Scientific writing – following guidelines

Authors


Scientific writing is not easy. In most cases, published articles are the result of several months (or even years) of hard work involving the input of a primary author, a crew of co-authors, a journal's editorial team and a scientific review panel. Although the task of publishing an article in the veterinary press can appear daunting, it is also extremely rewarding both for the authors and for the veterinary community as a whole. It is important that potential authors do not lose sight of the positive impact they are making when their manuscript is made available to the public. It is also important that potential authors are aware of the extensive support network that is available to assist them with their clinical research and scientific writing. The BSAVA and its official journal, The Journal of Small Animal Practice (JSAP), have an ongoing aim to encourage colleagues in general practice to remain actively involved in clinical research. Last year at the 2012 BSAVA Congress, there was a JSAP sponsored presentation about writing for publication that was accompanied by an editorial on scientific writing (Polton & Grant 2012). The lecture was well attended and the guidelines for scientific writing were so well received that a similar presentation will be made at this year's BSAVA Congress.

Hypothesis Testing

There are several important but common traits of high-quality clinical studies that are successfully published. Most frequently, the study begins with a clinically important question, or hypothesis. The person posing the question can either accept the task of designing an experiment that will allow the question to be answered or they can forward the question to another clinician. Before it is established if the hypothesis can be tested by that individual, a number of other questions must be answered. These include: (1) Has the question been answered before? (2) Do I have access to a representative study population? (3) Do I have access to the expertise and equipment necessary to test my hypothesis? and (4) Do I have the means to present my clinical study in an appropriate scientific format?

Has The Question Been Answered Before?

The first stage after identifying a clinically important hypothesis is to check whether the hypothesis has already been tested. This requires a systematic review of the current literature relevant to the hypothesis. This review helps establish what is known about the topic under investigation and may prevent unnecessary repetition of work already performed by others. The literature review will subsequently form the basis of the Introduction and Discussion sections of a scientific manuscript. The choice of web-based medical literature search engines is expanding, with common sources including Pubmed, Veterinary Information Network (VIN) and Google Scholar.

Do I Have Access To A Representative Study Population?

Access to an entire population (e.g. all dogs in Spain) is rarely possible. Therefore, it is usually necessary to identify a sample population of animals that appropriately represents the entire population. There are a number of pitfalls or biases that exist when a sample population is chosen. When we are obliged to infer features of the entire population from information about only some members of the population, randomness is usually the key component of obtaining a representative sample. In some cases, however, random sampling is not appropriate. In this month's JSAP, Tabar et al. (2013) report on the use of polymerase chain reaction (PCR) in the diagnosis of filariasis in a population of dogs evaluated in private practice in southeastern Spain. In this epidemiological study, the most important consideration is that the population is chosen from an area that is endemic for leishmaniosis and filariasis rather than from a random sample of all dogs in Spain. In clinical trials that require therapeutic intervention, random sampling and allocation to a treatment or control group are essential precautions in avoiding bias. The review process often identifies biases that can significantly reduce the clinical impact of a study. Problems in bias are best avoided early in the study design process by consultation with a statistician or epidemiology textbook (e.g. Dohoo et al. 2003). Essential information on sampling technique and the sample size required to achieve statistical significance should be collected at this stage. Consideration should also be given to potential legal issues, ethical review and owner consent as outlined in a previous editorial (Everitt 2013).

Do I Have Access To The Expertise And Equipment Necessary To Test My Hypothesis?

Practitioners should not be discouraged from testing a hypothesis because of a perceived lack of access to specialist facilities. Collaborations between practitioners and academics are always mutually beneficial. In the study by Tabar et al. (2013), for example, real-time PCR facilities were provided by the Clinical Veterinary Hospital at the University of Barcelona. Collaborations can also prove invaluable for journal access during the literature review process and in the design and writing of the manuscript itself.

Do I Have The Means To Present My Clinical Study In An Appropriate Scientific Format?

