Effect of Enhanced Physical Rehabilitation on Functional Outcomes After Traumatic Injury: A Narrative Review

Round 1

Reviewer 1 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

Dear Authors,

Thank you for the revised submission. While I appreciate the additional clarifications and improvements to structure and reporting, the central issue remains unresolved. The review ultimately includes only three RCTs, all of which have been synthesized in earlier reviews and guideline documents, and no new studies were identified through the updated search. As a result, the manuscript offers only an incremental update rather than a meaningful advancement of current knowledge. Because the evidence base is too limited to support new conclusions or clinical implications, the manuscript does not represent a sufficiently substantial contribution to warrant publication at this time.

Author Response

Comment 1:
While I appreciate the additional clarifications and improvements to structure and reporting, the central issue remains unresolved. The review ultimately includes only three RCTs, all of which have been synthesized in earlier reviews and guideline documents, and no new studies were identified through the updated search. As a result, the manuscript offers only an incremental update rather than a meaningful advancement of current knowledge. Because the evidence base is too limited to support new conclusions or clinical implications, the manuscript does not represent a sufficiently substantial contribution to warrant publication at this time.

Response:
Thank you for this feedback and highlighting the limitations of our original submission. In response, we have converted our systematic review into a narrative review to focus on the synthesis of existing evidence and providing context and interpretation without implying novel conclusions or clinical implications. We believe that this addresses the concerns raised whilst more accurately reflecting the current evidence.

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

This review asks a simple question - if ward base physiotherapy and OT are engaged early after traumatic injury do the patients do better.  After a solid PRISMA search, only 3 RCTs with 388 patients fit the bill.  That's the core problem straight away - the evidence base is tiny.

Across the three trials the rehab packages had little consistency (different timing, frequency and outcomes) and the intervention was inadequately detailed so it's hard to replicate what was done.  Because the function measures and timepoints didn’t line up, the authors could only pool length of stay and discharge destination — and there was no clear difference there.  As results were mixed and sample size small there is a significant risk of bias here.  As such we cannot answer the question is early/more better.  Suggestions for standardisation and further studies are required to be better examined in the conclusions before worthy of publication.

Author Response

Comment 1:
This review asks a simple question - if ward base physiotherapy and OT are engaged early after traumatic injury do the patients do better. After a solid PRISMA search, only 3 RCTs with 388 patients fit the bill. That's the core problem straight away - the evidence base is tiny.
Response:
Thank you for emphasising this important point. We agree that there is limited evidence with only three RCTs yielded. Given this constraint, we have converted the manuscript from a systematic review to a narrative review. The review synthesises the findings and contextualises the existing evidence rather than attempting to draw new clinical conclusions. This approach facilitates meaningful discussion of current knowledge, limitations and future directions.

Comment 2:
Across the three trials the rehab packages had little consistency (different timing, frequency and outcomes) and the intervention was inadequately detailed so it's hard to replicate what was done. Because the function measures and timepoints didn’t line up, the authors could only pool length of stay and discharge destination — and there was no clear difference there. As results were mixed and sample size small there is a significant risk of bias here. As such we cannot answer the question is early/more better. Suggestions for standardisation and further studies are required to be better examined in the conclusions before worthy of publication.
Response:
We appreciate your detailed assessment. We agree that there was considerable variation in intervention timing, frequency and outcomes across the studies which limits reproducibility. We also acknowledge that there are small sample sizes and risk of bias which limit conclusions on enhanced rehabilitation. Given this review is now presented as a narrative synthesis, pooling of length of stay and discharge destination data is no longer appropriate. Furthermore, we have revised the conclusions to emphasise the need for standardised intervention protocols, consistent outcome measures, and larger high quality studies to better address this question as suggested.

Manuscript update (Line 292-296): “Improved reporting of interventions using tools such as the Template for Intervention Description and Replication (TIDieR) checklist or the Consensus on Exercise Reporting Template (CERT) is recommended to facilitate reproducibility and implementation [38-40]. At a minimum, studies should clearly report intervention timing, dosage, and intensity to facilitate replication and implementation.”

Manuscript update (Line 342-348): “However, this heterogeneity reflects the true spectrum of presentations in clinical practice, which pragmatic designs can better address in future research. For example, a stepped-wedge cluster randomised trials implements tailored interventions sequentially across multiple centres, while maintaining methodological quality and preserving real-world applicability [44]. This enables comparison within and across centres and may assist in clarifying the relative contributions of specific intervention components and identifying which elements result in meaningful change across the various subgroups.”

Manuscript update (Line 369-373): “The uncertainty of optimal timing, intensity, and dosage impacts standardisation, adherence, and implementation of enhanced rehabilitation models of care. Future research should use standardised reporting frameworks to ensure that intervention details are clearly described, providing guidance to clinicians and trauma centres, including adequate allocation of resources such as staffing to deliver interventions.”

