosis-axis structure with a count-based Elixhauser burden derived from ICD-10-coded comorbid diagnoses. In this model, the cHIS main effect remained positive (OR 2.75, 95% CI 1.18–6.41), whereas the Elixhauser burden main effect (OR 0.77, 95% CI 0.13–4.49) and the cHIS × Elixhauser interaction (OR 0.82, 95% CI 0.34–1.98) were imprecise and did not clearly depart from the null. These estimates were centered near the null and were less precise than those observed in the HBOT moderation model. Figure 1 summarizes the final model coefficients and the observed cHIS–rSOFA patterns most relevant to interpretation. Additional model diagnostics and fitted-score distributions are provided in Additional file 2. The global main-effects model without moderation (Fig. 1 d) included HBOT, Elixhauser burden, and ATC3-derived drug axes as main effects only. This broader model retained a favorable HBOT-associated coefficient (OR 0.16, 95% CI 0.03–0.98) together with a positive cHIS coefficient (OR 3.06, 95% CI 1.43–6.57). However, the cardiovascular-drug coefficient expanded to an implausibly large value, again indicating that the main-effects model should be read as a broad contextual summary rather than a mechanistic explanation. Figure 1 . Main model results and observed distributions. Panels a–d show forest plots for HBOT moderation, comorbidity moderation, drug-axes moderation, and the global main-effects model without moderation. Points indicate odds ratios and horizontal lines indicate 95% confidence intervals on a logarithmic scale. Panel e shows observed cHIS versus observed respiratory SOFA stratified by HBOT versus control, with jittered points and median trend lines. Panel f shows observed cHIS versus observed respiratory SOFA with point color intensity indicating Elixhauser burden. Discussion The main finding of this post hoc secondary analysis is not that HBOT produced a straightforward, across-the-board increase in PaO₂/FiO₂. That claim would be incompatible with the original randomized report, which did not detect a statistically significant between-group difference in serial PaO₂/FiO₂ values [ 9 ]. Instead, the present analysis suggests something narrower but more clinically coherent: worsening hyperinflammatory burden tracked with worse respiratory severity across repeated assessments, and this association was markedly weaker in patients assigned to HBOT. This distinction matters because the HBOT literature in COVID-19 has often struggled with endpoint interpretation. On physiological grounds, HBOT is attractive in hypoxemic respiratory failure because it transiently increases dissolved oxygen in plasma while potentially influencing vascular and inflammatory pathways [ 10 , 14 , 15 ]. But transient tissue hyperoxia, modulation of inflammatory signaling and clinically meaningful respiratory improvement are not interchangeable phenomena. A trial can show lower oxygen requirements without proving a uniform shift in raw PaO₂/FiO₂, and a biomarker change does not by itself establish a treatment effect on respiratory trajectory. The present analysis is therefore best read as a reframing of the parent trial signal rather than as a retrospective rescue of a negative oxygenation endpoint. The central role of the reduced cHIS-derived score in our models is biologically plausible. Hyperferritinaemia, elevated D-dimer, LDH release, neutrophil-to-lymphocyte imbalance and cytokinaemia-related markers all map onto features of severe COVID-19 that have repeatedly been associated with progression to respiratory failure and death [ 5 – 8 , 21 , 22 ]. Within that framework, the HBOT moderation model is clinically coherent because it tests whether treatment assignment changes the relationship between inflammatory burden and respiratory trajectory rather than assuming that treatment acts independently of inflammatory context. The broader models help define the limits of interpretation. The global main-effects model also favored HBOT, but its treatment coefficient necessarily pooled inflammatory burden, comorbidity burden, and medication context into a single structure. The comorbidity model—now based on a count-based Elixhauser burden derived from ICD-10 comorbid diagnoses rather than ad hoc diagnosis axes—did not identify a robust cHIS-by-burden interaction and therefore did not provide a clearer explanation than the HBOT moderation model. That point is particularly important for the drug-axes and global context models. Antibiotic and immunomodulatory therapy were not randomly distributed in this cohort, and some exposure patterns were uncommon. Accordingly, the extreme coefficients in the drug-axes model and the implausibly inflated cardiovascular-drug coefficient in the global model should not be read as evidence of strong independent medication-class effects on respiratory course. They are better understood as markers of clinical context in a small inpatient dataset in which part of the observed signal may reflect treatment of severe COVID-19 itself. We therefore retained these analyses because they describe the structure present in the data, but we interpret them cautiously and do not treat them as causal findings. Our results fit the broader HBOT-in-COVID literature in a nuanced way. Randomized evidence remains limited and heterogeneous, with one trial reporting support for efficacy in severe hypoxemia and another failing to show benefit for major hard outcomes [ 10 , 11 ]. Systematic reviews have concluded that HBOT appears feasible and safe but that the evidence base is still insufficient for firm efficacy claims [ 12 , 13 ]. The present study is consistent with that literature if one avoids demanding a simple yes-or-no answer: it supports the possibility that HBOT may matter most where inflammatory burden and respiratory deterioration are tightly coupled, while stopping well short of proving a uniform treatment benefit. Because the present manuscript is a secondary analysis rather than a primary trial report, these findings should be interpreted alongside, not instead of, the parent trial publication and its CONSORT-based reporting [ 9 ]. The present analysis also has clear limitations. The cohort was small, some exposure patterns were sparse, and the derived rSOFA, reduced cHIS, Elixhauser-burden count, and ATC3 drug axes were code-derived summaries tailored to the available trial dataset rather than external instruments reproduced in full. Missing data were supplemented using LOCF, and the GEE models used an independence working-correlation structure to maintain stable population-averaged inference in a limited sample. These are appropriate exploratory choices for this dataset, but they necessarily constrain the strength of any claim that can be made from the results. Even with those limitations, the repeated-measures framework adds useful information to the parent trial. It suggests that in this dataset HBOT is most plausibly linked to respiratory course through attenuation of the inflammation–severity relationship rather than through a simple across-the-board oxygenation effect. Future studies should test this hypothesis prospectively in larger cohorts using respiratory-support trajectories and repeated inflammatory markers as paired repeated-measures endpoints. Conclusions In this repeated-measures analysis of hospitalized patients with COVID-19, higher hyperinflammatory burden was associated with worse respiratory severity over time, and this association appeared less pronounced in patients assigned to HBOT. This pattern may indicate that the relationship between inflammation and respiratory course differs according to treatment context, but the present study was not designed to establish a definitive treatment effect. The comorbidity- and medication-based models provided additional clinical context, although these results were less stable and should be interpreted cautiously. Because of the small sample size, sparse exposure patterns, exploratory post hoc modeling strategy, and the use of derived summary variables, these findings should be regarded as hypothesis-generating. Larger prospective studies with prespecified respiratory and inflammatory repeated-measures endpoints are needed