A Randomized, Controlled Trial of Methylprednisolone or Naloxone in the Treatment of Acute Spinal-Cord Injury (NASCIS II)
Editorial decision support · Structured manuscript triage and peer-review support. Editors and editorial teams screening manuscripts.
Structured editorial decision support. Not an editorial decision, and not a replacement for peer review.
Two engines read this manuscript and named different routes — first engine: Statistical or methods review recommended before external review; second engine: Early editorial assessment recommended before reviewer assignment. The route above is the one delivered after reconciliation and, where it applies, the numeric layer's protection.
Audience fit & interest
A multicenter double-blind RCT of pharmacologic treatment for acute spinal-cord injury is squarely within scope for an original-research general-medicine section.
Study validity & design
The randomized blinded multicenter design is strong, but inferential validity of the core early-treatment claim depends on prespecified subgroup, multiplicity, and center-stratified analyses not visible in the supplied text.
Conclusions vs. evidence
The overall motor comparison was null and the efficacy conclusion is drawn from the ≤8-hour subgroup, so a methods reviewer should confirm this subgroup was pre-specified as primary before the conclusion stands.
Statistical reporting & rigor
Multiple endpoints, timepoints and subgroups are tested with no stated multiplicity correction, and odds ratios/differences with confidence intervals crossing the null are narrated as consistent benefit.
Guideline & checklist compliancenot easily fixable
Current submission-package criteria are not met in the supplied materials because clinical-trial registration, protocol/SAP, data-sharing, and several disclosure statements are not located, despite IRB and funding statements.
Clarity & communication
Not assessed: no grounded writing or communication signal in this read. The dimension is reported only when one is.
The manuscript builds on NASCIS 1 and animal-dose rationale by testing higher-dose methylprednisolone with placebo and naloxone arms in a multicenter RCT; later landmark status is not independently assessed from the supplied context.
Journal and manuscript metadata
- Manuscript
- NASCIS-II-1990
- Article type
- Original research
- Journal section
- Original Article
- Prompt version
- editorial-evaluation-prompt-2026-08-29.1
Topic and audience fitClear
A randomized, double-blind, placebo-controlled multicenter trial of methylprednisolone and naloxone for acute spinal-cord injury matches the original_research article type and a general-medicine Original Article section.
- The primary end point was a change in neurologic function between base line and the follow-up examination.
- We evaluated the efficacy and safety of methylprednisolone and naloxone in a multicenter randomized, double-blind, placebo-controlled trial in patients with acute spinal-cord injury
Design, data, and conclusions fitConcern
The overall randomized comparison showed no significant motor benefit; the efficacy conclusion is derived from patients treated within eight hours, a stratified subgroup, which the reader must accept as pre-specified rather than data-driven.
- No comparable improvements in motor function were observed.
- The beneficial effect of methylprednisolone was limited to the patients treated within eight hours of their injury, supporting the hypothesis that
- Since two a priori hypotheses were that any effects of treatment would be influenced by how quickly the drug was given and by the severity of injury
- The primary end point was a change in neurologic function between base line and the follow-up examination.
- We conclude that in patients with acute spinal-cord injury, treatment with methylprednisolone in the dose used in this study improves neurologic recovery when the medication is given in the first eight hours.
Potential for improvement with current dataConcern
With the existing randomized trial data, authors could provide protocol/SAP documentation, multiplicity-adjusted or hierarchical analyses, center-stratified sensitivity models, confidence intervals, n/N denominators, and tighter functional interpretation. These revisions would not require a new study but could materially affect the strength and framing of the early methylprednisolone claim.
- We summarized the results using an analysis of variance for the effects of the protocol, the time the dose was received
- As a check to ensure that the conclusions were not influenced by the
- The analyses were then repeated for the patients who received the drug within the protocol's time limits.
- the improvements in neurologic function attributable to methylprednisolone seen in the present study cannot readily be translated into specific improvements in functional status.
Statistical and methodological alignmentConcern
The section written for the reader a clinical reviewer is not: threats to the inference that survive clinical peer review — unadjusted confounding and its direction, multiplicity without correction, immortal time, an odds ratio narrated as a risk, a stateable missing-data mechanism, unmodelled clustering, competing risks treated as censoring, a precision claim the sample cannot support, ITT and per-protocol conflated. Where one applies it is named here; the instance it rests on, and the reviewer competence that would settle it, are below.
