Data mining versus manual screening to select papers for inclusion in systematic reviews: a novel method to increase efficiency.

Systematic reviews rely on identification of studies, initially through electronic searches yielding potentially thousands of studies, and then reviewer-led screening studies for inclusion. This standard method is time- and resource-intensive. We designed and applied an algorithm written in Python involving computer-aided identification of keywords within each paper for an exemplar systematic review of arm impairment after stroke. The standard method involved reading each abstract searching for these keywords. We compared the methods in terms of accuracy in identification of keywords, abstracts' eligibility, and time taken to make a decision about eligibility. For external validation, we adapted the algorithm for a different systematic review, and compared eligible studies using the algorithm with those included in that review. For the exemplar systematic review, the algorithm failed on 72 out of 2,789 documents retrieved (2.6%). Both methods identified the same 610 studies for inclusion. Based on a sample of 21 randomly selected abstracts, the standard screening took 1.58 ±â€…0.26 min per abstract. Computer output screening took 0.43 ±â€…0.14 min per abstract. The mean difference between the two methods was 1.15 min ( P  < 0.0001), saving 73% per abstract. For the other systematic review, use of the algorithm resulted in the same studies being identified. One study was excluded based on the interpretation of the comparison intervention. Our purpose-built software was an accurate and significantly time-saving method for identifying eligible abstracts for inclusion in systematic reviews. This novel method could be adapted for other systematic reviews in future for the benefit of authors, reviewers and editors.

Home-based therapy programmes for upper limb functional recovery following stroke.

BACKGROUND: With an increased focus on home-based stroke services and the undertaking of programmes, targeted at upper limb recovery within clinical practice, a systematic review of home-based therapy programmes for individuals with upper limb impairment following stroke was required. OBJECTIVES: To determine the effects of home-based therapy programmes for upper limb recovery in patients with upper limb impairment following stroke. SEARCH METHODS: We searched the Cochrane Stroke Group's Specialised Trials Register (May 2011), the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2011, Issue 2), MEDLINE (1950 to May 2011), EMBASE (1980 to May 2011), AMED (1985 to May 2011) and six additional databases. We also searched reference lists and trials registers. SELECTION CRITERIA: Randomised controlled trials (RCTs) in adults after stroke, where the intervention was a home-based therapy programme targeted at the upper limb, compared with placebo, or no intervention or usual care. PRIMARY OUTCOMES were performance in activities of daily living (ADL) and functional movement of the upper limb. SECONDARY OUTCOMES were performance in extended ADL and motor impairment of the arm. DATA COLLECTION AND ANALYSIS: Two review authors independently screened abstracts, extracted data and appraised trials. We undertook assessment of risk of bias in terms of method of randomisation and allocation concealment (selection bias), blinding of outcome assessment (detection bias), whether all the randomised patients were accounted for in the analysis (attrition bias) and the presence of selective outcome reporting. MAIN RESULTS: We included four studies with 166 participants. No studies compared the effects of home-based upper limb therapy programmes with placebo or no intervention. Three studies compared the effects of home-based upper limb therapy programmes with usual care. PRIMARY OUTCOMES: we found no statistically significant result for performance of ADL (mean difference (MD) 2.85; 95% confidence interval (CI) -1.43 to 7.14) or functional movement of the upper limb (MD 2.25; 95% CI -0.24 to 4.73)). SECONDARY OUTCOMES: no statistically significant results for extended ADL (MD 0.83; 95% CI -0.51 to 2.17)) or upper limb motor impairment (MD 1.46; 95% CI -0.58 to 3.51). One study compared the effects of a home-based upper limb programme with the same upper limb programme based in hospital, measuring upper limb motor impairment only; we found no statistically significant difference between groups (MD 0.60; 95% CI -8.94 to 10.14). AUTHORS' CONCLUSIONS: There is insufficient good quality evidence to make recommendations about the relative effect of home-based therapy programmes compared with placebo, no intervention or usual care.

Predictors of upper limb recovery after stroke: a systematic review and meta-analysis.

OBJECTIVE: To systematically review and summarize the current available literature on prognostic variables relating to upper limb recovery following stroke. To identify which, if any variables predict upper limb recovery following stroke. DATA SOURCES: We completed searches in MEDLINE, EMBASE, AMED, CINAHL and Cochrane CENTRAL databases. Searches were completed in November 2010. REVIEW METHODS: Studies were included if predictor variables were measured at baseline and linked to an outcome of upper limb recovery at a future time point. Exclusion criteria included predictor variables relating to response to treatment and outcome measurements of very specific upper limb impairments such as spasticity or pain. Two independent reviewers completed data extraction and assessed study quality. RESULTS: Fifty-eight studies met the inclusion criteria. Predictor variables which have been considered within these studies include; age, sex, lesion site, initial motor impairment, motor-evoked potentials and somatosensory-evoked potentials. Initial measures of upper limb impairment and function were found to be the most significant predictors of upper limb recovery; odds ratio 14.84 (95% confidence intervals (CI) 9.08-24.25) and 38.62 (95% CI 8.40-177.53), respectively. CONCLUSIONS: Interpretation of these results is complicated by methodological factors including variations in study populations, upper limb motor outcome scales, timing of baseline and outcome assessments and predictors selected. The most important predictive factors for upper limb recovery following stroke appears to the initial severity of motor impairment or function.

