Adaptation to poststroke visual field loss: A systematic review

Introduction Appropriate reporting concerns the clear and complete presentation of what was planned, carried out and found in a particular study. Inappropriate reporting makes the sound application of study findings into clinical practice difficult, if not impossible. For instance, clinicians are often faced with reports of studies which do not provide enough detail of interventions for them to be implemented in clinical practice.1 Inadequate reporting may also be a problem for the synthesis of evidence. It is common for systematic reviewers or meta-analysts to exclude otherwise eligible studies owing to incomplete reporting. Given the human and financial resources needed to conduct clinical studies, it is clear that inadequate reporting has also important ethical and moral implications.2 To help improve the reporting of key items, reporting guidelines have been developed for various types of research design. The positive effect of reporting guidelines on the quality of reporting has been documented in several reviews.3–7 The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement was developed in a collaborative effort of epidemiologists, methodologists, statisticians, researchers and journal editors to improve the reporting of observational studies, including cohort, case–control and cross-sectional studies. The initiative was established in 2004 and subsequently developed in several meetings. STROBE was published in October 2007 simultaneously in several journals including leading biomedical journals such as The Lancet, BMJ, Annals of Internal Medicine and PLoS Medicine. The STROBE recommendations are accompanied by an explanation and elaboration document (E&E) which explains in detail the rationale of each item and provides examples of transparent reporting from published articles. To date, the STROBE Statement is endorsed by over 100 journals as well as by the International Committee of Medical Journal Editors.8 The purpose of the STROBE Statement ‘is to ensure clear presentation of what was planned, done, and found in an observational study’ and the authors of STROBE clearly stress that ‘the recommendations are not prescriptions for setting up or conducting studies, nor do they dictate methodology or mandate a uniform presentation.’9 Notwithstanding the clear statement of the purpose of STROBE by its authors, some journal editors are concerned that the STROBE recommendations may be inappropriately used as an assessment tool to judge study quality or that researchers may use STROBE as a guideline to set up or conduct observational studies.10 In the present study, we examined in which circumstances and context STROBE has been cited by the researchers (how, where and why), and the background of researchers who cited STROBE. We considered the use of STROBE other than in relation to its original purpose as a reporting guideline for observational studies. Our results were presented and discussed during the August 2010 STROBE group meeting, when the group met to discuss a possible revision of the STROBE recommendations in light of the group's experience, anecdotal reports, feedback from STROBE users and new evidence. Methods Literature search On 12 August 2010, we conducted a two-step literature search in the Web of Knowledge database (http://www.isiknowledge.com/). The first step consisted of a search for STROBE publications. Our search strategy to identify STROBE publications contained words used in the title of the STROBE article and author name (appendix 1). This first search was conducted in order to set up the second search, where we identified articles which had cited any of the STROBE publications identified in the first step, using the ‘create citation report’ tool available in the Web of Knowledge database. Article selection Eligibility criteria for our first search (ie, search for STROBE publications) consisted of any STROBE publication regardless of language or version of publication (ie, checklist or E&E). We excluded the Strengthening the Reporting of Genetic Association Studies (STREGA) recommendations, which is an extension of STROBE to genetic epidemiology. There were no eligibility restrictions for our second search (ie, search of articles which cited STROBE). Any publication type, such as comments, editorials, systematic reviews or observational research, was eligible for inclusion. We then randomly selected a sample of 100 articles from the final list of articles which cited STROBE for detailed assessment. Data collection We collected the following information for STROBE publications identified through our initial search: journal in which STROBE was published, language of publication and number of citations which STROBE received per year. For the data extraction from articles which cited STROBE, a data-extraction form was developed by the authors and piloted by two reviewers (BRdC and MC) on a sample of 40 studies which were not part of the randomly selected articles. A single reviewer (BRdC) then conducted data extraction on a randomly selected sample of 100 articles for the following items: verbatim text of sentence(s) which included STROBE citation; reason for citation; STROBE article cited; type of article that cited STROBE (observational research; systematic review or meta-analysis; comment, editorial or letter to the editor; methodological article; or recommendations or narrative review); and background of researchers. Categories for ‘reason of citation’ were defined during the piloting of the data-extraction form, and additional categories were defined during data extraction as necessary. Whenever the data extractor (BRdC) was in doubt about the classification of an article, he discussed it with other authors until a decision was reached. In case of disagreements, the final decision was made by the senior epidemiologist (ME). Classification of reasons for citation Results of our search for articles which cited STROBE are presented narratively. Our sample of 100 randomly selected articles were classified into eight groups according to reason for citation: STROBE used as a reporting guideline; STROBE used as a methodological (ie, research design or conduct) guideline; STROBE used as an assessment tool of reporting quality; STROBE used as an assessment tool of methodological quality; STROBE cited to stress the importance of reporting guidelines; ‘unclear’; and ‘other.’ We considered the use of STROBE in reports of observational research to be appropriate when STROBE was used as a reporting guideline to ensure a clear and complete report of the study's design, conduct and findings. In this case, it was clear from the text that one or more items of STROBE were used to guide the reporting of the study. Our definition of appropriate use of STROBE is in line with the stated intentions of its authors ‘solely to provide guidance on how to report research well’ and ‘not prescriptions for designing or conducting studies.’9 Accordingly, we considered inappropriate the use of STROBE as a methodological guideline (ie, guideline to design or conduct observational research or as an assessment tool of methodological quality of publications reporting observational research). An example of what we considered inappropriate is the use of STROBE to assign an overall score of the methodological quality of a study. As stressed by the authors, STROBE was not developed to explain how research should be done, and so cannot validly be used to assess methodological quality. In box 1, we present text examples of appropriate and inappropriate uses of STROBE. The remainder of the citations which did not fall into either category were considered neutral, that is, neither appropriate nor inappropriate. Only observational research, systematic reviews or meta-analyses could be classified as appropriate or inappropriate. Box 1 Text examples of appropriate and inappropriate use of Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Based on the original purpose of STROBE, we consider the following text examples as appropriate or inappropriate use of STROBE Appropriate ‘The reporting of this study conforms to the STROBE statement.’11 ‘The STROBE guidelines were used to ensure the reporting of this observational study.’12 ‘The paper is reported following the STROBE statement.’13 Inappropriate ‘This prospective study was designed following recommendation of the STROBE statement.’14 ‘The analysis followed the principles of the Strengthening the Reporting of Observational Studies in Epidemiology initiative.’15 ‘The quality of selected studies was assessed using a modified version of the STROBE statement. …The STROBE was modified by adding questions about the serologic method used to confirm dengue diagnosis, use of viral isolation, and whether the study was based on a single dengue outbreak or transmission season. Use of viral isolation increased the score whereas single outbreak studies received no additional points. The quality score was the number of items from the STROBE checklist addressed as a percentage of the total number of items applicable (minimum of 23 and maximum of 25). Studies with a quality assessment below 50% were excluded.’16 For exploratory purposes, we classified articles which appropriately and inappropriately used STROBE according to the affiliation of the authors as a proxy for their background. Authors affiliated with epidemiology or public-health departments were assumed to have methodological training in epidemiological research. Owing to the identified large proportion of systematic reviews which inappropriately cited STROBE, we examined these reviews further, extracting information on the language of publication (English or other) and type of journal (specialist or general medical journal). Results Search for STROBE publications We identified 16 STROBE publications. Eleven of these publications were published in English, three in Spanish and two in German. Ten of these publications were published in 2007, five in 2008 and one in 2009. Table 1 displays the journals where STROBE was published, the number of citations received, the impact factor of the journal, the language of the publication and the date of the publication. Table 1 Overview of Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) publications ordered by amount of citations received STROBE publication (journals) No of citations Impact factor* Language Date of publication The Lancet 110 30.8 English October 2007 Ann Intern Med 99 16.2 English October 2007 Ann Intern Med (E&E)† 97 16.2 English October 2007 BMJ 69 13.7 English October 2007 Plos Med (E&E)† 62 13.1 English October 2007 J Clin Epidemiol 51 3.0 English October 2008 Plos Med 43 13.1 English October 2007 Epidemiology (E&E)† 42 5.5 English November 2007 Epidemiology 21 5.5 English November 2007 Prev Med 15 3.2 English October 2007 Bull World Health Organ 14 5.3 English November 2007 Gaceta Sanitaria 7 1.2 Spanish March 2008 Internist 4 0.3 German June 2008 Gaceta Sanitaria (E&E)† 3 1.2 Spanish March 2009 Rev Esp Salud Publica 3 – Spanish May 2008 Notfall & Rettungsmedizin 0 0.6 German May 2008 * Impact factors retrieved from 2009 Journal Citation Report Science Edition. † Publication of the explanation and elaboration (E&E) document that accompanied the STROBE recommendations. Search for articles which cited STROBE We identified 643 citations of STROBE since its first publication in October 2007 until August 2010. Two-hundred and four citations (32%) concerned the STROBE E&E document, and 439 (68%) citations concerned the STROBE checklist. The STROBE checklist published in The Lancet was the most cited, receiving 110 (17%) citations. The number of citations was associated with the impact factor of the journal as shown in table 1, the main exception being the Journal of Clinical Epidemiology, which received a large number of citations despite a lower