Manuscript update (Line 376-383): “Variation in outcome measures further limits comparability across studies and may not capture meaningful changes in function. Implementing patient-centred outcomes, such as the Patient-Reported Outcomes Measurement Information System – Physical Function (PROMIS-PF), may help address domains that are meaningful to the individual. PROMIS-PF is a validated measure in orthopaedic trauma populations and is a standardised approach to evaluating physical function [45-47]. Larger studies in the future should also integrate adherence monitoring and economic evaluations to inform the implementation of sustainable enhanced rehabilitation models in acute trauma units.”

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

Please also make sure to carefully review the attached ‘Comments for Authors’ document, as it contains additional detailed points and requested clarifications.

The topic is clinically important and the review is generally well-structured; however, several issues must be addressed to ensure reproducibility and transparency. Please (1) correct and confirm the exact executed CINAHL search strategy (the current Supplement contains apparent syntax/step-referencing errors), and re-run/update the search/PRISMA flow/analyses if needed; (2) reconcile the PRISMA flow and narrative description of study identification/selection, explicitly describing how studies from prior systematic reviews were processed; (3) provide complete arm-level numbers used in each meta-analysis, particularly where results depend on author-provided data (e.g., discharge destination); and (4) clarify eligibility criteria regarding severe TBI vs included TBI populations (define “severe” operationally and ensure consistency across Methods/Results/Tables). In addition, please align any statements about safety with the extent of harms reporting (“not reported” vs “no events observed”) and consider tempering directional language for non-significant findings.

Comments for author File: Comments.pdf

Author Response

Comment 1:
The topic is clinically important and the review is generally well-structured; however, several issues must be addressed to ensure reproducibility and transparency.
The CINAHL search strategy in the Supplement appears to contain syntax/typographical problems (e.g., unclosed quotation marks and/or ambiguous step referencing such as “S44 NOT 45” rather than “S44 NOT S45”). Requested action: Please (a) correct the CINAHL search syntax exactly as executed, (b) confirm the final run steps and hit counts per step, and (c) if the published supplement differs from what was executed, re-run the corrected search and update the PRISMA flow, included studies (if changed), and analyses accordingly.
Response:
Thank you for these valuable comments. Accordingly, the full search strategies have been updated to include the final run steps and corresponding hit counts in the supplementary materials to ensure full transparency and reproducibility. Furthermore, the CINAHL search syntax has been reviewed carefully to correct the typographical and formatting errors identified including “S44 NOT S45”. These identified issues were limited to reporting inconsistencies in the supplemental material and did not reflect errors in the executed database search. As such, the PRISMA flow and included studies remain unchanged. Although the manuscript was transitioned to a narrative review, the detailed search strategy has been retained and repositioned to supplementary materials to ensure methodological clarity.
Manuscript update (Line 85-86): “Additional details about the methods of the evidence synthesis are provided in the Supplementary Material (S1).”

Comment 2:
Reconcile the PRISMA flow and narrative description of study identification/selection, explicitly describing how studies from prior systematic reviews were processed PRISMA flow and study identification/selection clarity (including prior SRs)
The manuscript’s narrative of additional records identified via citation searching/other methods is not fully reconciled with the PRISMA flow diagram and the described handling of studies derived from prior systematic reviews (e.g., prior SRs reportedly contained a larger set of included studies, but only a subset appears to proceed to full-text screening). Requested action: Please harmonize the PRISMA diagram and text (define clearly what is counted as “records identified,” “records screened,” “full-text assessed,” etc.) and explicitly describe how studies from prior SRs were processed (deduplication approach and reasons why only certain studies progressed to full-text assessment).
Response:
Thank you for this actionable feedback. While the manuscript is now a narrative review, we have retained and updated the PRISMA flow diagram (See Supplementary Material Figure 1) to show records identified, title/abstract screening, and full-text screening. The review describes the overall study identification and selection process, including the use of prior systematic reviews to identify relevant studies in the Methods section. Detailed step by step systematic screening and deduplication are no longer provided, however the PRISMA flow diagram alongside the narrative and Supplementary Methods ensure clarity and rigor.
Manuscript update (Line 85-86): “Additional details about the methods of the evidence synthesis are provided in the Supplementary Material (S1).”
Comment 3:
Provide complete arm-level numbers used in each meta-analysis, particularly where results depend on author-provided data (e.g., discharge destination). For at least one trial, discharge destination (e.g., home discharge) appears to rely on author-provided data, but the intervention/control cell counts are not fully presented in the supplementary extraction tables (only partial discharge categories are shown). Requested action: Please report the exact group-level numbers used in each meta-analysis (per study and per arm). If obtained from authors, label explicitly as “author-provided data,” include the correspondence date (optional), and provide the exact counts used for the RR calculations.
Response:
We appreciate the reviewer for highlighting this issue. As the manuscript has been converted to a narrative review, meta-analyses or pooled estimates are no longer included in the manuscript. The arm-level numbers with author-provided data to discharge home have been added to Table 11: Summary of Discharge Destination (Secondary outcomes) in the supplementary material. Findings for discharge destination are now described narratively.
Manuscript update (Line 164-167):
“The relationship between discharge destination and enhanced rehabilitation was explored by Calthorpe et al. [20] and Wu et al. [21] (S1: Table 11). Both studies found that adults receiving enhanced rehabilitation were more likely to be discharged directly home, with a lower proportion requiring inpatient rehabilitation compared to control groups.”