Three neurologic measures are tested at two timepoints across multiple strata (≤8/>8 hours, complete/incomplete, per-protocol) with no stated correction for multiplicity; the primary overall motor comparison is null while significance emerges within subgroups, the classic pattern peer review passes over.
- Analysis of variance was used to test the hypothesis that the change in score was not different across the three treatment groups.
- the odds of improving by a full category after six weeks were always higher in the
- For motor function, the odds ratio, as adjusted for initial level of injury, was 2.04 (95 percent confidence interval, 0.81 to 5.12); for the sensation of pinprick, the odds ratio was 2.93
- The primary end point was a change in neurologic function between base line and the follow-up examination.
- We summarized the results using an analysis of variance for the effects of the protocol, the time the dose was received (≤8 or >8 hours from injury), and the degree of neurologic loss (complete or incomplete).
- Within each center the three treatments were randomized in blocks of nine.
- The study’s blocked design was disturbed at three centers because one patient in each block was given an incorrectly assigned drug.
Statistical/methods follow-up needConcern
A statistical reviewer should confirm whether the ≤8-hour stratum and the per-protocol reanalysis were pre-specified primary analyses, assess the multiplicity across endpoints/timepoints/subgroups, and evaluate whether small subgroup cells (n as low as 5) support the categorical claims.
- the analysis was also stratified on the basis of time to loading dose (≤8 vs.
- The analyses were then repeated for the patients who received
- No. of patients 5 12 6 5 11 6
- Analysis of variance was used to test the hypothesis that the change in score was not different across the three treatment groups.
- In addition to the expanded neurologic score, the five categories of injury were analyzed according to whether the patient's condition improved, remained the same, or regressed.
What was read
This evaluation read one manuscript document, which is what an editorial evaluation takes — 39,375 characters of extracted text.
The manuscript refers to Table 1, Table 2, Table 3, Table 4 and Table 5. The deterministic layer transcribed 55 structured values from the text and recomputed 33 checks.
Where a table is an image rather than selectable text, its cells cannot be transcribed and the checks that would depend on them do not run. This report does not distinguish a table with nothing to flag from a table it could not read — so a short list of checks is not, on its own, evidence that the numbers are sound.
Recomputed checks
Statistics recomputed from the manuscript's own printed numbers — arithmetic only, run independently of both appraisers. A consistent check means the printed value is reproducible from the values beside it; it does not assess the underlying data. The arithmetic on any one number is exact and repeats; which numbers this layer could pull out of the manuscript does not, so the count below is what this run reached rather than a fixed property of the paper. 33 checks recomputed · 31 consistent · 2 flagged.
Mean vs SD skew screenpresentation (mild): 4 of 18 mean±SD summaries of non-negative quantities have mean − 2·SD < 0, so those distributions are likely skewed (Mean expanded motor score, Mean expanded motor score, Mean expanded motor score, Time accident to admission (hr)) — median (IQR) would represent these data better
- screened
- 18
- flagged
- 4
- quantities
- Mean expanded motor score, Mean expanded motor score, Mean expanded motor score, Time accident to admission (hr)
- rows
- (1) quantity: Mean expanded motor score · mean: 23.7 · sd: 17.4 · spread label: sd · n: 161 · implied from sem: no; (2) quantity: Mean expanded motor score · mean: 24.9 · sd: 18.2 · spread label: sd · n: 153 · implied from sem: no; (3) quantity: Mean expanded motor score · mean: 24 · sd: 19.6 · spread label: sd · n: 170 · implied from sem: no; (4) quantity: Time accident to admission (hr) · mean: 3.1 · sd: 2.6 · spread label: sd · n: — · implied from sem: no
Inconsistent — flagged in findings
Percentages reported without denominatorspresentation (mild): 10 percentage(s) are reported without a stated denominator (e.g. "95 percent of whom were treated within 14 hours of injury" — Abstract) — report numerator/denominator (n/N) for each percentage
- count
- 10
- examples
- Abstract: 95% — “95 percent of whom were treated within 14 hours of injury” · Results: 80% — “80 percent of the patients received their study drug within the protocol's time limits” · Results: 92.1% — “92.1 percent received the drug according to the protocol's dose schedule” · Results: 60% — “About 60 percent of the patients had complete injuries” · Results: 92% — “lines were placed in 92 percent of the patients” · Results: 40% — “Bone fragments were seen in over 40 percent” · Discussion: 6% — “the overall mortality rate of 6 percent in this study” · Discussion: 97.9% — “97.9 percent underwent a neurologic examination after six weeks”
Inconsistent — flagged in findings
Consistent, nothing to report: Subgroup counts vs analytic N; Impossible p-values; Threshold-only p-value reporting; Over-precise p-values; Unlabeled dispersion (SD vs SEM); Cross-location value consistency (7); Effect estimate vs its confidence interval (6); Category percentages sum to 100 (12); Participant flow arithmetic.