journal impact factor. Half of the articles in our sample of 100 randomly selected articles were observational studies (32%) or systematic reviews/meta-analyses (19%). Table 2 displays the characteristics of these 51 research articles identified in our random sample. Table 2 Characteristics of 51 classifiable articles citing Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Classification Reason for citation Observational research Systematic review, meta-analysis Appropriate Guideline for reporting of study 26 0 Inappropriate Guideline for design and conduct of study 3 0 Tool to assess methodological quality 0 10 Neutral Tool to assess reporting quality 0 5 Example to stress the importance of reporting guidelines 0 1 Other* 2 1 Unclear 1 2 Total 32 19 * This category includes observational studies which cited STROBE as a guideline without making it clear whether the guidance related to reporting or methodology, or articles which mentioned STROBE in a commentary. Comments, editorials and letters accounted for 15% of the articles, methodological articles for 8%, and recommendations and narrative reviews for 26%. Of the 32 observational studies assessed, 26 (81%) made an appropriate use of STROBE, and three (10%) uses were considered inappropriate. Of the 19 systematic reviews or meta-analyses assessed, 10 (53%) made an inappropriate use of STROBE. Thus, of the 51 research articles, 26 (51%) used STROBE appropriately, and 13 (25%) used STROBE inappropriately. Finally, out of the 26 studies which used STROBE appropriately, 13 (50%) had authors affiliated with epidemiology or public-health departments, compared with six (46%) among the 13 studies with inappropriate use of STROBE. As for the 10 systematic reviews that inappropriately used STROBE as a tool to assess study quality, none were published in the Cochrane Database of Systematic Reviews, nine were published in medical specialist journals, five had at least one of the authors affiliated with epidemiology or public health departments, and all were published in English. The 10 articles related to cardiology, endocrinology, psychology, neurology, paediatrics, ophthalmology, internal medicine, periodontology, pharmacology and infectious diseases. Discussion STROBE has been cited over 600 times since its publication in 2007. Our citation analysis showed that most of the observational studies used STROBE as a guideline for reporting, whereas about half of the systematic reviews and meta-analyses used STROBE inappropriately, as a methodological quality-assessment tool. The inappropriate interchangeable use of the terms ‘methodological quality’ and ‘reporting quality’ is common and may explain why some researchers used STROBE as a tool for the assessment of methodological quality or as a guideline to design and conduct observational research. Reporting quality refers to the completeness with which a study is presented and whether major items for the proper appraisal of internal and external validity of findings are clearly reported.17 Methodological quality refers to the appropriateness of the methods employed in the design and conduct of epidemiological research, which determines the reliability of findings (ie, internal validity). Even though the authors of STROBE clearly state that its purpose is to guide reporting of observational research, we found that it is commonly used as a methodological quality-assessment tool in systematic reviews and sometimes as a guideline to design and conduct observational studies. Some journal editors have expressed concern that the STROBE recommendations may be inappropriately used as an assessment tool to judge study quality, or that researchers may use STROBE as a guideline to set up or conduct observational studies.10 Our results show that these concerns were justified, although we feel that some items of STROBE might be useful when designing or conducting an observational study. In particular, the explanation and elaboration STROBE document (E&E) could be useful to inform methodological decisions, especially for researchers with little formal training in epidemiology. Although some items of STROBE may be related to risk of bias, many of the items are exclusively related to transparent reporting. For instance, the first item of STROBE asks researchers to indicate in the title of their manuscript the design of their study. Obviously, whether authors report the design of their study in the title of their manuscript will not influence the effect estimates of their study or compromise the generalisability of their findings. To illustrate the pitfalls of using STROBE as a methodological assessment tool, we refer to the last text example we provided as an example of inappropriate use of STROBE in box 1. In this example, the authors of a systematic review transformed STROBE into a scale by assigning scores to each of its items, and a final score was used to determine the eligibility of studies for inclusion into the systematic review. By doing so, authors may have introduced bias to their findings, as different tools used to score methodological quality may result in different conclusions regarding the overall appropriateness of the methods employed in a particular study.18 The misuse of STROBE as a tool to assess methodological quality may be explained by the lack of validated and accepted tools for such assessments. As a consequence, authors who want to assess methodological quality of studies may turn to reporting guidelines. The absence of reliable tools that are based on sound empirical evidence to assess the quality of observational studies was documented by Sanderson et al in 2007.19 There is an extensive literature indicating that the misuse of findings of scientific articles is common across different fields of healthcare.7 20–29 In a similar analysis, Moher et al reported that the Consolidated Standards of Reporting Trials statement, a reporting guideline for randomised controlled trials, has