Comment 4:
Clarify eligibility criteria regarding severe TBI vs included TBI populations (define “severe” operationally and ensure consistency across Methods/Results/Tables). Eligibility criteria ambiguity: severe TBI exclusion vs included TBI populations. The methods state exclusion of severe neurological/cognitive impairment (including severe TBI), yet the included evidence base is described including TBI populations.
Requested action: Please specify the operational definition of “severe TBI” (e.g., GCS thresholds, ICU/mechanical ventilation, imaging criteria) and confirm whether included TBI trials excluded severe cases. Consider adding TBI severity information to study characteristics (Table 1) and/or a short clarification in Methods and Results.
Response:
Thank you for this constructive comment. Eligibility criteria has been described in the supplementary material in Supplementary Methods to align with the narrative review.
As suggested, the TBI severity information for the included study has been added to align with the eligibility criteria of “severe TBI” via Supplementary Materials in Table 6: Study characteristics and participant demographics of RCTs” in the manuscript, specifically the admission diagnosis/injury to “Moderate TBI (GCS 9-12)”. Furthermore, supplementary methods including further details on inclusion and exclusion criteria have been added to the supplementary materials.
Comment 5:
Safety outcomes and handling of deaths/adverse events. Several trials appear to have limited or inconsistent adverse event reporting; in at least one instance, deaths are mentioned with exclusion/attribution language that is difficult to interpret from the review alone. In addition, please align any statements about safety with the extent of harms reporting (“not reported” vs “no events observed”) and consider tempering directional language for non-significant findings.
Requested action: Please (a) standardize safety reporting across included studies (reported vs not reported), (b) clarify how mortality data were treated (and whether original trials used ITT), and (c) soften any “safe” language to reflect limitations in harms reporting.
Response:
Thank you for this helpful comment. The manuscript has been revised to standardise safety reporting and clarify the limitations in adverse events and mortality data.
Manuscript update (Line 170-178): “Evidence for mortality and adverse events associated with enhanced rehabilitation following traumatic injury is limited and inconsistently reported. Across the included studies, no increases in mortality were reported with enhanced rehabilitation compared to usual therapy (S1: Table 12). Most studies did not report on adverse events, and this absence of reported harms should not be interpreted as indicating that no events occurred. In the studies that did report outcomes, events were generally minimal. Calthorpe et al. [20] reported no deaths during and did not report adverse events. Fan et al. [19] reported a single death from other reasons and no adverse events associated with enhance rehabilitation, while Wu et al. [21] did not report on mortality or adverse events.”

Comment 6:
Protocol amendment after the final search date: clarify what changed and why. The manuscript notes protocol registration and a later update intended to “clarify eligibility/intervention definitions,” but the specific before/after changes and their impact are not sufficiently transparent.
Requested action: Please provide a concise description (preferably a small table) of what was amended, the rationale, and whether any studies/outcomes would have been included/excluded under the pre-amendment criteria.
Response:
Thank you for this comment. The review protocol was prospectively registered to PROSPERO and later updated to clarify eligibility criteria and definitions of interventions. As the manuscript was revised to a narrative review, these details are repositioned in the Supplementary Materials under the Supplementary Methods section which includes a Summary of Protocol Amendments. However, to ensure transparency, this has been added to the Supplementary Materials document under Supplementary Methods where we have noted that these amendments did not change the studies included.
Comment 7:
The PRISMA checklist indicates reporting bias and certainty assessments, but these elements are not clearly identifiable in the main manuscript as presented.
Requested action: If performed, present the methods and results (e.g., publication bias feasibility, certainty assessment such as GRADE). If not performed (e.g., too few studies), state this explicitly and adjust the PRISMA checklist mapping accordingly
Response:
We thank the reviewer for this feedback. Although the manuscript was revised to a narrative review, the PRISMA checklist has been omitted from the manuscript. Furthermore, formal assessments of risk of bias were conducted for all included RCTs using the Cochrane ROB 2 tool and the results were presented narratively (Supplementary Materials – Summary of Methodological Quality and Risk of Bias and Figures 2-7). A formal certainty assessment was not performed due to the limited number of studies and the narrative synthesis approach. The narrative summaries and RoB figure outputs have been transferred from the manuscript to supplementary materials to better align with the format of a narrative review.