Do vascular risk factors explain the association between socioeconomic status and stroke incidence: a meta-analysis.

BACKGROUND: Reduced socioeconomic status (SES) is associated with an increased risk of stroke, although the mechanism is not clear. It may be that those with lower SES have a greater burden of classic vascular risk factors. METHODS: Our aim was to quantify the extent to which classic vascular risk factors explain the association between SES and stroke incidence. We conducted a systematic review and meta-analysis of studies examining the association of SES and stroke incidence, where classic vascular risk factors were considered. Searching MEDLINE, EMBASE and the Cochrane Library from 1980 onwards we identified 17 studies, 12 of these studies provided sufficient information to allow a meta-analysis. From each study the increased risk of stroke incidence, where the lowest socioeconomic category was compared with the highest, was recorded and pooled. The stroke incidence risks, adjusted for grouped classic risk factors, were also pooled. Review Manager 5 software was used for all analyses and results were analysed using hazard ratios (HR, 95% confidence interval, 95% CI) with a random effects model. RESULTS: Those with a lower SES were more likely to have a stroke (HR 1.67; 95% CI 1.46-1.91). Additional risk was reduced, but not eliminated, when classic vascular risk factors were adjusted for (HR 1.31; 95% CI 1.16-1.48). CONCLUSION: Low SES is associated with an increased risk of stroke that is partly explained by known classic vascular risk factors.

Simultaneous bilateral training for improving arm function after stroke.

BACKGROUND: Simultaneous bilateral training, the completion of identical activities with both arms simultaneously, is one intervention to improve arm function and reduce impairment. OBJECTIVES: To determine the effects of simultaneous bilateral training for improving arm function after stroke. SEARCH STRATEGY: We searched the Cochrane Stroke Trials Register (last searched August 2009) and 10 electronic bibliographic databases including the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library Issue 3, 2009), MEDLINE, EMBASE, CINAHL and AMED (August 2009). We also searched reference lists and trials registers. SELECTION CRITERIA: Randomised trials in adults after stroke, where the intervention was simultaneous bilateral training compared to placebo or no intervention, usual care or other upper limb (arm) interventions. PRIMARY OUTCOMES were performance in activities of daily living (ADL) and functional movement of the upper limb. SECONDARY OUTCOMES were performance in extended activities of daily living and motor impairment of the arm. DATA COLLECTION AND ANALYSIS: Two authors independently screened abstracts, extracted data and appraised trials. Assessment of methodological quality was undertaken for allocation concealment, blinding of outcome assessor, intention-to-treat, baseline similarity and loss to follow up. MAIN RESULTS: We included 18 studies involving 549 relevant participants, of which 14 (421 participants) were included in the analysis (one within both comparisons). Four of the 14 studies compared the effects of bilateral training with usual care. PRIMARY OUTCOMES: results were not statistically significant for performance in ADL (standardised mean difference (SMD) 0.25, 95% confidence interval (CI) -0.14 to 0.63); functional movement of the arm (SMD -0.07, 95% CI -0.42 to 0.28) or hand (SMD -0.04, 95% CI -0.50 to 0.42). SECONDARY OUTCOMES: no statistically significant results. Eleven of the 14 studies compared the effects of bilateral training with other specific upper limb (arm) interventions. PRIMARY OUTCOMES: no statistically significant results for performance of ADL (SMD -0.25, 95% CI -0.57 to 0.08); functional movement of the arm (SMD -0.20, 95% CI -0.49 to 0.09) or hand (SMD -0.21, 95% CI -0.51 to 0.09). SECONDARY OUTCOMES: one study reported a statistically significant result in favour of another upper limb intervention for performance in extended ADL. No statistically significant differences were found for motor impairment outcomes. AUTHORS' CONCLUSIONS: There is insufficient good quality evidence to make recommendations about the relative effect of simultaneous bilateral training compared to placebo, no intervention or usual care. We identified evidence that suggests that bilateral training may be no more (or less) effective than usual care or other upper limb interventions for performance in ADL, functional movement of the upper limb or motor impairment outcomes.

Motor recovery after stroke: a systematic review.

Loss of functional movement is a common consequence of stroke for which a wide range of interventions has been developed. In this Review, we aimed to provide an overview of the available evidence on interventions for motor recovery after stroke through the evaluation of systematic reviews, supplemented by recent randomised controlled trials. Most trials were small and had some design limitations. Improvements in recovery of arm function were seen for constraint-induced movement therapy, electromyographic biofeedback, mental practice with motor imagery, and robotics. Improvements in transfer ability or balance were seen with repetitive task training, biofeedback, and training with a moving platform. Physical fitness training, high-intensity therapy (usually physiotherapy), and repetitive task training improved walking speed. Although the existing evidence is limited by poor trial designs, some treatments do show promise for improving motor recovery, particularly those that have focused on high-intensity and repetitive task-specific practice.