also been misused by 4–10% of assessed articles as a guideline for quality assessment.7 30 Taken together, these results indicate a failure of not only authors but also peer-reviewers in recognising and avoiding misuse of guidelines. The peer-review process should ideally prevent misuses that are of major concern, in particular when misuses may dictate the outcomes of research as discussed above. We observed an association between the numbers of citations received and the impact factor of the journal in which STROBE was published. Even though the contents of many STROBE publications are identical, researchers preferred to cite a STROBE article published in a journal with a higher impact factor. This association was also observed by Perneger, who concluded that citations received by an article are not merely explained by scientific merit.31 Although reasons directing authors' choices are not clear, for some, a high impact factor may suggest the acceptance by a broader audience. The main exception to this association was the STROBE checklist published by the Journal of Clinical Epidemiology, which was also highly cited. This may have occurred because STROBE is particularly relevant for the daily work of the readers of this journal. We found no clear pattern regarding affiliations of authors to an institution with a methodological focus. In particular, we found no difference in affiliations between authors who used STROBE appropriately and authors who used STROBE inappropriately. Similarly, no clear patterns emerged when we examined more closely the characteristics of the 10 systematic reviews that inappropriately cited STROBE. A limitation of our findings is the fact that we included only articles which cited STROBE. This may have resulted in selection bias, since some researchers may have used STROBE for their study but not formally cited it. Such studies could not be identified by our search strategy. However, it is unclear to us in which direction this possible selection bias may have influenced our findings. Citations to the STROBE checklist have been steeply increasing since its first publication. STROBE is commonly used according to its original purpose, that is, a reporting guideline of observational research. However, despite the clear statement of its objective by its authors, STROBE is misused by some researchers reporting observational studies and about half of authors reporting systematic reviews, because they used it either as a tool to assess methodological quality or as a guideline on how to design and conduct observational studies. Further studies are required to define the consequences of the inappropriate use of reporting recommendations in clinical and epidemiological research, in particular the use of these checklists as an instrument to assess the methodological quality of studies. Supplementary Material Supporting Appendix Supporting Statement Author's manuscript Reviewer comments

Background Vision loss after optic neuropathy is considered irreversible. Here, repetitive transorbital alternating current stimulation (rtACS) was applied in partially blind patients with the goal of activating their residual vision. Methods We conducted a multicenter, prospective, randomized, double-blind, sham-controlled trial in an ambulatory setting with daily application of rtACS (n = 45) or sham-stimulation (n = 37) for 50 min for a duration of 10 week days. A volunteer sample of patients with optic nerve damage (mean age 59.1 yrs) was recruited. The primary outcome measure for efficacy was super-threshold visual fields with 48 hrs after the last treatment day and at 2-months follow-up. Secondary outcome measures were near-threshold visual fields, reaction time, visual acuity, and resting-state EEGs to assess changes in brain physiology. Results The rtACS-treated group had a mean improvement in visual field of 24.0% which was significantly greater than after sham-stimulation (2.5%). This improvement persisted for at least 2 months in terms of both within- and between-group comparisons. Secondary analyses revealed improvements of near-threshold visual fields in the central 5° and increased thresholds in static perimetry after rtACS and improved reaction times, but visual acuity did not change compared to shams. Visual field improvement induced by rtACS was associated with EEG power-spectra and coherence alterations in visual cortical networks which are interpreted as signs of neuromodulation. Current flow simulation indicates current in the frontal cortex, eye, and optic nerve and in the subcortical but not in the cortical regions. Conclusion rtACS treatment is a safe and effective means to partially restore vision after optic nerve damage probably by modulating brain plasticity. This class 1 evidence suggests that visual fields can be improved in a clinically meaningful way. Trial Registration ClinicalTrials.gov NCT01280877

Partial blindness after brain injury has been considered non-treatable. To evaluate whether patients with visual-field defects can profit from computer-based visual restitution training (VRT), two independent clinical trials were conducted using patients with optic nerve (n = 19) or post-chiasmatic brain injury (n = 19). In post-chiasma patients, VRT led to a significant improvement (29.4%) over baseline in the ability to detect visual stimuli; in optic nerve patients, the effects were even more pronounced (73.6% improvement). Visual-field enlargements were confirmed by the observation of a visual-field expansion of 4.9 degrees-5.8 degrees of visual angle and improved acuity in optic nerve patients. Ninety five percent of the VRT-treated patients showed improvements, 72.2% confirmed visual improvements subjectively. Patients receiving a placebo training did not show comparable improvements. In conclusion, VRT with a computer program improves vision in patients with visual-field defects and offers a new, cost-effective therapy for partial blindness.