Supplementary Material update: “Methodological Quality of Included Studies
The RCTs reviewed varied in methodological quality (Figures 2-7) [51]. Although randomisation was methodologically sound across all three studies, small sample sizes limited the precision of effect estimates and the generalisability of the findings. One study by Calthorpe et al. [20] was well conducted and appeared largely free from bias, supporting confidence in the reported results. In contrast, the other two RCTs [19, 21] were of lower quality, with several outcome measures affected by missing data, inconsistencies in intervention delivery, and challenges in outcome measurement. Overall, these methodological limitations suggest that the current evidence should be interpreted cautiously. Formal assessment of publication bias was not feasible due to the small number of included studies and narrative synthesis.”

Comment 8:
Data conversions (median→mean; unit conversions) and sensitivity analysis. The review uses statistical conversions (e.g., median/IQR to mean/SD) and unit conversions (e.g., minutes to days).
Requested action: Please specify exactly which studies/outcomes were converted, provide the converted inputs (or a clear trail in the supplement), and, where feasible, include a sensitivity analysis excluding converted estimates or using alternative reasonable conversion assumptions.
Response:
We thank the reviewer for these comments. Following the transition from a systematic review to a narrative review, no statistical or unit conversions were performed. Thus no formal sensitivity analyses were required. All data presented is as reported in the original studies.
Comment 9:
Intervention characterization and contamination risk. “Enhanced rehabilitation” versus “usual care” remains difficult to compare across trials due to incomplete reporting and the possibility that usual care may include similar services “as needed.” Suggest adding a concise table summarizing the actual differences (frequency, duration, disciplines involved, start time) and explicitly discussing contamination/crossover as a limitation.
Response:
We thank the reviewer for these valuable suggestions. Table 1 in the manuscript summarises the reported components of enhanced rehabilitation and usual care across the studies. As the original studies provided limited details regarding usual care, no additional information could be added. However, we have clarified this
limitation in the manuscript, including explicitly addressing contamination and crossover risks.

Manuscript update (Line 101-103): “Therapies provided for control groups also varied (Table 1) across trials, with frequent limited reporting of components and an overlap in type of intervention delivered between groups.”

Manuscript update (Line 357-368): “Across the included studies there was substantial heterogeneity in both intervention and control therapies, including differences in intervention type, timing, and dosage. Subsequently, this limits comparisons across studies and the interpretation of findings. Usual therapy commonly included PT, whilst OT was delivered as required. In the RCT by Fan et al. [19], the type of intervention provided in usual care was the same as the enhanced rehabilitation group, with the only difference being timing and frequency of delivery. Additionally, the individuals receiving usual care in the RCT by Wu et al. [21] could receive enhanced rehabilitation if indicated due to ethical reasons. Thus, these study designs present the risk of contamination and crossover between groups, which may reduce between group comparison and lead to an underestimation of intervention effects. As enhanced rehabilitation becomes more similar to usual care, it may be difficult to differentiate between intervention and control groups.
Comment 10:
Consistency and precision in language for non-significant findings. Where effects are not statistically significant, consider using “no clear evidence of difference” rather than implying directional effects without qualification. If directionality is described, label explicitly as a trend.
Response:
We appreciate the constructive feedback. Language throughout the manuscript has been reviewed and updated accordingly for non-significant findings in the narrative synthesis.
Comment 11:
Risk-of-bias summary consistency. Ensure that any narrative summaries of RoB (counts of outcomes at low/high risk) exactly match the RoB figure/table outputs; small discrepancies can distract readers and reviewers.
Response:
We thank the reviewer for this feedback. The narrative summaries and RoB figure outputs have been transferred from the manuscript to supplementary materials to better align with the format of a narrative review.
Supplementary Material update:
Supplementary Material update:
“Methodological Quality of Included Studies
The RCTs reviewed varied in methodological quality (Figures 2-7) [51]. Although randomisation was methodologically sound across all three studies, small sample sizes limited the precision of effect estimates and the generalisability of the findings. One study by Calthorpe et al. [20] was well conducted and appeared largely free from bias, supporting confidence in the reported results. In contrast, the other two RCTs [19, 21] were of lower quality, with several outcome measures affected by missing data, inconsistencies in intervention delivery, and challenges in outcome measurement. Overall, these methodological limitations suggest that the current evidence should be interpreted cautiously. Formal assessment of publication bias was not feasible due to the small number of included studies and narrative synthesis.”

