How to write research papers and grants: 2011 Asian Pacific Society for Respirology Annual Scientific Meeting Postgraduate Session
Peter R. Eastwood, Department of Pulmonary Physiology and Sleep Medicine, Sir Charles Gairdner Hospital, Hospital Avenue, Nedlands, WA 6009, Australia. Email: Peter.Eastwood@health.wa.gov.au
This review article summarizes the content of a series of interrelated workshop presentations from the Annual Scientific Meeting of the Asian Pacific Society of Respirology held in Shanghai in November, 2011. The article describes tips and strategies for writing research papers and research grant applications and includes discussion of: the role of pulmonologists in research; the debates around the use of the journal impact factor; tips for writing manuscripts and publishing research in high-impact journals; how journals assess manuscripts and the most common reasons editors reject manuscripts; how to write grant applications and what grant panels look for in successful proposals; and how to undertake research in resource-limited countries.
Y.C. Gary Lee
Medical advances rely heavily, if not solely, on research. In the 21st century, patients and clinicians are still faced with numerous incurable illnesses everyday around the world. Indeed, respiratory illnesses are claiming more lives than ever before. Lung cancer kills more people than any other malignancy,1 and chronic obstructive pulmonary disease (COPD) is soon expected to be the third leading cause of death worldwide.2 Mortality and morbidity from centuries-old problems, such as tuberculosis3 and pleural infection,4 once thought to be ‘under control’, still remain substantial or are even rising. Advances in pulmonology are failing to meet the increasing challenges from the diverse respiratory illnesses.5 Now, more than ever, we need high quality, translatable, patient-oriented respiratory research.
This review article summarizes the content of a series of interrelated workshop presentations from the Annual Scientific Meeting of the Asian Pacific Society of Respirology held in Shanghai in November, 2010. The workshop and the following sections focus on the role of pulmonologists in research—why they should get involved and how they can fund their research and publish their results in the most suitable journal.
RESPIRATORY RESEARCH FOR THE CLINICIAN
Y.C. Gary Lee
Pulmonologists have a critical role and duty in driving respiratory research. Resources for research are limited, and prioritization is essential. There are endless questions on disease aetiology, prevention, management to improve various outcomes (from survival to better quality of life) and cost-effectiveness for every respiratory illness. Clinicians working at the frontline are the logical persons to determine the most pressing areas for research.5
Crucial breakthroughs in medicine have often originated from astute clinical observations. Patterns observed from clinical practice or hypotheses developed from thought-provoking cases have led to many major discoveries. For example, Sir Richard Doll followed up on a survey that revealed lung cancer patients were almost always smokers. He conducted the landmark epidemiology study that linked smoking with lung cancer;6 understanding this relationship has saved millions of lives. Wagner et al. made the connection between asbestos exposure and mesothelioma after describing mesothelioma in 32 asbestos miners in South Africa,7 and Hippocrates' observation that ‘if empyema does not rupture, death follows’ heralded pleural drainage as the key principle of empyema management.8
In the modern healthcare system, clinicians are faced with large clinical loads and often regarded as a ‘technician’ in a production line focussing on fast turnover of patients. In most settings, research is not a performance indicator, and we see a diminishing worldwide interest in research among clinicians. There are many reasons for this including the poorer financial return of academic research compared with clinical practice, the growing debts of doctors at the time of graduation, and the increasingly tough competition for research funding.
However, to engage in research is a pulmonologist's responsibility and privilege. Clinicians gain great pride in their research work which, no matter how big or small, represents a unique contribution to medicine. It is the best way to ensure that clinicians do not become mere ‘factory-line workers’, blindly following diagnostic algorithms and dispensing guidelines day-in and day-out. Most who have published will always remember the pride and satisfaction in seeing their name in print, a feeling that remains with each successive publication. Research still carries significant prestige and provides opportunities to present data at conferences, to build theses for higher degrees, to improve one's curriculum vitae and enhance promotion prospects in teaching medical centres. It satisfies intellectual curiosity and can be fun.
Many pulmonologists cite the lack of time and resource support as prohibitions to research participation. However, there are many ways to engage in research that even busy clinicians can accommodate. For example, recruitment of patients and collection of samples (e.g. blood, sputum, etc) for multi-centre clinical projects can represent extremely useful contributions.
A good clinician will regularly ask ‘what is it?’ and ‘why is it?’. Clinical medicine and academic research are complimentary, not mutually exclusive. Most distinguished clinical researchers are good practitioners motivated by the intellectual pursuit for better understanding of clinical observations. What made this smoker develop lung cancer when most don't? Why did this patient with pneumonia develop empyema when others fully recovered?
The research idea
‘Every battle is won or lost before it was fought’ (Sun Tzu, The Art of War). The novelty and clinical relevance of the idea underpinning the research question are the absolute keys to a high-impact project. Impact in clinical research is gauged by whether (and by what magnitude) the result changes practice. Research must aim to contribute new information to existing literature, and to avoid repetition of prior studies. A thorough literature review, and (if possible) discussions with experts in the field, are prudent. A good project should be contributory to medicine whether the results are positive or negative.