Non-obvious weaknessesConcern
In the plegic-with-partial-sensory-loss stratum the six-month methylprednisolone change scores are at or below placebo (motor 23.0 vs 26.5; touch 0.0 vs 5.2), running counter to a uniform benefit narrative and resting on very small cells (n=5).
- 23.0 (0.652) 28.9 (0.711) 26.5 (R)
- 0.0 (0.479) 13.5 (0.181) 5.2 (R)
Two of three category-improvement odds ratios (motor 0.81–5.12; touch 0.72–3.94) cross 1 and the motor proportion difference CI (−2.9 to 25.5) crosses 0, yet the text characterizes improvement as consistently higher; interpretation should track the intervals.
- the odds of improving by a full category after six weeks were always higher in the
- a larger proportion had improved (as opposed to stable or worsening) motor function
The manufacturer of the favorable drug (Upjohn/methylprednisolone) supplied the study drug and placebo; a reviewer should note this alignment even though it is disclosed, as no formal COI statement contextualizes the funder's role.
- The study drugs and placebos were provided by the Upjohn Corporation (methylprednisolone) and the Dupont Corporation (naloxone).
Several non-negative quantities are summarized as mean±SD where mean−2·SD is negative (e.g., expanded motor score 23.7±17.4; time to admission), indicating skew for which median (IQR) would represent the data better.
- Mean expanded motor score 23.7±17.4
Multiple percentages are reported without explicit numerator/denominator (e.g., the 95% treated within 14 hours in the abstract), reducing verifiability of the reported proportions.
- 95 percent of whom were treated within 14 hours of injury
- For example, in 33 percent of the patients given methylprednisolone, pinprick sensation improved by at least one category after six weeks, as compared with 17 percent of the placebo group.
The principal favorable claim is supported by multiple neurologic outcomes, two follow-up times, and time-to-dose/severity strata, with several P values near 0.05 and no visible correction or hierarchical testing rule. A trial statistician should verify prespecification and multiplicity handling against the protocol/SAP.
- The primary end point was a change in neurologic function between base line and the follow-up examination.
- Among the patients treated within eight hours of their injury, those receiving methylprednisolone recovered more motor function than those given placebo (16.0 vs. 11.2; P = 0.033), and they also had greater sensory function (pinprick, 11.4 vs. 6.6; P = 0.016; and touch, 8.9 vs. 4.3; P = 0.030).
Randomization was blocked within center and the block design was disturbed at some centers, but the described ANOVA summary does not state whether center, block, or center-by-treatment variation was modeled. A trial statistician should check whether center-adjusted or sensitivity analyses are needed.
- Within each center the three treatments were randomized in blocks of nine.
- The study’s blocked design was disturbed at three centers because one patient in each block was given an incorrectly assigned drug.
- We summarized the results using an analysis of variance for the effects of the protocol, the time the dose was received (≤8 or >8 hours from injury), and the degree of neurologic loss (complete or incomplete).
Journal-specific publishing criteriaFlag
NEJM instructions require trial registration for original_research clinical trials; no registration identifier is present (the manuscript predates prospective registration mandates).
The supplied guidelines require protocol plus statistical analysis plan at submission for clinical trials; neither is provided, and it is central to resolving the subgroup pre-specification question.
A data-sharing statement is required for clinical trials under the supplied instructions and is not present.
A formal conflict-of-interest statement is required; the manuscript discloses commercial provision of drugs but no structured COI declaration, and the funder-drug alignment (Upjohn/methylprednisolone) is the favorable arm.