Comment 12:
Supplementary material polish. Please proofread the supplementary tables for numbering, quotation marks, and step references (especially search strategies), as these are central to reproducibility.
Response:
Thank you for this comment. As the manuscript has been converted to a narrative review, detailed search strategies including numbered step references have been removed from the supplementary material. The remaining supplementary tables have been proofread for formatting consistencies. The search process is now described narratively and we have retained the PRISMA flow diagram (See Figure 1 between line 140-141) for clarity.
Comment 13:
Future research implications. The call for improved intervention reporting (e.g., using TIDieR principles) is appropriate; consider adding 2–3 concrete minimum reporting items (dose, timing, intensity, co-interventions) as guidance.
Response:
We thank the reviewer for this helpful suggestion. The discussion has been expanded to incorporate.
Manuscript update (Line 292-296):
“Improved reporting of interventions using tools such as the Template for Intervention Description and Replication (TIDieR) checklist or the Consensus on Exercise Reporting Template (CERT) is recommended to facilitate reproducibility and implementation [38-40]. At a minimum, studies should clearly report intervention timing, dosage, and intensity to facilitate replication and implementation.”
Manuscript update (Line 369-373):
“The uncertainty of optimal timing, intensity, and dosage impacts standardisation, adherence, and implementation of enhanced rehabilitation models of care. Future re-search should use standardised reporting frameworks to ensure that intervention details are clearly described, providing guidance to clinicians and trauma centres, including adequate allocation of resources such as staffing to deliver interventions.”

Round 2

Reviewer 1 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

.

Author Response

Nil comments noted.

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

The authors have responded to the critique of the previous submission.  The paper does not extend too much and give conclusions that are definite.  It is a valuable area and this paper shows there is a dearth of literature in it, with only the 3 reasonable papers.   For this reason it is more suitable for publication now.

Author Response

Comment 1:
The authors have responded to the critique of the previous submission. The paper does not extend too much and give conclusions that are definite. It is a valuable area and this paper shows there is a dearth of literature in it, with only the 3 reasonable papers. For this reason it is more suitable for publication now.
Response:
We are pleased that the revisions adequately addressed previous concerns. We thank the reviewer for their re-review and for recommending the manuscript for publication.

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

Thank you for the careful revision of the manuscript. The restructuring toward a narrative synthesis has improved clarity and strengthened the clinical focus of the paper. The cautious interpretation of findings and acknowledgement of the limited evidence base are appropriate. Overall, the manuscript addresses an important clinical question and is suitable for publication pending minor revisions.

I have only two minor points that should be addressed before acceptance:

1. Language describing non-significant findings
   In several sections (e.g., length of stay and discharge destination), the wording may imply directional effects despite statistical non-significance (e.g., “more likely,” “longer,” “reduced”). Please revise such statements to clearly reflect the absence of statistically significant differences (e.g., “no clear evidence of difference” or “a non-significant trend toward…” where appropriate). This will ensure precision and avoid potential over-interpretation.

2. Applicability to severe TBI populations
   The eligibility criteria mention exclusion of severe neurological impairment, including severe TBI, yet TBI populations are included in the review. Please clarify how “severe TBI” was operationally defined, or alternatively, include a brief statement acknowledging that the applicability of findings to patients with severe TBI remains uncertain. This will improve clarity regarding the scope of the conclusions.

With these minor clarifications, I believe the manuscript will be appropriate for publication.

Author Response

Comment 1:
Thank you for the careful revision of the manuscript. The restructuring toward a narrative synthesis has improved clarity and strengthened the clinical focus of the paper. The cautious interpretation of findings and acknowledgement of the limited evidence base are appropriate. Overall, the manuscript addresses an important clinical question and is suitable for publication pending minor revisions. I have only two minor points that should be addressed before acceptance. […] With these minor clarifications, I believe the manuscript will be appropriate for publication.
Response:
We sincerely thank the reviewer for their constructive feedback and for their kind acknowledgement of revisions made. We have addressed both minor revisions in the following responses.

