This systematic review protocol was developed using guidance from the Cochrane Collaboration . The review protocol was registered with the PROSPERO International Prospective Register of Systematic Reviews  (Registration number CRD42012003120) to reduce the variability in the review process and increase the validity of the results [15, 17]. To increase the impact of our review , a committee of stakeholders (neurotrauma rehabilitation program clinicians, users, managers) has been asked to assist the review process by collaborating during specific steps.
We will search electronic databases using a strategy developed by the researchers using the medical subject headings pertaining to each database. The stakeholders’ committee will examine the search strategy for its comprehensiveness (key words and data sources). The quality of this strategy will be assessed by at least two independent information specialists using the Peer Review of Electronic Search Strategies Evidence-Based Checklist (PRESS EBC) [19, 20].
A professional librarian will perform the primary search in nine electronic databases: PubMed, EMBASE, CINAHL, AMED, PsychINFO, the Cochrane Library, FirstSearch, Web of Science, ProQuest dissertations and theses to retrieve primary studies published from January 1990 to October 2012. A secondary search will be performed by checking the reference lists of the studies included in the review. Unpublished, non-peer-reviewed sources (gray literature) will also be searched. For example, we will explore websites and publications of organizations dedicated to individuals with neurotrauma, official agencies, Google, and so on.
Studies examining 1) the association between length of stay or intensity of inpatient rehabilitation services and outcomes (level of function, participation, return to work), or 2) the determinants of inpatient rehabilitation length of stay in adults with neurotrauma (moderate and/or severe TBI or SCI), and 3) published after January 1990 will be considered. In a first selection phase, two review authors (CG, MEL) will independently screen the title and abstract of all identified studies for inclusion against the eligibility criteria.
Languages: English and French.
Publication types: experimental and epidemiological study designs including randomized control trials, non-randomized controlled trials, quasi-experimental, before and after studies, prospective and retrospective cohort studies, case control studies and analytical cross-sectional studies. Gray literature, qualitative and observational studies will also be taken into consideration.
Study population: adult (18 to 65 years old without geriatric profile) with moderate or severe traumatic SCI, adult with moderate or severe TBI.
Intervention setting: inpatient rehabilitation.
Publication types: editorial, case report, anecdotal report.
Study population: non-traumatic injuries (for example cancer, stroke, poliomyelitis, and so on), alternate level of care, children only, older people only.
Intervention setting: acute rehabilitation, outpatient rehabilitation, and community integration setting.
In a second selection phase, two independent reviewers (MEL, CG) will confirm the eligibility of the studies by reviewing the full text of the selected studies. A third party (LN) will consider all discrepancies not resolved by discussion in the course of the two selection phases and make a final decision about eligibility. A kappa statistic and the percent agreement will be used to calculate interrater agreement in both selection phases.
The stakeholders’ committee will be informed of the progress of study selection and invited to complete the list of selected studies by suggesting additional studies or results from gray literature.
Two reviewers (MEL, ASA) will independently extract data from the studies regarding 1) general information, 2) study characteristics, 3) participants’ characteristics, 4) outcomes, and 5) results, using a standardized form. A first extraction will be performed on 10 articles to pilot the extraction form and insure the comprehensiveness and reliability of the data extraction. The results of this pilot extraction will be examined to refine the process. Disagreements between the reviewers will be resolved by discussion and consensus. If a consensus cannot be reached, a third researcher will make the final decision (CG).
Risk of bias assessment
The Cochrane Collaboration's tool for assessment of risk of bias will be used to assess the risk of bias of randomized studies. The Transparent Reporting of Evaluations with Nonrandomized Design (TREND) grid  will be used to assess risks of bias in non-randomized studies. A risk of bias table will be generated with the principal biases and the methodological quality of the studies.
A narrative synthesis method  will first be used to describe the results of the systematic review. All included studies will be summarized in narrative form, and summary tables will be created showing key study characteristics (population characteristics, study outcomes, sample sizes, settings, statistics used, related results, and any other important aspect related to each research question of interest).
Then, when possible, meta-analysis methods will be applied . The tau-squared and I2 statistic will be used to quantify the statistical heterogeneity between studies examining our outcomes of interest, where P <0.10 and where I2 is larger than 50%, indicating a high level of statistical heterogeneity between the studies. If characteristics of the studies are homogeneous enough, we will group them and perform meta-analysis of the pooled data. If outcomes of interest of each included study were reported using different outcome measures on a continuous scale, the summary measures of effect in the form of standardized mean difference (SMD) for each outcome will be generated. All analyses will be conducted with R software (available at http://www.r-project.org).
We will ask two independent experts to assess the quality of our systematic review using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) tool [24, 25] and its development and interpretation guide .
The partners of the stakeholders’ committee will receive a synthesis of the narrative synthesis and meta-analysis results, to serve as a basis for formal discussion in a one-day meeting. This knowledge exchange activity will provide guidance for the interpretation of the results of the systematic review . It will increase our understanding of the impacts of the length of stay on actual rehabilitation practice and hopefully facilitate knowledge dissemination and appropriation by the principal clinical users.