There are many journals that publish veterinary clinical work. Each journal has a number of formats that can showcase a clinician's work (e.g. case report, clinical paper, research paper, brief communication, what is your diagnosis?). All journals have a set of instructions for authors that explicitly state how the manuscript should be formatted for submission. Potential authors are encouraged to follow the directions carefully and to make full use of the journal's guidelines as well as the numerous writing support systems available online. Before starting to write a manuscript, it is a good idea to read some scientific papers that have been written in the format of the journal that has been selected for submission. Expect to modify a manuscript several times before it is even near to ready for submission. Avoid relying on editorial help from the journal to whip a manuscript into shape as this may not be forthcoming – a badly and hastily written manuscript is more likely to be rejected outright. Here are some general tips for construction of your manuscript:

Title

Each paper should begin with a title that succinctly describes the contents of the manuscript. Bear in mind that the majority of readers will find your paper via electronic database searches and web-based search engines’ key words found in the title. If a title contains insufficient detail, few people will find and access the paper.

Introduction

The introduction must answer the following questions: “What was I studying? Why was it an important question? What did we know about this subject before I completed this study? How will this study advance our knowledge?” It is a good practice to complete this section before designing and completing the clinical study as the process of writing the introduction frequently highlights problems that need to be accounted for when designing the study.

Materials and methods

In this section, how the study was carried out must be clearly explained as if verbally describing the study to a colleague. The broad aim is to provide sufficient detail such that other clinicians could repeat the work and verify the findings. Scientific journals share common guidelines on the use of the past tense, third person and use of active verbs. For example, a statement in the materials and methods section might read, “All samples were submitted for quantitative PCR.” Inclusion of unnecessarily complicated words is actively discouraged. According to research conducted by Oppenheimer (2005), the use of unnecessarily complex vernacular exerts a consistently negative effect on readers. In a series of studies, Oppenheimer examined the complexity of the vocabulary used in various written passages. A group of students were asked to read the samples and rate the intelligence of the person who wrote them. Simpler language resulted in significantly higher ratings of intelligence. Thus, if an author intends to create an impression of lower than average intelligence, inclusion of unnecessarily complex vocabulary is a good way to realize that aim.

Results

The function of the results section is to objectively present your results, without interpretation, in an orderly sequence using a combination of text and illustrations (i.e. tables and figures). Each paragraph should open with a statement that tells the reader which hypothesis is being tested. Summaries of any statistical analyses may appear either in the text or in the relevant tables or figures. Tables and figures should summarize results succinctly, and should not be used for presentation of large amounts of raw data.

Discussion

The function of the discussion is to interpret the results in the context of what was already known about the subject of the investigation, and to explain how our understanding of the problem has been altered after taking the results into consideration. Although the discussion is not intended as a section for restating of the results, an opening paragraph containing a concise summary of the important results is useful. A statement can be made regarding whether the hypotheses were accepted or rejected before discussing the clinical relevance of the findings (i.e. how did the actual results compare with the expected results?).

Citations, reference lists and the review process

It is essential to credit published papers for work mentioned in your manuscript. The format chosen for citations and references depends on the policy of the chosen journal. Plagiarism includes close paraphrasing of other's words, which can be easily avoided by adequate referencing throughout the manuscript. Other considerations before submission for peer-review include construction of an abstract and title, proof reading and writing of a letter to the editor explaining the unique scientific merit of the manuscript. The title and abstract are generally the last sections to be written because they require a completed form of the paper before they can be constructed. It is advisable to ask experienced colleagues to critique your paper before submission. When the paper is reviewed by the journal's panel of experts, you will receive detailed advice on the suitability of the paper for publication, and recommendations on changes that should be made to improve the manuscript. It is never easy when negative comments are received, and the reaction of most authors is to feel affronted by the challenge to their expertise. The best policy is to take a minimum of 1 week to consider the recommendations, and to respond in a polite fashion even when you disagree with the reviewer's opinion. Reviewers give their time for free, and will usually respond much better to a civil well-reasoned argument than to an angry knee-jerk response.

Mike graduated from the Royal Veterinary College (University of London) in 1997. He went on to complete internships at the Universities of Bristol and Edinburgh. Mike developed an early interest in anaesthesia and pain management. He has worked in the USA (University of Minnesota) and Switzerland (University of Zurich) as a veterinary anaesthetist. Mike spent some time working in general practice in the UK and in Australia. He attained the RCVS Certificate in Surgery whilst working at the PDSA. In 2003, Mike joined the University of Glasgow as a resident in Small Animal Surgery (Orthopaedics). He gained his European Diploma in Small Animal Surgery in 2007. He remained at the University of Glasgow as a registrar and lecturer, and recently spent a year working in Sydney as a referral surgeon. Mike joined Fitzpatrick Referrals in November 2010 and maintains a strong ongoing interest in clinical research.

Ancillary