to determine whether the observed pattern is reproducible and clinically meaningful. They should be interpreted within the limits of a small post hoc secondary analysis of a single randomized trial. Abbreviations ATC3: Anatomical Therapeutic Chemical Classification System, 3rd level; cHIS: COVID-19-associated hyperinflammatory syndrome; CI: confidence interval; COVID-19: coronavirus disease 2019; EudraCT: European Union Drug Regulating Authorities Clinical Trials Database; GEE: generalized estimating equation; HBOT: hyperbaric oxygen therapy; ICD-10: International Statistical Classification of Diseases and Related Health Problems, 10th Revision; IL-6: interleukin 6; LDH: lactate dehydrogenase; LOCF: last observation carried forward; NLR: neutrophil-to-lymphocyte ratio; OR: odds ratio; PaO₂/FiO₂: arterial oxygen partial pressure to inspired oxygen fraction ratio; rSOFA: modified respiratory Sequential Organ Failure Assessment. Declarations Ethics approval and consent to participate The parent trial was conducted in accordance with the Declaration of Helsinki and approved by the Bioethics Committee of the Military Institute of Medicine – National Research Institute (approval no. 25/WIM/2020). All participants provided written informed consent in the original trial [9]. Consent for publication Not applicable. Availability of data and materials The datasets analysed during the current study are not publicly available because they derive from a completed clinical trial dataset containing potentially identifiable clinical information, but are available from the corresponding author on reasonable request. Competing interests The authors declare that they have no competing interests. Funding The parent randomized trial was funded by the Polish Medical Research Agency (grant 2020/ABM/COVID19/0043). No additional external funding was obtained for the present post hoc repeated-measures analysis. Authors' contributions Conceptualization: J.S., K.B. and J.K.; methodology: J.S., K.B., N.J. and J.K.; formal analysis: N.J., J.S., K.B. and J.K.; investigation: J.S., K.B., A.L. and K.K.; writing—original draft preparation: N.J., J.S. and J.K.; writing—review and editing: N.J., J.K, J.S., K.B., A.L. and K.K.; funding acquisition, J.S. and J.K. All authors read and approved the final manuscript. Acknowledgements The authors are especially indebted to all employees of the designated COVID-19 clinics of the Military Medical Institute involved in the implementation of this clinical trial. Authors' information Not applicable. References Wiersinga WJ, Rhodes A, Cheng AC, Peacock SJ, Prescott HC. Pathophysiology, transmission, diagnosis, and treatment of coronavirus disease 2019 (COVID-19): a review. JAMA. 2020;324:782–93. 10.1001/jama.2020.12839 . Varga Z, Flammer AJ, Steiger P, et al. Endothelial cell infection and endotheliitis in COVID-19. Lancet. 2020;395:1417–8. 10.1016/S0140-6736(20)30937-5 . ARDS Definition Task Force, Ranieri VM, Rubenfeld GD, et al. Acute respiratory distress syndrome: the Berlin definition. JAMA. 2012;307:2526–33. 10.1001/jama.2012.5669 . Gattinoni L, Chiumello D, Caironi P, et al. COVID-19 pneumonia: different respiratory treatments for different phenotypes? Intensive Care Med. 2020;46:1099–102. 10.1007/s00134-020-06033-2 . Coomes EA, Haghbayan H. Interleukin-6 in COVID-19: a systematic review and meta-analysis. Rev Med Virol. 2020;30:e2141. 10.1002/rmv.2141 . Aziz M, Fatima R, Assaly R. Elevated interleukin-6 and severe COVID-19: a meta-analysis. J Med Virol. 2020;92:2283–5. 10.1002/jmv.25948 . Lucas C, Wong P, Klein J, et al. Longitudinal analyses reveal immunological misfiring in severe COVID-19. Nature. 2020;584:463–9. 10.1038/s41586-020-2588-y . Webb BJ, Peltan ID, Jensen P, et al. Clinical criteria for COVID-19-associated hyperinflammatory syndrome: a cohort study. Lancet Rheumatol. 2020;2:e754–63. 10.1016/S2665-9913(20)30343-X . Siewiera J, Brodaczewska K, Jermakow N, et al. Effectiveness of hyperbaric oxygen therapy in SARS-CoV-2 pneumonia: the primary results of a randomised clinical trial. J Clin Med. 2023;12:8. 10.3390/jcm12010008 . Cannellotto M, Duarte M, Keller G, et al. Hyperbaric oxygen as an adjuvant treatment for patients with COVID-19 severe hypoxaemia: a randomised controlled trial. Emerg Med J. 2022;39:88–93. 10.1136/emermed-2021-211253 . Kjellberg A, Douglas J, Pawlik MT, et al. Five sessions of hyperbaric oxygen for critically ill patients with COVID-19-induced ARDS: a randomised, open label, phase II trial. Respir Med. 2024;232:107744. 10.1016/j.rmed.2024.107744 . Boet S, Martin L, Cheng-Boivin O, et al. Efficacy and safety of hyperbaric oxygen treatment in SARS-CoV-2 (COVID-19) pneumonia: a systematic review. Diving Hyperb Med. 2021;51:271–81. 10.28920/dhm51.3.271-281 . Boet S, Martin L, Chen Y, et al. Efficacy and safety of hyperbaric oxygen treatment to treat COVID-19 pneumonia: a living systematic review update. Diving Hyperb Med. 2022;52:126–35. 10.28920/dhm52.2.126-135 . De Maio A, Hightower LE. COVID-19, acute respiratory distress syndrome (ARDS), and hyperbaric oxygen therapy (HBOT): what is the link? Cell Stress Chaperones. 2020;25:717–20. 10.1007/s12192-020-01121-0 . Xu SW, Ilyas I, Weng JP. Endothelial dysfunction in COVID-19: an overview of evidence, biomarkers, mechanisms and potential therapies. Acta Pharmacol Sin. 2023;44:695–709. 10.1038/s41401-022-00998-0 . Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:13–22. 10.1093/biomet/73.1.13 . Pan W. Akaike's information criterion in generalized estimating equations. Biometrics. 2001;57:120–5. 10.1111/j.0006-341X.2001.00120.x . Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. Intensive Care Med. 1996;22:707–10. 10.1007/BF01709751 . WHO Working Group on the Clinical Characterisation and Management of COVID-19 infection. A minimal common outcome measure set for COVID-19 clinical research. Lancet Infect Dis. 2020;20:e192–7. 10.1016/S1473-3099(20)30483-7 . Ospina-Tascón GA, Calderón-Tapia LE, García AF, et al. Effect of high-flow oxygen therapy vs conventional oxygen therapy on invasive mechanical ventilation and clinical recovery in patients with severe COVID-19: a randomized clinical trial. JAMA. 2021;326:2161–71. 10.1001/jama.2021.20714 . Simadibrata DM, Calvin J, Wijaya AD, Ibrahim NAA. Neutrophil-to-lymphocyte ratio on admission to predict the severity and mortality of COVID-19 patients: a meta-analysis. Am J Emerg Med. 2021;42:60–9. 10.1016/j.ajem.2021.01.006 . Chi L, Wang X, Liu J, et al. Predictive value of C-reactive protein for disease severity and survival in COVID-19 patients: a systematic review and meta-analysis. Clin Exp Med. 2023;23:2001–8. 10.1007/s10238-022-00948-4 . Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–83. 10.1016/0021-9681(87)90171-8 . Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36:8–27. 10.1097/00005650-199801000-00004 . Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43:1130–9. 10.1097/01.mlr.0000182534.19832.83 . Hollingworth S, Kairuz T. Measuring medicine use: applying ATC/DDD methodology to real-world data. Pharm (Basel). 2021;9:60. 10.3390/pharmacy9010060 . Additional Declarations No competing interests reported. Supplementary Files Additionalfile1.png Additional file 1 File format: .png Title: CONSORT flow diagram. Description: CONSORT flow diagram of the parent randomized clinical trial, including screening, randomization, follow-up, and the subset included in the present post hoc secondary analysis. Additionalfile2.png Additional file 2 File format: .png Title: Supplementary fitted-score distributions and descriptive summaries. Description: Figure S1 shows observed respiratory SOFA values overlaid on fitted-score distributions across models, fitted score distributions stratified by observed respiratory SOFA category, median fitted score across cHIS values for each model, and counts across observed respiratory SOFA categories. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 22 May, 2026 Reviews received at journal 17 May, 2026 Reviewers agreed at journal 08 May, 2026 Reviewers invited by journal 08 May, 2026 Editor invited by journal 30 Apr, 2026 Editor assigned by journal 13 Apr, 2026 Submission checks completed at journal 13 Apr, 2026 First submitted to journal 11 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9389851","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":641484671,"identity":"5811554b-9114-4257-b0ef-5cf06028199f","order_by":0,"name":"Natalia Jermakow","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA20lEQVRIiWNgGAWjYFCCAyAigYGBvQFIG1iQooUHxDCQINoqoBaJBBCDCC3mjIePfbpRkyZncPP51Q0/CiQY+Nu7E/BqsWw4ljw751iOscHtnLKbPUCHSZw5uwGvFoMDZ4yZc9gqErfdzkm7wQPUYiCRS0jL+c/MOf+AWm6eSbv5hzgtZ5iZc9tyErfdYD92m0hbjhkz5/alGdufyWG7LWMgwUPYLzcOP2bO+ZYsJ9l+/NnNN39s5Pjbe/FrYZA4AGPxGIBJ/MpBgL8BxmJ/QFj1KBgFo2AUjEgAAMmTTWmaDH+FAAAAAElFTkSuQmCC","orcid":"","institution":"Military Institute of Medicine – National Research Institute","correspondingAuthor":true,"prefix":"","firstName":"Natalia","middleName":"","lastName":"Jermakow","suffix":""},{"id":641484672,"identity":"256df41d-a934-400b-a908-8c1368387c91","order_by":1,"name":"Klaudia Brodaczewska","email":"","orcid":"","institution":"Military Institute of Medicine – National Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Klaudia","middleName":"","lastName":"Brodaczewska","suffix":""},{"id":641484673,"identity":"d891f3b1-a890-4fd2-b9ab-0d148ba28c0c","order_by":2,"name":"Jacek Kot","email":"","orcid":"","institution":"Medical University of Gdańsk","correspondingAuthor":false,"prefix":"","firstName":"Jacek","middleName":"","lastName":"Kot","suffix":""},{"id":641484674,"identity":"5e7266e3-3031-431c-a365-13e8f69df4e4","order_by":3,"name":"Arkadiusz Lubas","email":"","orcid":"","institution":"Military Institute of Medicine – National Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Arkadiusz","middleName":"","lastName":"Lubas","suffix":""},{"id":641484675,"identity":"1a646e00-9fb9-4dcb-9913-a6e988291dd2","order_by":4,"name":"Krzysztof Kłos","email":"","orcid":"","institution":"Military Institute of Medicine – National Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Krzysztof","middleName":"","lastName":"Kłos","suffix":""},{"id":641484676,"identity":"39690dc6-898f-4b0b-8172-37c3e43ab54a","order_by":5,"name":"Jacek Siewiera","email":"","orcid":"","institution":"Military Institute of Medicine – National Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Jacek","middleName":"","lastName":"Siewiera","suffix":""}],"badges":[],"createdAt":"2026-04-11 17:53:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9389851/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9389851/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109481736,"identity":"cb2e8c94-0b16-4d43-99a1-5034e8b87832","added_by":"auto","created_at":"2026-05-18 15:18:05","extension":"png","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":136740,"visible":true,"origin":"","legend":"\u003cp\u003eAdditional file 1\u003c/p\u003e\n\u003cp\u003eFile format: .png\u003c/p\u003e\n\u003cp\u003eTitle: CONSORT flow diagram.\u003c/p\u003e\n\u003cp\u003eDescription: CONSORT flow diagram of the parent randomized clinical trial, including screening, randomization, follow-up, and the subset included in the present post hoc secondary analysis.\u003c/p\u003e","description":"","filename":"Additionalfile1.png","url":"https://assets-eu.researchsquare.com/files/rs-9389851/v1/9e1254feb81e2103baa5f46a.png"},{"id":109481737,"identity":"6000d054-435b-4518-b813-61e151d06436","added_by":"auto","created_at":"2026-05-18 15:18:05","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":548466,"visible":true,"origin":"","legend":"\u003cp\u003eAdditional file 2\u003c/p\u003e\n\u003cp\u003eFile format: .png\u003c/p\u003e\n\u003cp\u003eTitle: Supplementary fitted-score distributions and descriptive summaries.\u003c/p\u003e\n\u003cp\u003eDescription: Figure S1 shows observed respiratory SOFA values overlaid on fitted-score distributions across models, fitted score distributions stratified by observed respiratory SOFA category, median fitted score across cHIS values for each model, and counts across observed respiratory SOFA categories.\u003c/p\u003e","description":"","filename":"Additionalfile2.png","url":"https://assets-eu.researchsquare.com/files/rs-9389851/v1/d1f753360ca1dbf7dbbe96ca.png"}],"financialInterests":"No competing interests reported.","formattedTitle":"HBOT, hyperinflammatory burden, comorbidity burden and drug exposure in relation to repeated-measures respiratory severity in hospitalized adults with COVID-19: a post hoc secondary analysis of a randomized clinical trial","fulltext":[{"header":"Background","content":"\u003cp\u003eCoronavirus disease 2019 (COVID-19) is now understood as a systemic illness in which respiratory failure emerges from an interaction between alveolar injury, endothelial dysfunction, vascular abnormalities and dysregulated inflammation rather than from gas-exchange impairment alone [\u003cspan additionalcitationids=\"CR2 CR3\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Patients with apparently similar levels of hypoxemia can follow different clinical trajectories because oxygen requirement, inflammatory activity and supportive care intensity change over time. This heterogeneity is clinically important and also creates a methodological problem for treatment studies that rely on a single oxygenation index as the primary lens through which respiratory course is interpreted.\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThe inflammatory component of COVID-19 has substantial prognostic relevance. Higher interleukin-6 concentrations and related inflammatory abnormalities are associated with severe disease, and longitudinal profiling has shown that adverse trajectories are characterized not only by high inflammatory peaks but also by persistence of maladaptive immune programs [\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Webb and colleagues formalized this concept in the COVID-19-associated hyperinflammatory syndrome (cHIS) score, a pragmatic framework linking hyperferritinaemia, coagulopathy, cytokinaemia and hematologic dysfunction to worsening oxygen requirement, mechanical ventilation and mortality [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. That framework is clinically attractive because it captures a multidimensional inflammatory state rather than any single laboratory abnormality.\u003c/p\u003e\u003cp\u003eHyperbaric oxygen therapy (HBOT) has been explored as an adjunctive treatment in COVID-19 because it increases dissolved oxygen content in plasma and may also influence inflammatory and endothelial responses [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Clinical evidence, however, remains mixed. Our randomized trial demonstrated feasibility and safety, lower normobaric oxygen requirements in the HBOT arm, and favorable changes in selected inflammatory and immune variables, but it did not show a statistically significant between-group difference in serial PaO₂/FiO₂ values [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Other randomized studies and systematic reviews likewise suggest that HBOT is biologically plausible and feasible, while still falling short of establishing a uniform treatment effect across all clinical endpoints [\u003cspan additionalcitationids=\"CR11 CR12\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThat inconsistency makes endpoint choice critical. If treatment influences oxygen support needs or changes the coupling between inflammation and respiratory deterioration, then reliance on raw PaO₂/FiO₂ alone may dilute clinically relevant information. A patient whose oxygenation is maintained with lower respiratory support may be improving even when the PaO₂/FiO₂ ratio itself does not show a dramatic between-group divergence. This issue is especially relevant in COVID-19, where respiratory severity is closely intertwined with supportive strategies such as high-flow nasal oxygen therapy [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan additionalcitationids=\"CR19\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThe present study was designed as a post hoc secondary analysis of that randomized trial. Instead of asking only whether HBOT improved a single oxygenation metric, we examined whether HBOT was associated with the time-varying distribution of a modified respiratory SOFA-type severity score derived from PaO₂/FiO₂ and advanced respiratory support, and whether this relationship varied with hyperinflammatory burden over repeated assessments. Because the original dataset was small, we also summarized comorbidities and medication exposures into clinically named binary axes derived from ICD-10 diagnoses and ATC3 drug classes rather than entering many sparse variables separately [\u003cspan additionalcitationids=\"CR24 CR25\" citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eOur hypothesis was deliberately conservative. We did not assume that HBOT would exert a uniform main effect across the entire cohort. Rather, we asked whether the signal in this dataset was better captured as attenuation of the association between hyperinflammatory burden and respiratory worsening over time. To address that question, we compared four repeated-measures ordinal GEE model families: an HBOT moderation model, a disease model, a drug-exposure model, and a broader global model without interaction terms.