Comment 2:
Language describing non-significant findings
In several sections (e.g., length of stay and discharge destination), the wording may imply directional effects despite statistical non-significance (e.g., “more likely,” “longer,” “reduced”). Please revise such statements to clearly reflect the absence of statistically significant differences (e.g., “no clear evidence of difference” or “a non-significant trend toward…” where appropriate). This will ensure precision and avoid potential over-interpretation.
Response:
Thank you for this actionable advice. The language describing non-significant findings has been adjusted accordingly in the manuscript.
Manuscript update (Line 155-158): “Subgroup analyses suggested a non-significant trend towards lower pain scores in those requiring subacute rehabilitation in the enhanced rehabilitation group compared to the control group at discharge from subacute rehabilitation and at follow-up [21].”
Manuscript update (Line 161-165): “The acute length of stay across studies (S1: Table 10) following enhanced rehabilitation was longer than that of those who underwent usual therapy, whilst intervention subgroups that received inpatient rehabilitation had shorter subacute and combined hospital length of stays compared to controls [20, 21]. Nonetheless, there was no clear evidence to support these trends.”
Manuscript update (Line 166-170): “The relationship between discharge destination and enhanced rehabilitation was explored by Calthorpe et al. [20] and Wu et al. [21] (S1: Table 11). Both studies found that adults receiving enhanced rehabilitation were discharged directly home at a higher proportion, with fewer requiring inpatient rehabilitation compared to control groups, though these findings were not statistically significant.”

Comment 3:
Applicability to severe TBI populations
The eligibility criteria mention exclusion of severe neurological impairment, including severe TBI, yet TBI populations are included in the review. Please clarify how “severe TBI” was operationally defined, or alternatively, include a brief statement acknowledging that the applicability of findings to patients with severe TBI remains uncertain. This will improve clarity regarding the scope of the conclusions.
Response:
We appreciate the reviewer for raising this point. Severe TBI was operationally defined as a GCS ≤8, and studies including patients at this level were excluded. TBI populations in this review therefore include mild-to-moderate severity with GCS 9-12. We acknowledge that the applicability of findings to patients with severe TBI remains uncertain and have added a statement in the manuscript discussion under 7.2 Generalisability of Enhanced Rehabilitation.

Supplementary Material 1: Supplementary Methods section under Population: “Studies with adults who were unable to participate or tolerate interventions due to severe neurological or cognitive impairment, or medical instability (e.g. spinal cord injuries, or severe traumatic brain injuries (TBI with GCS≤8) were excluded.”

Manuscript update (Line 272-274): “Adults with severe traumatic brain injury (Glasgow Coma Scale ≤8) were outside the scope of this review. Therefore, the extent of which the review findings specifically apply to severe TBI populations warrants dedicated investigation.”

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors, 

Although the topic is potentially of interest, I believe that the overall research methodology presents major flaws that undermine the validity of the work. The concept of traumatic injury, as the authors themselves acknowledge, encompasses several different aspects. Moreover, the concept of acute hospitalization is highly variable, depending both on the patient's clinical condition and on the healthcare system of each country. In light of this, several concerns arise. How can patients with such heterogeneous clinical conditions be compared in terms of the usefulness and effectiveness of early rehabilitation? How can a mild traumatic brain injury be compared with a severe one? Or how can a complex polytrauma (for example, involving multiple fractures, internal organ damage, and severe traumatic brain injury) be compared with a condition resulting from a single bone fracture? Likewise, how can a patient with central nervous system involvement be compared with one presenting peripheral nervous system lesions when discussing early rehabilitation?

In my opinion, an essential selection criterion is missing — one that accounts for the severity, complexity, and diagnosis of the patients included in the analysis. The term traumatic injury is far too broad. It might be advisable to revise the inclusion and exclusion criteria, however, this would require a complete reworking of the entire systematic review. Furthermore, the authors set as their primary objective the assessment of the effectiveness of early intervention, including in the review three randomized controlled trials and comparing the effects of experimental versus control treatments. This methodological approach could have been appropriate if the included studies had compared a group starting rehabilitation early (experimental) with another beginning later (control), or if the intent of the review had been to compare the effects of specific rehabilitation protocols conducted by physiotherapists and occupational therapists during the acute hospitalization phase. Even in this case, however, it would necessitate a major revision and a substantial reconsideration of the entire systematic review.

In light of these considerations, I must unfortunately recommend rejection of the manuscript, as — despite its potentially relevant topic — it is methodologically inadequate and lacks several fundamental elements required for publication.

Author Response

Comment 1:
The concept of traumatic injury, as the authors themselves acknowledge, encompasses several different aspects. The term traumatic injury is far too broad. An essential selection criterion is missing accounting for the severity, complexity, and diagnosis of the patients included in the analysis. It might be advisable to revise the inclusion and exclusion criteria.

Response:
Thank you for these constructive comments to ensure that the population selection criteria are explicit and accurately reflects the characteristics of the included patients. While the inclusion and exclusion criteria were agreed upon among the screening authors during the initial submission, we acknowledge that these were not clearly described in the manuscript.

To address this, the inclusion and exclusion criteria have now been revised in the Methods section to account for severity, complexity, and diagnosis as follows:
Inclusion criteria: (line 182-187)
“[…] admitted as inpatients to an acute hospital immediately after sustaining unintentional traumatic injury were included. In rehabilitation research, it is recommended to select a population based on their clinical presentation rather than restricting to a specific diagnosis [21]. Traumatic injuries are inherently heterogenous therefore broader inclusion criteria were applied to reflect the nature of presentations in most trauma units.”