- The study drugs and placebos were provided by the Upjohn Corporation (methylprednisolone) and the Dupont Corporation (naloxone).
An author-contributions statement is required for original_research; personnel are listed by role but no contribution statement is provided.
The supplied instructions require a structured abstract with Background/Methods/Results/Conclusions under 250 words; the abstract as presented is not shown with those section headers.
- Studies in animals indicate that methylprednisolone and naloxone are both potentially beneficial in acute spinal-cord
The supplied instructions require a structured abstract with Background, Methods, Results, and Conclusions headings. The manuscript has an abstract, but the supplied text does not show those required headings; this is a formatting compliance issue.
- Abstract
The manuscript is a clinical trial and reports IRB protocol approval, but the supplied manuscript/package does not show trial registration, protocol file, statistical analysis plan, or data-sharing statement required by the supplied current original-research instructions. Under those criteria, this is a package-level gate before routine reviewer assignment.
- We evaluated the efficacy and safety of methylprednisolone and naloxone in a multicenter randomized, double-blind, placebo-controlled trial in patients with acute spinal-cord injury
- Institutional review boards at each center approved the study protocol.
Ethics and funding statements are present, but author-contribution, AI-disclosure, explicit COI, trial-registration, and data-sharing statements were not located in the supplied text. The office should request the missing statements rather than assume absence or compliance.
- Institutional review boards at each center approved the study protocol.
- Supported by a grant (NS-15078) from the National Institute of Neurological Disorders and Stroke. The study drugs and placebos were provided by the Upjohn Corporation (methylprednisolone) and the Dupont Corporation (naloxone).
Five tables are visible, apparently within the supplied max_tables_figures limit, but body word count and complete reference style/order cannot be verified from the OCR/transcribed text provided for this screen.
- NASCIS 2 — Tables 1 to 5
Author-instruction and reporting-guideline complianceConcern
Core RCT features are reported, including randomization, blinding, and outcome definitions, but the supplied text lacks a CONSORT-style participant flow, sample-size rationale, trial registration/protocol accessibility, and consistently denominatored percentages. A CONSORT checklist and missing reporting elements would materially improve reviewability.
- Within each center the three treatments were randomized in blocks of nine.
- Of the surviving patients, 97.9 percent underwent a neurologic examination after six weeks, and 96.5 percent after six months.
- All phases of the study (preparation and administration of the drugs, neurologic examinations, and statistical analyses) were carried out in a blinded fashion.
- The primary end point was a change in neurologic function between base line and the follow-up examination.
Place in the literatureMinor
The trial follows NASCIS 1 with a higher methylprednisolone dose and an added naloxone and placebo arm, and became a widely cited basis for early high-dose steroid use; the definitive efficacy claim nonetheless rests on the early-treatment subgroup.
- The present trial, NASCIS 2, was undertaken to study this higher dose of methylprednisolone. A placebo arm was added
- In an earlier trial (the National Acute Spinal Cord Injury Study, or NASCIS 1) we compared a 1000-mg infusion of methylprednisolone sodium succinate with a 100-mg dose of methylprednisolone given as a bolus and daily thereafter for 10 days.
Strengths worth preserving
- Randomized, double-blind, placebo-controlled multicenter design with blinding maintained across drug preparation, examination, and analysis.
- Near-complete follow-up (97.9% at six weeks, 96.5% at six months) with mortality data on all patients, limiting attrition bias.
- Baseline characteristics well balanced across the three arms with no meaningful differences.
- Sensitivity checks reported (bootstrap standard error comparison; left- vs right-side concordance).
- Transparent safety reporting of wound infection and gastrointestinal bleeding with p-values across arms.
- Multicenter randomized, double-blind, placebo-controlled design with separate active/placebo systems to maintain blinding.
- Defined eligibility criteria and systematic neurologic outcome assessments at baseline, six weeks, and six months.
- High follow-up among surviving patients and complete mortality ascertainment are reported.
- The manuscript distinguishes all-randomized analyses from protocol-compliant analyses rather than relying only on per-protocol results.
- The discussion explicitly cautions that neurologic score improvements cannot be directly translated into specific functional outcomes.
