JAMA Network Open corrects multiple data display errors in remote monitoring trial after infection and sepsis
JAMA Network Open issued a correction to the Original Investigation “Remote Monitoring Approaches to Reduce Readmissions After Infection and Sepsis: A Randomized Clinical Trial,” originally published June 11, 2026. The correction addresses several presentation errors across the article’s figures, table, and supplement, all in a study centered on postdischarge care after infection and sepsis.
The revised record matters because this was a randomized clinical trial, the kind of study in which small reporting errors can distort how readers interpret baseline risk, eligibility, and the balance of illness severity between groups. In a sepsis population, even a mislabeled age category or comorbidity heading can change the way clinicians judge the fragility of the enrolled cohort. Remote monitoring strategies live or die on these details: who was enrolled, how sick they were at baseline, and how carefully the trial’s patient mix was represented.
One correction involved the exclusion data in Figure 1. Another involved incorrect age category and Charlson Comorbidity Index headings in Figure 3. The percentages for intensive care unit stay were also incorrect in Table 1. In Supplement 2, the final five columns of data in eTable 8 were incorrect, and footnote f had to be consolidated with footnote d in both Table 1 and eTable 8.
Those are presentation errors, not a new clinical result.
Still, presentation errors in a trial of post-sepsis follow-up are not trivial. Sepsis survivors often leave the hospital with persistent physiologic instability, impaired functional reserve, and a high risk of returning for acute care. A remote monitoring program is designed to catch deterioration early, when vital signs, symptoms, or recovery patterns begin drifting before the next readmission. If the table that characterizes the cohort contains wrong intensive care unit percentages, or if the figures misstate exclusion pathways, the reader may misread the severity profile that shaped the intervention’s performance.
The Charlson Comorbidity Index is especially relevant in this setting because it gives a quick measure of chronic disease burden. In a trial of remote monitoring after infection and sepsis, comorbidity mix influences nearly every downstream outcome: readmission risk, outpatient complexity, response to follow-up, and the threshold for returning to the emergency department. An incorrect heading can therefore mislead readers who are trying to understand whether the intervention was tested in a relatively stable population or in patients carrying a dense burden of chronic illness.
The same is true for age categories. Older patients with sepsis often have different recovery trajectories than younger adults, with slower mobilization, more medication complexity, and a narrower physiologic margin after discharge. If age grouping is mislabeled, even if the underlying dataset is unchanged, the clinical portrait of the trial becomes harder to trust at a glance. In a field where postdischarge monitoring is increasingly linked to digital tools, nurse outreach, and early warning systems, accuracy in cohort description is part of the intervention’s credibility.
Table 1 and supplementary tables usually do more than list numbers. They frame the baseline architecture of the trial. ICU stay percentages tell readers how many patients required the highest level of acute care before discharge, which in turn hints at the severity of the index hospitalization and the expected burden of postacute complications. Incorrect percentages in that context may not alter the trial’s conclusion, but they do affect how confidently clinicians can compare the enrolled cohort with their own patients.
The correction also reached Supplement 2, where the final five columns of eTable 8 were corrected and footnote f was merged with footnote d. Supplementary materials often carry the granular details that support secondary analyses, subgroup exploration, or more technical interpretation of the study dataset. When those columns are wrong, even if only in the supplement, researchers and clinicians lose a clean line of sight into the data structure that underpins the published paper.
For bedside clinicians, the practical meaning of this correction is straightforward: the article’s presentation has been repaired, and the trial should now be read using the corrected figures and tables. For researchers, the larger lesson lies in the fragility of data display in complex randomized studies. Sepsis is not a simple exposure, and remote monitoring after hospitalization is not a simple intervention. Both depend on accurate baseline characterization, because the usefulness of any readmission-reduction strategy depends on whether it was tested in a cohort with the severity and comorbidity pattern that the tables claim to represent.
Corrections like this do not change the fact that the study asked a highly relevant question: whether remote monitoring approaches can reduce readmissions after infection and sepsis. They do, however, reinforce the need to read the corrected version rather than rely on the original tables, figures, or supplement. In a high-stakes area of care where recurrence, decompensation, and rehospitalization remain common, precision in reporting is part of the evidence itself.