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy design and patients\u003c/h2\u003e \u003cp\u003eThis manuscript reports a post hoc secondary analysis of a previously conducted prospective randomized clinical trial comparing standard care alone with standard care plus HBOT in hospitalized adults with COVID-19 pneumonia. The primary clinical outcomes of the parent randomized trial were reported previously [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. The trial was registered in EudraCT (2020-002722-90), approved by the local ethics committee, and conducted after written informed consent.\u003c/p\u003e \u003cp\u003eFrom 1 March 2021 to 3 February 2022, 30 adults hospitalized with RT-PCR-confirmed COVID-19 pneumonia at the Military Institute of Medicine \u0026ndash; National Research Institute in Warsaw were randomly assigned to HBOT plus standard care or standard care alone. Two patients were excluded after randomization because they did not meet eligibility criteria, leaving 28 participants (14 per group) for the present analysis. HBOT consisted of five daily sessions at 2.5 ATA for 75 minutes. Standard of care included anticoagulation and corticosteroids in all patients. No serious HBOT-related adverse events were recorded. Three deaths occurred in the control arm and none in the HBOT arm; these were attributed to COVID-19 progression and occurred after day 10. The participant flow of the parent trial is shown in a CONSORT flow diagram [see Additional file 1].\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eOutcome and inflammatory predictor\u003c/h3\u003e\n\u003cp\u003eRespiratory severity was quantified using a modified respiratory Sequential Organ Failure Assessment score (rSOFA) derived from the Horowitz index (PaO₂/FiO₂) together with advanced respiratory support [\u003cspan additionalcitationids=\"CR19\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. High-flow nasal oxygen therapy and invasive mechanical ventilation were treated as advanced respiratory support because both reflect clinically meaningful escalation of respiratory care in contemporary COVID-19 management [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. As implemented in the analysis code, rSOFA was assigned as 0 when PaO₂/FiO₂ exceeded 400, 1 for values of 301\u0026ndash;400, 2 for values\u0026thinsp;\u0026le;\u0026thinsp;300 without advanced respiratory support, 3 for advanced respiratory support with PaO₂/FiO₂ \u0026le;200, and 4 for advanced respiratory support with PaO₂/FiO₂ \u0026le;100. The resulting ordinal outcome therefore ranged from 0 to 4, with higher categories indicating worse respiratory severity.\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eInflammatory burden was summarized using a reduced cHIS-derived score based on ferritin\u0026thinsp;\u0026ge;\u0026thinsp;700 \u0026micro;g/L, D-dimer\u0026thinsp;\u0026ge;\u0026thinsp;1.5 \u0026micro;g/mL, LDH\u0026thinsp;\u0026ge;\u0026thinsp;400 U/L, NLR\u0026thinsp;\u0026ge;\u0026thinsp;10, and IL-6\u0026thinsp;\u0026ge;\u0026thinsp;15 pg/mL or CRP\u0026thinsp;\u0026ge;\u0026thinsp;15 mg/dL [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. This reduced score preserves the principal inflammatory domains of the original cHIS framework while omitting the fever component, which was not consistently available in the trial dataset [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\n\u003ch3\u003eComorbidity burden and medication axes\u003c/h3\u003e\n\u003cp\u003eComorbidity burden was operationalized in code as a patient-level count of Quan/Elixhauser-style categories derived from ICD-10-coded secondary diagnoses [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. ICD-10 subcodes were normalized and matched by prefix/range rules to Elixhauser-style categories in the analysis script; in this cohort, the mapped burden categories contributing to the final count were hypertension, chronic pulmonary disease, pulmonary-circulation disorders, arrhythmia, and obesity. The burden covariate used in modeling was the count of present categories per patient. Uncomplicated diabetes was not double-counted when complicated diabetes was present, and non-metastatic solid tumor was not double-counted when metastatic cancer was present. ATC3-derived treatment data were collapsed into four binary exposure axes defined directly in code by class prefixes: cardiovascular drugs (B01A, C01, C02, C03, C07, C08, C09, C10), respiratory drugs (R03, R05, R06), immunomodulatory drugs (H02A, L04A), and antibiotics (J01) [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Vaccination data were merged descriptively from the trial vaccine file but were not included in the ATC3-only drug models.\u003c/p\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eFor the present analysis, day 1 was defined as the baseline pre-HBOT assessment, whereas day 10 represented follow-up after completion of five HBOT sessions. Intermediate time points reflected repeated in-hospital assessments during treatment. Missing data were supplemented using the last observation carried forward (LOCF) approach.\u003c/p\u003e \u003cp\u003eData processing and modeling were performed in Python using pandas, NumPy, statsmodels, matplotlib, and seaborn. Population-averaged ordinal generalized estimating equation models were fitted with the statsmodels OrdinalGEE implementation, patient-level clustering, and an independence working-correlation structure [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Four models were compared: an HBOT moderation model including cHIS, HBOT and their interaction; a comorbidity moderation model including Elixhauser burden and a cHIS-by-burden interaction; a drug-axes moderation model including ATC3-derived cardiovascular, respiratory, immunomodulatory, and antibiotic axes together with cHIS-by-drug interactions; and a global model including HBOT, Elixhauser burden, and ATC3-derived drug axes as main effects only. Effects are reported as odds ratios with 95% confidence intervals. Because this manuscript reports a post hoc secondary analysis of an already completed randomized trial, no new sample size calculation was performed for the present analysis; all effect estimates were interpreted as exploratory and hypothesis-generating rather than confirmatory.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eOf the 30 randomized patients, 28 were analyzed (14 HBOT, 14 control). Baseline characteristics were similar between groups (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), including age (52.8\u0026thinsp;\u0026plusmn;\u0026thinsp;13.5 vs. 56.1\u0026thinsp;\u0026plusmn;\u0026thinsp;14.0 years; P\u0026thinsp;=\u0026thinsp;0.52), sex (14.3% vs. 21.4% women; P\u0026thinsp;\u0026gt;\u0026thinsp;0.99), NEWS (median 2 in both groups; P\u0026thinsp;=\u0026thinsp;0.42), SpO2 (96.8\u0026thinsp;\u0026plusmn;\u0026thinsp;2.8% vs. 93.4\u0026thinsp;\u0026plusmn;\u0026thinsp;9.3%; P\u0026thinsp;=\u0026thinsp;0.27), CRP (3.8\u0026thinsp;\u0026plusmn;\u0026thinsp;4.4 vs. 5.4\u0026thinsp;\u0026plusmn;\u0026thinsp;4.3; P\u0026thinsp;=\u0026thinsp;0.27), and PCT (0.09\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07 vs. 0.18\u0026thinsp;\u0026plusmn;\u0026thinsp;0.21; P\u0026thinsp;=\u0026thinsp;0.22). Confirmed COVID-19 vaccination was documented in 2 HBOT-treated patients and 1 control patient (P\u0026thinsp;=\u0026thinsp;1.000). Three deaths occurred in the control group and none in the HBOT group (21.4% vs. 0%; P\u0026thinsp;=\u0026thinsp;0.217), all were attributed to COVID-19 progression or its complications and occurred after day 10.