Small projects can make a big impact. Massive funding of trials is of secondary importance to originality and clinical applicability of the research study. A classic example is the study published in the New England Journal of Medicine by Diem et al. which showed that television drama series portray unrealistic and over-optimistic outcomes of cardiopulmonary resuscitation (CPR) and distort public perception.9 To come to this conclusion, the investigators watched hundreds of hours of prime time medical dramas and recorded the survival rates of all CPR episodes—a task many consider ‘entertainment’ rather than hard work. Yet the findings have significant ‘real-world’ implications for clinicians worldwide when discussing ‘do not resuscitate’ orders with patients.
When considering a research idea, the local prevalence of a condition and the accompanying clinical expertise can often provide new investigators an advantage (also see How to do Research in Resource-limited Countries). In developing an idea, it is important to spend time digesting and thinking through the subject and discussing it widely with experienced colleagues. It is essential not to take any criticism personally as it is far better to be aware of any flaws in your research plan before starting. In the end, it is important to ‘just do it!’ and recognize that while no study is perfect, information from every study helps advance knowledge.
Great research ideas are worthless unless they are executed. Despite new investigators often having little funds and support, it is important to recognize that any assistance will be helpful. Registrars, medical students or even friends and patient volunteers may be helpful in this regard. It is worthwhile writing to local charities and to apply for all funding opportunities, big or small, as every dollar helps. One should estimate carefully how much time it will take to complete the study, then double or triple it; projects always take longer than anticipated. Once started, you must see the project to completion. Otherwise, you will exhaust the goodwill of your helpers and sponsors and unlikely find future support.
‘Research not published is as good as research never performed’. The hard work of the project will come to nothing if the results are not disseminated through publication. Negative results are as important as positive findings, in many cases. Findings from small projects may provide a critical piece of the jigsaw to complete the big picture.
How best to write a manuscript is outside the scope of this paper, but has been a topic of many forums (For example, see the Asian Pacific Society of Respirology Research Paper Writing Symposium: http://www.apsresp.org/education/writing-symposium.html).
IMPACT FACTOR—ITS USE, THE DEBATE AND ALTERNATIVES
The idea of a journal impact factor (IF) was first mentioned in Science magazine by Eugene Garfield in 1955,10 although it was not until the early 1960s that it was more formally proposed as a method of selecting and comparing journals regardless of their size and citation frequency.11 Today, the IF is published annually by Thomson Reuters in their Journal Citation Reports and is the citation metric most often used to describe a journal's influence and impact. Despite its limitations,12 it is a useful metric for comparing journals within a given specialization or sub-specialization.13 However, a major concern to all fields of research, including those focused on respiratory medicine, is the growing trend for use of the IF to rate the quality and importance of individual research publications and the academic performance of individual researchers. An understanding of the method of calculation is required in order to understand why such a concern is warranted.
A journal's IF is calculated by dividing the number of citations received in any given year by the number of original articles and review articles published in the preceding 2 years. For example, in 2010 a total of 679 citations were received for all Respirology articles published in 2008 and 2009. Because a total of 364 original articles and reviews were published in these same 2 years, Respirology's 2011 IF was 1.865 (=679/364). This means that, on average, recent articles from Respirology have been cited 1.8 times each.
It is apparent from this method of calculation that the IF of a journal can potentially be influenced by a few articles that attract a disproportionately large number of citations.14 In such a circumstance the overall journal IF will underestimate the impact of these highly cited articles and overestimate the impact of other articles that attracted few or zero citations. Also, a journal can also adopt strategies to increase its IF and as a result lessen the relationship between the journal's IF and the significance of specific articles. Such strategies include: ‘front-loading’ the journal so that most manuscripts appear in the beginning of the year thereby optimizing the time for any given manuscript to be cited; removing case reports; increasing the number of review articles, which are generally cited more often than original articles; decreasing the acceptance rate; preferentially citing other articles from within the same journal; and decreasing the time between acceptance of an article and its publication.
These concerns have resulted in an increasing number of influential groups advocating strongly against the use of the journal IF to assess the quality of an individual's research publications in order to assess their academic performance. Two such groups include: the International Respiratory Journal Editors Roundtable15 who issued a statement proposing that the IF calculated for individual journals should not be used as a basis for evaluating the significance of an individual scientist's past performance or scientific potential; and the Australian National Health and Medical Research Council16 who removed journal IFs from peer review of individual research grant and fellowship applications, noting that the IF of a journal does not describe the impact, importance or quality of any individual manuscript and therefore was not logically sound to equate the ‘impact’ of the journal with the ‘impact’ of each manuscript in that journal.
