The first afternoon of the Faculty of Health’s writing week at the University of Canberra included a workshop facilitated by Louise Ada from the University of Sydney. Louise is the Scientific Editor of the Australian Journal of Physiotherapy (AJP). She was introduced to the workshop participants by Gordon Waddington.
Louise noted in her introduction that a journal review process includes:
- An editor deciding whether or not to send a paper out to review.
- The possibility that reviewers can reject a paper after 1st, 2nd, 3rd versions.
- An editor having the authority to override reviewers’ recommendations.
Louise emphasised the importance of responding carefully to reviewers’ comments in the submission process. She identified two steps
Step 1: getting past the editor to the reviewer
Louise noted how busy editors are. She indicated that some large journals publish 10% of the papers they receive. This indicates a significant filter process prior to review and Louise discussed the impact of rejection at this stage in the process. Large journals tend to return papers within two weeks. Only general feedback is received at this point. She encouraged anyone who received this kind of rejection to turn the paper around within a week and submit to another journal in the appropriate format
Louise indicated that the following factors help to refer a paper to review:
- Good title
- Correct title page
- Abstract according to guidelines
- Appropriate checklists
- Cover letter explaining the above
She noted the importance of the abstract and the cover letter (comments to the editor). The etiquette of a cover letter requires: (1) a one-line statement of why the paper is important; (2) why you have sent it to the specific journal.
Louise clarified that titles of submitted papers should include as much information as possible ad recommended that ‘fun titles’ be reserved for oral presentations.
Louise noted the importance of the PICO system in determining a paper’s title: Participants, Intervention, Comparison, Outcome (note Cochrane‘s advice). She recommended that the design and results are included in the title of the paper. An example:
Mime therapy improves facial symmetry in people with facial nerve paresis: a randomised placebo-controlled trial.
The AJP requests the following Title/Cover Page details:
- Running Head
- Key words
- Word Count
- tables Figures
- Source(s) of support
- Competing Interests
Louise recommended that running head should use the PICO system too.
Abstract requirements from AJP are specified:
- Outcome Measures
- Trial registration
Question What is the effect of mime therapy on facial symmetry and severity of paresis in people with facial nerve paresis? Design Randomised controlled trial. Participants 50 people recruited from the Outpatient department of two metropolitan hospitals with facial nerve paresis for more than nine months. Intervention The experimental group received three months of mime therapy consisting of massage, relaxation, inhibition of synkinesis, and co-ordination and emotional expression exercises. The control group was placed on a waiting list. Outcome measures Assessments were made on admission to the trial and three months later by a measurer blinded to group allocation. Facial symmetry was measured using the Sunnybrook Facial Grading System. Severity of paresis was measured using the House-Brackmann Facial Grading System. Results After three months of mime therapy, the experimental group had improved their facial symmetry by 20.4 points (95% CI 10.4 to 30.4) on the Sunnybrook Facial Grading System compared with the control group. In addition, the experimental group had reduced the severity of their paresis by 0.6 grade (95% CI 0.1 to 1.1) on the House-Brackmann Facial Grading System compared with the control group. These effects were independent of age, sex, and duration of paresis. Conclusion Mime therapy improves facial symmetry and reduces the severity of paresis in people with facial nerve paresis.
[Buerskens CHG and Heymans PG (2006) Mime therapy improves facial symmetry in people with long-term facial nerve paresis: A randomised controlled trial. Australian Journal of Physiotherapy 177–183]
Louise identified two big mistakes in writing an Abstract: Mistake 1: start the abstract as if it is the paper – too much background uses up too many words. Mistake 2: not giving the results! Note that you must keep within the word limit for the Abstract (in most journals up to 250 words).
Louise pointed to the benefits of using a CONSORT (Consolidated Standards of Reporting Trials) statement linked to the study design. These can be downloaded as Word document from AJP and it would be impressive to include with a submitted paper. This works for PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) and STARD (STAndards for the Reporting of Diagnostic accuracy studies) too (and for observational studies).
Louise discussed the use of a PEDro Scale as a self-reporting tool in the cover letter to editors.
Step 2: get a good review
Louise recommended that you should have your research question at the end of the paper’s Introduction. This avoids burying the question in the paper! If the question is indented it adds to its clarity. Louise then discussed at length aims, research questions, and hypotheses. A research question: What? Does? Are? ends with a question mark. PICO is most helpful in this context. Louise talked about the discipline of writing for AJP.
This submission framework is on the AJP web site. Yet many authors submit paper without a research question. The guidelines are intended to make a paper easier for a reader (first of all the reviewer and then if accepted for professional colleagues). Note how many resources are available for authors on the AJP web site.
Louise recommends keeping introductions short! She suggested that overlong introductions tend to come from theses. She advises authors to plan a three paragraph introduction: first paragraph is the broadest (problem to be addressed) and narrowed down; second paragraph is what we know already (literature); the third paragraph is what we do not know and therefore the research question is … (the research questions are) …
Louise confirmed that writing a paper is not an opportunity to share your pain! (Other journals offer opportunities to discuss research issues, see, for example, Contemporary Clinical Trials)
AJP gives guidance about length of papers.
Louise the discussed writing up the Method (note singular) section of the paper. She recommended starting the section with a design subheading to address how you will answer the research question(s) identified in the introduction to the paper. She proposed this approach to:
- Allow a reader to see how you have answered there search question
- Outline the design elements (eg randomisation, blinding, counterbalancing) in chronological order
Louise presented an example of a CONSORT flow diagram with a time line for a clinical trial with six measurement periods. She emphasised how a well crafted flow diagram replaces verbal account of chronology.
She urged the workshop attendees to sort out what goes in the Method section and what goes in the Results section. For example, participants are described in the Method but where is it best to put a table of characteristics of population? Where to write about intervention: design in Method, compliance in Results. Louise noted that current better practice of reporting trials has criteria for inclusion and exclusion for patients and therapist (care giver) and for venues.
Louise turned her attention to Outcome Measures (the measure of the outcome that will answer your question). She noted that authors tend to want to include anything they have measured! She argues that authors should identify primary outcome measures (those that you live or die by, what you really want to know).
Louise suggested three levels of outcome: impairment/activity, limitation/participation (ICF); procedure/tool; units. She noted how often units of measurement are missing from research reports. She gave an example of clarity in this statement; spasticity was measured using the Tardieu Scale (0 to 4 where 0 is no spasticity).
In her discussion of the Results section of the paper, Louise suggested:
- Present the findings that will answer the research questions.
- Present findings in a way that makes sense to readers in a clinical journal.
- Include dense tables that are organised logically.
AJP includes individual data in Addenda (it was one of the first journals to do so).
She recommended presenting findings as size effect such as mean or odds ratios and observed that we are looking at clinical effect sizes. Over four weeks of training, the experimental group increased their walking speed …’
Louise provided an example of a dense table from a clinical trial using CONSORT guidelines about participants and therapists. There are templates for tables on the AJP web site. She presented an example of a table with a logical format: outcome measures, unit, with a caption to describe content.
Louise encouraged authors to ensure that graphs add value to survive the editorial process. She usrged the use of a full range of scale on Y and X axes.
Louise offered guidelines for captions and rules for their use: what is presented? what is the outcome (Y and X axes)? Louise noted the difference in presenting graphs in an oral presentation (talk to the graph legend rather than a caption). A paper requires clarity in a caption in order to provide reader with appropriate information.
Louise concluded a fascinating workshop with the exhortation to write clearly and repeat key terms, especially Participants, Intervention, Outcome measures.