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBaseline characteristics of randomized participants by treatment group. Legend: Values are shown as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD unless otherwise indicated. NEWS is presented as median (IQR).\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eControl\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHBOT\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge, mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e56.1\u0026thinsp;\u0026plusmn;\u0026thinsp;14.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e52.8\u0026thinsp;\u0026plusmn;\u0026thinsp;13.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 (21.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2 (14.3)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeaths, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 (21.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVaccinated, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1 (7.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2 (14.3)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNEWS, median (IQR)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2 (1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2 (0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBaseline SpO₂\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e93.4\u0026thinsp;\u0026plusmn;\u0026thinsp;9.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e96.8\u0026thinsp;\u0026plusmn;\u0026thinsp;2.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBaseline CRP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.4\u0026thinsp;\u0026plusmn;\u0026thinsp;4.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.8\u0026thinsp;\u0026plusmn;\u0026thinsp;4.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBaseline PCT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.18\u0026thinsp;\u0026plusmn;\u0026thinsp;0.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.09\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eFor repeated-measures modeling, 28 patients contributed 168 repeated observations (baseline, day 2\u0026ndash;5, follow-up on day 10). A descriptive summary of the patient-level and observation-level variables entered into the final models is provided in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. Observation-level rSOFA was concentrated in category 2 (140/168, 83.3%), with fewer observations in categories 0, 1, 3, and 4. Observation-level cHIS values spanned the full 0\u0026ndash;5 range, most commonly categories 1\u0026ndash;3. At the patient level, Elixhauser burden was 0 in 14 patients, 1 in 11 patients, and \u0026ge;\u0026thinsp;2 in 3 patients; ATC3-derived cardiovascular, respiratory, immunomodulatory, and antibiotic axes were present in 28, 18, 26, and 21 patients, respectively.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSummary of variables included in the repeated-measures model dataset. rSOFA and cHIS counts are observation-level counts across the 168-row repeated-measures dataset. Treatment allocation, Elixhauser burden, ATC3-derived drug axes, and vaccination are patient-level counts across 28 randomized participants.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDomain\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSamples n (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eTreatment allocation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHBOT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e14 (50.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eControl\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e14 (50.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e \u003cp\u003eObserved rSOFA category\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003erSOFA 0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5 (3.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003erSOFA 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e13 (7.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003erSOFA 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e140 (83.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003erSOFA 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6 (3.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003erSOFA 4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4 (2.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003eObserved cHIS score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecHIS 0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e28 (16.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecHIS 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e37 (22.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecHIS 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e53 (31.5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecHIS 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e28 (16.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecHIS 4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e14 (8.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecHIS 5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e8 (4.8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003ePatient-level comorbidity burden\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eElixhauser count 0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e14 (50.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eElixhauser count 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e11 (39.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eElixhauser count\u0026thinsp;\u0026ge;\u0026thinsp;2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3 (10.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003ePatient-level ATC3 drug axes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCardiovascular drugs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e28 (100.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRespiratory drugs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e18 (64.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eImmunomodulatory drugs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e26 (92.9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAntibiotics\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e21 (75.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDescriptive context\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVaccinated\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3 (10.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe clearest signal was observed in the HBOT moderation model (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea,e). Higher cHIS was associated with higher odds of worse respiratory severity (OR 5.17, 95% CI 2.44\u0026ndash;10.94), whereas the cHIS \u0026times; HBOT interaction was negative (OR 0.20, 95% CI 0.08\u0026ndash;0.49). The HBOT main effect alone was not clearly different from the null (OR 1.89, 95% CI 0.25\u0026ndash;14.14). The negative interaction term indicates a shallower association between hyperinflammatory burden and worsening respiratory severity in the HBOT arm than in the standard-care arm.\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThe comorbidity moderation model (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb,f) replaced the earlier diagnosis-axis structure with a count-based Elixhauser burden derived from ICD-10-coded comorbid diagnoses. In this model, the cHIS main effect remained positive (OR 2.75, 95% CI 1.18\u0026ndash;6.41), whereas the Elixhauser burden main effect (OR 0.77, 95% CI 0.13\u0026ndash;4.49) and the cHIS \u0026times; Elixhauser interaction (OR 0.82, 95% CI 0.34\u0026ndash;1.98) were imprecise and did not clearly depart from the null. These estimates were centered near the null and were less precise than those observed in the HBOT moderation model.