Suggested reviewer or statistical-review focus
- Statistician to determine whether the ≤8-hour stratum and per-protocol reanalysis were pre-specified primary analyses versus post-hoc, and to assess multiplicity across three measures, two timepoints, and multiple strata.
- Statistician to evaluate whether the small subgroup cells (n as low as 5) support the categorical/odds-ratio conclusions and to reconcile CIs that cross the null with the narrative.
- Clinical reviewer to judge whether the neurologic change-score improvements are clinically meaningful given the authors' own caution that they cannot readily be translated into functional status.
- Reviewer to assess the funder-arm alignment (Upjohn provided methylprednisolone) against the strength of the efficacy claims.
- Trial statistician: verify protocol/SAP prespecification of the primary endpoint, six-month timepoint, three neurologic measures, eight-hour cutoff, severity strata, and multiplicity plan.
- Trial statistician: request or reproduce center-adjusted and sensitivity analyses accounting for center blocking and the reported block disturbances.
- Statistical reviewer: require confidence intervals and n/N denominators for key percentages, categorical improvements, and harms.
- CONSORT-oriented reviewer/editorial staff: check participant flow, sample-size rationale, registration/protocol/SAP availability, data-sharing statement, and checklist completion.
- Clinical SCI reviewer: assess whether neurologic-score changes and one-category improvements are framed appropriately relative to unmeasured functional outcomes.
- Editorial office: clarify Upjohn/Dupont role and obtain author COI disclosures given topically relevant commercial provision of study drugs and placebos.
Highest-yield author queries
- Was the ≤8-hour subgroup the pre-specified primary analysis, and if so, why is the overall (all-patients) motor comparison not the primary result? Please provide the protocol and statistical analysis plan.
- How was multiplicity across the three neurologic measures, two follow-up timepoints, and multiple strata handled, and were any p-values adjusted?
- Please clarify the role of Upjohn and Dupont beyond drug provision and provide a formal conflict-of-interest statement.
- Provide numerator/denominator for the reported percentages and consider median (IQR) for skewed non-negative measures.
- Given the reversed direction and tiny cell sizes in the plegic-with-partial-sensory-loss stratum, how robust are the category-level conclusions?
- Provide the protocol and statistical analysis plan showing the prespecified primary endpoint, timepoint, outcome hierarchy, eight-hour subgroup, severity strata, and multiplicity strategy.
- Provide the trial registration record, or a dated explanation if no prospective registration exists, plus any protocol/SAP submission files required by the journal.
- Provide reanalyses or sensitivity analyses that account for center/blocking and the reported block disturbances, with confidence intervals for key treatment effects.
- Add n/N denominators for percentages in the abstract, outcomes, follow-up, protocol adherence, and adverse-event reporting.
- Provide a CONSORT checklist, participant flow information, sample-size/power rationale, and data-sharing statement.
- Provide author-contribution, COI, funding-role, and AI-disclosure statements; specifically clarify the roles of Upjohn and Dupont beyond supplying drugs/placebos.
- Revise conclusion language to keep functional implications clearly bounded to measured neurologic outcomes unless functional outcome data are supplied.
Limitations of this evaluation
- Tables were transcribed from an image-only PDF, so exact numeric fidelity and footnote alignment could not be independently confirmed.
- No protocol or statistical analysis plan was supplied, so pre-specification of the 8-hour stratum and endpoint hierarchy cannot be verified.
- Body word count against the 2700-word limit could not be reliably measured from the extracted text.
- The private guidelines reflect current NEJM instructions applied to a 1990 manuscript; several requirements (registration, data-sharing, AI disclosure) postdate the study, so compliance flags are contextual rather than actionable defects.
- No separate disclosure or funding forms were provided beyond the in-text acknowledgment.
- The supplied text appears to be OCR/transcribed from an image-only PDF with extraction artifacts, so exact word count, layout compliance, and complete reference order/style could not be verified.
- Protocol, statistical analysis plan, trial-registration record, CONSORT checklist, data-sharing statement, disclosure forms, and supplementary files were not supplied.
- This evaluation did not have individual participant data and could not reproduce ANOVA, log-linear, bootstrap, survival, center-adjusted, or multiplicity-adjusted analyses.
- Independent later literature context was not supplied, so contribution placement beyond the manuscript’s own comparison with NASCIS 1 and animal evidence was not assessed.