Exclusion criteria: (188-195)
“Studies were excluded if the principal diagnosis was for non-traumatic related conditions (e.g. stroke). Studies with adults who were unable to participate or tolerate interventions due to severe neurological or cognitive impairment, or medical instability (e.g. spinal cord injuries, or severe traumatic brain injuries (TBI)) were excluded. Studies were excluded if participants presented with an isolated injury (e.g. single fracture) or had no physical injuries after assessment. Studies that examined adults with pre-existing conditions or comorbidities impacting function (e.g. previously dependent for mobility, living in a nursing home, or required assistance to mobilise) were excluded.”

Comment 2:
The concept of acute hospitalization is highly variable, depending both on the patient's clinical condition and on the healthcare system of each country.

Response: Thank you for highlighting this important point. We acknowledge that variations in healthcare systems influence the interpretation and generalisability of acute hospitalisation. This has now been explicitly addressed as a limitation in the Discussion section (line 753-756):
“In conjunction with the heterogeneous nature of traumatic injuries, variations in healthcare systems, particularly the acute hospital structure and rehabilitation services, present challenges for meaningful comparative analysis.”

Comment 3:
How can patients with such heterogeneous clinical conditions be compared in terms of the usefulness and effectiveness of early rehabilitation? How can a mild traumatic brain injury be compared with a severe one? Or how can a complex polytrauma (for example, involving multiple fractures, internal organ damage, and severe traumatic brain injury) be compared with a condition resulting from a single bone fracture? Likewise, how can a patient with central nervous system involvement be compared with one presenting peripheral nervous system lesions when discussing early rehabilitation?

Response:
We thank the reviewer for these thoughtful questions. As noted, traumatic injuries are inherently heterogeneous therefore, broader inclusion criteria were applied to reflects the clinical diversity typically seen in trauma units. Despite this, we recognise this heterogeneity as a limitation and have discussed this in the Discussion section in the limitations of this review.

The revised text states: (line 650-676)
“[…] the benefits of enhanced rehabilitation have been shown to be effective in a wide range of other clinical presentations. For example, enhanced rehabilitation delivered in an acute hospital setting is associated with improvements in functional outcomes following heart and/or lung transplantation [39], following hip and knee operations [36, 40, 41], and in geriatric populations [42]. Despite the variation in presentation, it is important to acknowledge that while the magnitude of effects may differ between conditions, the direction of benefit is consistent. A quasi-experimental study that implemented enhanced rehabilitation across the acute hospital, encompassing adults admitted under neurosciences, oncology, heart and lung transplant, surgical, and other medical services demonstrated improvements in functional independence, a significantly increased likelihood of direct discharge home from the acute hospital, and a reduced requirement for subacute inpatient rehabilitation [43]. As previously mentioned, similar findings were identified by Gabbe et al. [30], and the associated evaluation study [38] with improvements in physical function and adjusted odds of discharging home following traumatic injury. Therefore, despite the heterogenous nature of traumatic injuries, the generalisability of enhanced rehabilitation may also apply to adults following traumatic injury. Further high-quality RCTs are required to enable subgroup meta-analysis to meaningfully address the heterogeneity of traumatic injury.”

In addition, the amended selection criteria (see Response 1) clarify the population parameters to address the reviewer’s concern regarding severity and injury type (e.g. central vs peripheral nervous system involvement).

Comment 4:
The authors set as their primary objective the assessment of the effectiveness of early intervention […]. This methodological approach could have been appropriate if the included studies had compared a group starting rehabilitation early (experimental) with another beginning later (control), or if the intent of the review had been to compare the effects of specific rehabilitation protocols […]

Response:
We thank the reviewer for the opportunity to clarify our methodological intent. The concept of early intervention has now been refined to ‘enhanced’ rehabilitation to more accurately reflect the scope of our study, consistent with recent clinical trial practice guidelines by the Society of Critical Care Medicine (USA). These guidelines also include a heterogenous population though within the intensive care setting. Enhanced rehabilitation is defined as: (line 198-200)

“[…] any physical intervention delivered on the ward that is more than usual care (e.g. timing of intervention or increased dosage) during the acute inpatient admission by PTs, OTs, or both disciplines.” 