\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e summarizes the final model coefficients and the observed cHIS\u0026ndash;rSOFA patterns most relevant to interpretation. Additional model diagnostics and fitted-score distributions are provided in Additional file 2.\u003c/p\u003e\u003cp\u003eThe global main-effects model without moderation (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed) included HBOT, Elixhauser burden, and ATC3-derived drug axes as main effects only. This broader model retained a favorable HBOT-associated coefficient (OR 0.16, 95% CI 0.03\u0026ndash;0.98) together with a positive cHIS coefficient (OR 3.06, 95% CI 1.43\u0026ndash;6.57). However, the cardiovascular-drug coefficient expanded to an implausibly large value, again indicating that the main-effects model should be read as a broad contextual summary rather than a mechanistic explanation.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Main model results and observed distributions. Panels a\u0026ndash;d show forest plots for HBOT moderation, comorbidity moderation, drug-axes moderation, and the global main-effects model without moderation. Points indicate odds ratios and horizontal lines indicate 95% confidence intervals on a logarithmic scale. Panel e shows observed cHIS versus observed respiratory SOFA stratified by HBOT versus control, with jittered points and median trend lines. Panel f shows observed cHIS versus observed respiratory SOFA with point color intensity indicating Elixhauser burden.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe main finding of this post hoc secondary analysis is not that HBOT produced a straightforward, across-the-board increase in PaO₂/FiO₂. That claim would be incompatible with the original randomized report, which did not detect a statistically significant between-group difference in serial PaO₂/FiO₂ values [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Instead, the present analysis suggests something narrower but more clinically coherent: worsening hyperinflammatory burden tracked with worse respiratory severity across repeated assessments, and this association was markedly weaker in patients assigned to HBOT.\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThis distinction matters because the HBOT literature in COVID-19 has often struggled with endpoint interpretation. On physiological grounds, HBOT is attractive in hypoxemic respiratory failure because it transiently increases dissolved oxygen in plasma while potentially influencing vascular and inflammatory pathways [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. But transient tissue hyperoxia, modulation of inflammatory signaling and clinically meaningful respiratory improvement are not interchangeable phenomena. A trial can show lower oxygen requirements without proving a uniform shift in raw PaO₂/FiO₂, and a biomarker change does not by itself establish a treatment effect on respiratory trajectory. The present analysis is therefore best read as a reframing of the parent trial signal rather than as a retrospective rescue of a negative oxygenation endpoint.\u003c/p\u003e\u003cp\u003eThe central role of the reduced cHIS-derived score in our models is biologically plausible. Hyperferritinaemia, elevated D-dimer, LDH release, neutrophil-to-lymphocyte imbalance and cytokinaemia-related markers all map onto features of severe COVID-19 that have repeatedly been associated with progression to respiratory failure and death [\u003cspan additionalcitationids=\"CR6 CR7\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Within that framework, the HBOT moderation model is clinically coherent because it tests whether treatment assignment changes the relationship between inflammatory burden and respiratory trajectory rather than assuming that treatment acts independently of inflammatory context.\u003c/p\u003e\u003cp\u003eThe broader models help define the limits of interpretation. The global main-effects model also favored HBOT, but its treatment coefficient necessarily pooled inflammatory burden, comorbidity burden, and medication context into a single structure. The comorbidity model\u0026mdash;now based on a count-based Elixhauser burden derived from ICD-10 comorbid diagnoses rather than ad hoc diagnosis axes\u0026mdash;did not identify a robust cHIS-by-burden interaction and therefore did not provide a clearer explanation than the HBOT moderation model.\u003c/p\u003e\u003cp\u003eThat point is particularly important for the drug-axes and global context models. Antibiotic and immunomodulatory therapy were not randomly distributed in this cohort, and some exposure patterns were uncommon. Accordingly, the extreme coefficients in the drug-axes model and the implausibly inflated cardiovascular-drug coefficient in the global model should not be read as evidence of strong independent medication-class effects on respiratory course. They are better understood as markers of clinical context in a small inpatient dataset in which part of the observed signal may reflect treatment of severe COVID-19 itself. We therefore retained these analyses because they describe the structure present in the data, but we interpret them cautiously and do not treat them as causal findings.\u003c/p\u003e\u003cp\u003eOur results fit the broader HBOT-in-COVID literature in a nuanced way. Randomized evidence remains limited and heterogeneous, with one trial reporting support for efficacy in severe hypoxemia and another failing to show benefit for major hard outcomes [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Systematic reviews have concluded that HBOT appears feasible and safe but that the evidence base is still insufficient for firm efficacy claims [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. The present study is consistent with that literature if one avoids demanding a simple yes-or-no answer: it supports the possibility that HBOT may matter most where inflammatory burden and respiratory deterioration are tightly coupled, while stopping well short of proving a uniform treatment benefit. Because the present manuscript is a secondary analysis rather than a primary trial report, these findings should be interpreted alongside, not instead of, the parent trial publication and its CONSORT-based reporting [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThe present analysis also has clear limitations. The cohort was small, some exposure patterns were sparse, and the derived rSOFA, reduced cHIS, Elixhauser-burden count, and ATC3 drug axes were code-derived summaries tailored to the available trial dataset rather than external instruments reproduced in full. Missing data were supplemented using LOCF, and the GEE models used an independence working-correlation structure to maintain stable population-averaged inference in a limited sample. These are appropriate exploratory choices for this dataset, but they necessarily constrain the strength of any claim that can be made from the results.\u003c/p\u003e\u003cp\u003eEven with those limitations, the repeated-measures framework adds useful information to the parent trial. It suggests that in this dataset HBOT is most plausibly linked to respiratory course through attenuation of the inflammation\u0026ndash;severity relationship rather than through a simple across-the-board oxygenation effect. Future studies should test this hypothesis prospectively in larger cohorts using respiratory-support trajectories and repeated inflammatory markers as paired repeated-measures endpoints.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eIn this repeated-measures analysis of hospitalized patients with COVID-19, higher hyperinflammatory burden was associated with worse respiratory severity over time, and this association appeared less pronounced in patients assigned to HBOT. This pattern may indicate that the relationship between inflammation and respiratory course differs according to treatment context, but the present study was not designed to establish a definitive treatment effect. The comorbidity- and medication-based models provided additional clinical context, although these results were less stable and should be interpreted cautiously. Because of the small sample size, sparse exposure patterns, exploratory post hoc modeling strategy, and the use of derived summary variables, these findings should be regarded as hypothesis-generating. Larger prospective studies with prespecified respiratory and inflammatory repeated-measures endpoints are needed to determine whether the observed pattern is reproducible and clinically meaningful. They should be interpreted within the limits of a small post hoc secondary analysis of a single randomized trial.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eATC3: Anatomical Therapeutic Chemical Classification System, 3rd level; cHIS: COVID-19-associated hyperinflammatory syndrome; CI: confidence interval; COVID-19: coronavirus disease 2019; EudraCT: European Union Drug Regulating Authorities Clinical Trials Database; GEE: generalized estimating equation; HBOT: hyperbaric oxygen therapy; ICD-10: International Statistical Classification of Diseases and Related Health Problems, 10th Revision; IL-6: interleukin 6; LDH: lactate dehydrogenase; LOCF: last observation carried forward; NLR: neutrophil-to-lymphocyte ratio; OR: odds ratio; PaO₂/FiO₂: arterial oxygen partial pressure to inspired oxygen fraction ratio; rSOFA: modified respiratory Sequential Organ Failure Assessment.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe parent trial was conducted in accordance with the Declaration of Helsinki and approved by the Bioethics Committee of the Military Institute of Medicine \u0026ndash; National Research Institute (approval no. 25/WIM/2020). All participants provided written informed consent in the original trial [9].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets analysed during the current study are not publicly available because they derive from a completed clinical trial dataset containing potentially identifiable clinical information, but are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe parent randomized trial was funded by the Polish Medical Research Agency (grant 2020/ABM/COVID19/0043). No additional external funding was obtained for the present post hoc repeated-measures analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization: J.S., K.B. and J.K.; methodology: J.S., K.B., N.J. and J.K.; formal analysis: N.J., J.S., K.B. and J.K.; investigation: J.S., K.B., A.L. and K.K.; writing\u0026mdash;original draft preparation: N.J., J.S. and J.K.; writing\u0026mdash;review and editing: N.J., J.K, J.S., K.B., A.L. and K.K.; funding acquisition, J.S. and J.K. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors are especially indebted to all employees of the designated COVID-19 clinics of the Military Medical Institute involved in the implementation of this clinical trial.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eWiersinga WJ, Rhodes A, Cheng AC, Peacock SJ, Prescott HC. Pathophysiology, transmission, diagnosis, and treatment of coronavirus disease 2019 (COVID-19): a review. JAMA. 2020;324:782\u0026ndash;93. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jama.2020.12839\u003c/span\u003e\u003cspan address=\"10.1001/jama.2020.12839\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVarga Z, Flammer AJ, Steiger P, et al. Endothelial cell infection and endotheliitis in COVID-19. Lancet. 2020;395:1417\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S0140-6736(20)30937-5\u003c/span\u003e\u003cspan address=\"10.1016/S0140-6736(20)30937-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eARDS Definition Task Force, Ranieri VM, Rubenfeld GD, et al. Acute respiratory distress syndrome: the Berlin definition. JAMA. 2012;307:2526\u0026ndash;33. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jama.2012.5669\u003c/span\u003e\u003cspan address=\"10.1001/jama.2012.5669\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGattinoni L, Chiumello D, Caironi P, et al. COVID-19 pneumonia: different respiratory treatments for different phenotypes? Intensive Care Med. 2020;46:1099\u0026ndash;102. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00134-020-06033-2\u003c/span\u003e\u003cspan address=\"10.1007/s00134-020-06033-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCoomes EA, Haghbayan H. Interleukin-6 in COVID-19: a systematic review and meta-analysis. Rev Med Virol. 2020;30:e2141. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/rmv.2141\u003c/span\u003e\u003cspan address=\"10.1002/rmv.2141\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAziz M, Fatima R, Assaly R. Elevated interleukin-6 and severe COVID-19: a meta-analysis. J Med Virol. 2020;92:2283\u0026ndash;5. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jmv.25948\u003c/span\u003e\u003cspan address=\"10.1002/jmv.25948\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLucas C, Wong P, Klein J, et al. Longitudinal analyses reveal immunological misfiring in severe COVID-19. Nature. 2020;584:463\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41586-020-2588-y\u003c/span\u003e\u003cspan address=\"10.1038/s41586-020-2588-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWebb BJ, Peltan ID, Jensen P, et al. Clinical criteria for COVID-19-associated hyperinflammatory syndrome: a cohort study. Lancet Rheumatol. 2020;2:e754\u0026ndash;63. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S2665-9913(20)30343-X\u003c/span\u003e\u003cspan address=\"10.1016/S2665-9913(20)30343-X\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSiewiera J, Brodaczewska K, Jermakow N, et al. Effectiveness of hyperbaric oxygen therapy in SARS-CoV-2 pneumonia: the primary results of a randomised clinical trial. J Clin Med. 2023;12:8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/jcm12010008\u003c/span\u003e\u003cspan address=\"10.3390/jcm12010008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCannellotto M, Duarte M, Keller G, et al. Hyperbaric oxygen as an adjuvant treatment for patients with COVID-19 severe hypoxaemia: a randomised controlled trial. Emerg Med J. 2022;39:88\u0026ndash;93. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/emermed-2021-211253\u003c/span\u003e\u003cspan address=\"10.1136/emermed-2021-211253\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKjellberg A, Douglas J, Pawlik MT, et al. Five sessions of hyperbaric oxygen for critically ill patients with COVID-19-induced ARDS: a randomised, open label, phase II trial. Respir Med. 2024;232:107744. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.rmed.2024.107744\u003c/span\u003e\u003cspan address=\"10.1016/j.rmed.2024.107744\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoet S, Martin L, Cheng-Boivin O, et al. Efficacy and safety of hyperbaric oxygen treatment in SARS-CoV-2 (COVID-19) pneumonia: a systematic review. Diving Hyperb Med. 2021;51:271\u0026ndash;81. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.28920/dhm51.3.271-281\u003c/span\u003e\u003cspan address=\"10.28920/dhm51.3.271-281\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoet S, Martin L, Chen Y, et al. Efficacy and safety of hyperbaric oxygen treatment to treat COVID-19 pneumonia: a living systematic review update. Diving Hyperb Med. 2022;52:126\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.28920/dhm52.2.126-135\u003c/span\u003e\u003cspan address=\"10.28920/dhm52.2.126-135\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Maio A, Hightower LE. COVID-19, acute respiratory distress syndrome (ARDS), and hyperbaric oxygen therapy (HBOT): what is the link? Cell Stress Chaperones. 2020;25:717\u0026ndash;20. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s12192-020-01121-0\u003c/span\u003e\u003cspan address=\"10.1007/s12192-020-01121-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu SW, Ilyas I, Weng JP. Endothelial dysfunction in COVID-19: an overview of evidence, biomarkers, mechanisms and potential therapies. Acta Pharmacol Sin. 2023;44:695\u0026ndash;709. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41401-022-00998-0\u003c/span\u003e\u003cspan address=\"10.1038/s41401-022-00998-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:13\u0026ndash;22. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/biomet/73.1.13\u003c/span\u003e\u003cspan address=\"10.1093/biomet/73.1.13\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePan W. Akaike's information criterion in generalized estimating equations. Biometrics. 2001;57:120\u0026ndash;5. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.0006-341X.2001.00120.x\u003c/span\u003e\u003cspan address=\"10.1111/j.0006-341X.2001.00120.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. Intensive Care Med. 1996;22:707\u0026ndash;10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/BF01709751\u003c/span\u003e\u003cspan address=\"10.1007/BF01709751\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWHO Working Group on the Clinical Characterisation and Management of COVID-19 infection. A minimal common outcome measure set for COVID-19 clinical research. Lancet Infect Dis. 2020;20:e192\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S1473-3099(20)30483-7\u003c/span\u003e\u003cspan address=\"10.1016/S1473-3099(20)30483-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOspina-Tasc\u0026oacute;n GA, Calder\u0026oacute;n-Tapia LE, Garc\u0026iacute;a AF, et al. Effect of high-flow oxygen therapy vs conventional oxygen therapy on invasive mechanical ventilation and clinical recovery in patients with severe COVID-19: a randomized clinical trial. JAMA. 2021;326:2161\u0026ndash;71. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jama.2021.20714\u003c/span\u003e\u003cspan address=\"10.1001/jama.2021.20714\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSimadibrata DM, Calvin J, Wijaya AD, Ibrahim NAA. Neutrophil-to-lymphocyte ratio on admission to predict the severity and mortality of COVID-19 patients: a meta-analysis. Am J Emerg Med. 2021;42:60\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ajem.2021.01.006\u003c/span\u003e\u003cspan address=\"10.1016/j.ajem.2021.01.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChi L, Wang X, Liu J, et al. Predictive value of C-reactive protein for disease severity and survival in COVID-19 patients: a systematic review and meta-analysis. Clin Exp Med. 2023;23:2001\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10238-022-00948-4\u003c/span\u003e\u003cspan address=\"10.1007/s10238-022-00948-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCharlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373\u0026ndash;83. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/0021-9681(87)90171-8\u003c/span\u003e\u003cspan address=\"10.1016/0021-9681(87)90171-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eElixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36:8\u0026ndash;27. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/00005650-199801000-00004\u003c/span\u003e\u003cspan address=\"10.1097/00005650-199801000-00004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQuan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43:1130\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/01.mlr.0000182534.19832.83\u003c/span\u003e\u003cspan address=\"10.1097/01.mlr.0000182534.19832.83\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHollingworth S, Kairuz T. Measuring medicine use: applying ATC/DDD methodology to real-world data. Pharm (Basel). 2021;9:60. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/pharmacy9010060\u003c/span\u003e\u003cspan address=\"10.3390/pharmacy9010060\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":false,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-infectious-diseases","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"infd","sideBox":"Learn more about [BMC Infectious Diseases](http://bmcinfectdis.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/infd","title":"BMC Infectious Diseases","twitterHandle":"#bmcinfectdis","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"COVID-19, hyperbaric oxygen therapy, respiratory failure, respiratory SOFA, cHIS, comorbidity burden","lastPublishedDoi":"10.21203/rs.3.rs-9389851/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9389851/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eRespiratory failure in COVID-19 reflects both impaired gas exchange and systemic inflammation, and treatment effects may be missed when only single oxygenation metrics are assessed.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe performed a post hoc repeated-measures secondary analysis of 28 randomized adults from a clinical trial of standard care alone or standard care plus hyperbaric oxygen therapy (HBOT). Respiratory severity was represented by a modified respiratory SOFA (rSOFA) score derived from PaO₂/FiO₂ and advanced respiratory support, inflammatory burden by a reduced cHIS-derived score, comorbidity burden by a count of Quan/Elixhauser-style ICD-10 categories, and medication context by ATC3-derived drug axes. Four ordinal generalized estimating equation models were fitted: HBOT moderation, comorbidity moderation, drug-axes moderation, and a global main-effects model.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eIn the HBOT moderation model, higher cHIS was associated with worse respiratory severity (OR 5.17, 95% CI 2.44\u0026ndash;10.94), whereas the cHIS \u0026times; HBOT interaction was negative (OR 0.20, 95% CI 0.08\u0026ndash;0.49). In the comorbidity model, the cHIS effect remained positive (OR 2.75, 95% CI 1.18\u0026ndash;6.41), while the Elixhauser main effect and interaction were imprecise. The ATC3 drug-axes model showed the widest coefficient spread and was interpreted cautiously.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eIn this exploratory repeated-measures reanalysis, the most coherent signal was a weaker association between hyperinflammatory burden and respiratory worsening in the HBOT arm than in the standard-care arm. These findings are hypothesis-generating and should not be interpreted as proof of efficacy.\u003c/p\u003e\u003ch2\u003eTrial registration:\u003c/h2\u003e \u003cp\u003eEudraCT 2020-002722-90, registered 3 May 2020.\u003c/p\u003e","manuscriptTitle":"HBOT, hyperinflammatory burden, comorbidity burden and drug exposure in relation to repeated-measures respiratory severity in hospitalized adults with COVID-19: a post hoc secondary analysis of a randomized clinical trial","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-18 15:18:01","doi":"10.21203/rs.3.rs-9389851/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"146185038960693695964036611593738853447","date":"2026-05-22T11:47:43+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-17T09:44:53+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"115538527681016688299678373807347733785","date":"2026-05-08T06:32:57+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-05-08T05:37:19+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-04-30T09:12:08+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-13T05:50:39+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-13T05:50:26+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Infectious Diseases","date":"2026-04-11T17:37:08+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-infectious-diseases","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"infd","sideBox":"Learn more about [BMC Infectious Diseases](http://bmcinfectdis.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/infd","title":"BMC Infectious Diseases","twitterHandle":"#bmcinfectdis","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b9037480-8b1e-4d8f-a019-0550954be024","owner":[],"postedDate":"May 18th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"146185038960693695964036611593738853447","date":"2026-05-22T11:47:43+00:00","index":54,"fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-17T09:44:53+00:00","index":43,"fulltext":""},{"type":"reviewerAgreed","content":"115538527681016688299678373807347733785","date":"2026-05-08T06:32:57+00:00","index":39,"fulltext":""},{"type":"reviewersInvited","content":"25","date":"2026-05-08T05:37:19+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-18T15:18:01+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-18 15:18:01","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9389851","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9389851","identity":"rs-9389851","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}