While this concept was implicit in the original design, we recognise that it was not clearly conveyed in the manuscript. The revised definition now explicitly clarifies that the review compared enhanced rehabilitation to usual care rather than strictly early versus delayed rehabilitation.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,
This focused PRISMA-based review of early ward-based PT/OT after traumatic injury is methodologically sound but provides limited incremental novelty. Its conclusions—only three small RCTs, heterogeneous interventions/outcomes, inconclusive functional effects, and no significant differences in LOS/discharge—largely restate prior reviews and the 2022 NICE guidance that evidence remains insufficient for specific acute protocols. To strengthen its contribution, the manuscript should (i) temper the novelty claim; (ii) correct numerous language/formatting errors (notably the abbreviations table); and (iii) situate findings alongside recent observational “allied health model of care” studies that suggest potential benefits but require RCT confirmation. 

Author Response

Comment 1:
Limited incremental novelty. Its conclusions—only three small RCTs, heterogeneous interventions/outcomes, inconclusive functional effects, and no significant differences in LOS/discharge—largely restate prior reviews and the 2022 NICE guidance that evidence remains insufficient for specific acute protocols. Manuscript should temper the novelty claim.

Response:
Thank you for these valuable comments. We have revised the manuscript to temper the novelty claim and clarify the contribution of this review. The following changes have been incorporated in the Introduction (line 92-97):

“Collectively, these previous reviews examined enhanced rehabilitation across the continuum of care, with limited focus on rehabilitation delivered specifically within the acute hospital ward. Building on this existing literature, this review aimed to synthesise the effect of enhanced physical rehabilitation following traumatic injury, delivered by PTs and/or OTs in the acute hospital ward, on functional outcomes compared to control groups.”

Additionally, the Discussion section (line 526-528) emphasises that, while evidence remains limited, this review:

“[…] provides a more detailed understanding of the acute phase of care which was previously encompassed by broader analyses. Whilst this adds nuance to the existing evidence, the findings of this review support the need for high-quality studies.”

Comment 2:
Correct numerous language/formatting errors (notably the abbreviations table)

Response:
We appreciate this feedback. Language and formatting have been carefully reviewed and corrected throughout the manuscript, including revisions to the abbreviations table.

Comment 3:
Situate findings alongside recent observational “allied health model of care” studies that suggest potential benefits but require RCT confirmation.

Response:
We thank the reviewer for this helpful suggestion. The discussion has been expanded to incorporate relevant observational evidence highlighting the potential benefits of enhanced rehabilitation, while emphasising the ongoing need for RCT confirmation. These revisions are located in the Discussion section as follows:

Line 529-537:

“However, subsequent to the publication of the NICE guidelines, a single-centre observational study conducted by Gabbe et al. [30] conducted in a purpose-built ward at an Australian level 1 trauma centre reported a clinically important mean difference of 5.8 [31] in physical function, measured by the mILOA, following enhanced PT and OT as part of an intensive allied health model of care. These findings support the mILOA scores observed in the RCT by Calthorpe et al. [22] included in this review. This may raise the possibility of potential benefits of enhanced rehabilitation in the acute ward setting for improving physical function following traumatic injury although high-quality RCTs are still needed to support this assertion.”

Line 606-611:

“Furthermore, the single-centre observational study aforementioned by Gabbe et al. [30] concluded a non-significant reduction in hospital-acquired complication rates from baseline to the enhanced rehabilitation intervention phase. This may suggest that enhanced rehabilitation in the acute hospital ward following traumatic injury is safe, yet confirmation through RCTs is required.”

Line 631-636:

“In contrast, the observational study aforementioned by Gabbe et al. [30] reported a reduced median acute length of stay from baseline to the enhanced rehabilitation intervention phase; however, this was not significant. This may suggest that while consistently reduced acute hospital length of stay may not be expected following enhanced rehabilitation, it may contribute to a subsequent reduction in overall hospital length of stay.”

Line 638-646:

Furthermore, this trend is supported in the observational study by Gabbe et al. [30] which determined a significantly higher proportion of patients with traumatic injury discharged home compared to usual care following enhanced rehabilitation in a purpose-built ward. A study evaluating the observational study results concluded a 53% increase in the likelihood of discharging home at 12 months rather than inpatient rehabilitation or an acute transfer after accounting for demographics, injury type and Charlson Comorbidity Index [38]. These findings may indicate a potential association between enhanced rehabilitation and discharge destination, however further RCTs are warranted to confirm these findings.

Line 671-676:

“As previously mentioned, similar findings were identified by Gabbe et al. [30], and the associated evaluation study [38] with improvements in physical function and adjusted odds of discharging home following traumatic injury. Therefore, despite the heterogenous nature of traumatic injuries, the generalisability of enhanced rehabilitation may also apply to adults following traumatic injury. Further high-quality RCTs are required to enable subgroup meta-analysis to meaningfully address the heterogeneity of traumatic injury.”

Line 715-726:

“More specifically, the observational study by Gabbe et al. [30] demonstrated that implementing a cost-effective enhanced rehabilitation model in a Level I trauma center was feasible using a before and after design with a historical control group.”