The development of metrics other than the journal IF has been a response to the inadequacy of the journal IF to assess individual scientific performance.17 Alternative metrics such as the number of citations, number of weighted citations including Thomson's Eigenfactor and Elsevier's SCImago Journal Rank, the h-index and the immediacy index are all aimed at improving the ability to assign quality and standing to an individual publication and individual researcher. Such measures are now easily calculated thanks to modern computing and the availability of online databases such as the Web of Science from Thompson Reuters, Google Scholar and Scopus from Elsevier. This rapidly increasing web-based capacity also enables real-time monitoring of article downloads, another potential metric for assessing an article's impact.18
Ideally, the significance of an individual researcher's work should be assessed by soliciting evaluations of the significance of their work from scientific peers who are carefully selected to be both highly qualified as well as being at ‘arms-length’ from the candidate.15 While possible to apply such a process in the setting of grant review panels, it is much more difficult to do so in other settings that require such an evaluation. In this case, judicious application of these alternative metrics, but not the journal IF, can help provide quantitative assessments of an individual's quality and contribution to their field.
HOW TO PUBLISH YOUR RESEARCH IN A HIGH-IMPACT FACTOR JOURNAL
Everyone who undertakes a research project is keen that others should learn about what they have done, and the best way to do this is by publishing your findings. It is obvious that if you think your work is worthwhile you will wish to publish in a journal which has the highest regard by your colleagues and one which they are most likely to read. Despite its limitations, the system of classifying journals by their IFs (see Impact factor—its use, the debate and alternatives) is now well established, and there are clear winners and losers in this approach. As someone who has published in very high and much lower impact journals, I have learned the hard way what works and what does not.19–24 The following sections briefly review which journals would be perceived as having a high IF, what their strengths and limitations are and how you might organize your work to maximize the chances of a successful publication.
Which are the high-impact factor journals?
Although the winners and losers in any given field fluctuate from year to year, there is some consistency about which journals are perceived as being of high impact. The general medical journals, such as the New England Journal of Medicine, The Lancet, the Journal of the American Medical Association, Annals of Internal Medicine and to a lesser extent the British Medical Journal are all perceived as having wide readership and high impact, both in terms of citations and in changing clinical practice. Within the respiratory field, the leading journal is the American Journal of Respiratory and Critical Care Medicine, the only journal with an IF above 10. In approximate terms, this level of impact is the marker of success in all the leading journals in the specialist fields of internal medicine. Some way behind in our field come Thorax and Chest, both of which have very similar IFs around 6.5, closely followed by the European Respiratory Journal with an IF of 5.9. Respiratory Research which is an online journal has an IF approaching 3 and this is the area that Respirology is now approaching, along with Respiratory Medicine. All of these journals are peer-reviewed, electronically accessed, and their research papers often cited. There are a number of other specialist journals, some also with quite high IFs. For instance, Sleep, the journal of the American Academy of Sleep Medicine, has an IF only slightly below the European Respiratory Journal, while there are other less quoted but nonetheless important journals serving the needs of communities interested in COPD and asthma. There is some overlap in the area of asthma with allergy, and the leading allergy journal, the Journal of Allergy and Clinical Immunology, also has a strong IF but not as high as the American Journal of Respiratory and Critical Care Medicine.
Thus, there is a wide choice of high IF journals to consider when submitting a manuscript. However by their nature, they are very competitive and the acceptance rate varies between 7% and 25%. Indeed, one way for editors to improve the IF of their journal is to be very selective about which papers are published and to try and minimize the number of articles never cited after publication. A consequence of this strategy is that the barrier to publication increases rather than decreases over time.
Which journal should you choose?
While it is an important goal to have a wide audience for your published research, it is also important to be realistic about where you send your manuscript for consideration for publication. It can be very time consuming to format a manuscript for one journal and then have it rejected without external review, as increasingly happens. This is more likely to occur if you send an interesting but specialized manuscript to one of the general journals who are looking for topics of broader interest or major scientific importance. It is always helpful to discuss where you want to publish with an experienced colleague, either before you format the manuscript or at the very least before you submit it. Sometimes you can be lucky, and your subject area becomes very topical, but simply sending everything you write to one of the high IF journals is likely to waste both the journal's time and yours.
What do you need to succeed?
Inevitably, your research idea needs to be a good one—the more original, the better. You should be looking at an important question and have applied good clinical and research methods which you can justify in detail in your manuscript. If yours is a clinical trial, it is a good idea to involve a statistician early in the design of the study rather than as an afterthought when you wonder whether your findings are significant. It is essential that you have a good grasp of the research which has already been done in your area. Acquiring this information is much easier now than in the past decades with the availability of excellent electronic library and search facilities. If you are aiming for one of the high-impact general journals, then it is also helpful to be topical. These journals are businesses as well as scientific enterprises, and they know that topics which catch the attention of their readership, as well as the wider community, are likely to promote their best interests. Of course, everyone always needs a slice of good luck, and this area is no different from any other aspect of professional life.
Preparing your manuscript
The time spent revising and improving your manuscript at the outset is likely to be more than repaid in the long run. There are some obvious things you can do. You should always follow the instructions to authors, which vary from journal to journal. You should not exceed the word count for the abstract or the manuscript—some journals now send manuscripts immediately back to the authors without review if they exceed their stated total word target. Occasionally, editors will allow an exception to these rules, but electronic data upload makes it quite difficult to even get to the stage of having a dialogue with the editorial team. It is important that your paper is presented clearly and logically. It is essential that the reader understands the population you have studied. Your introduction should not be more than four paragraphs long and should contain a clear statement of a hypothesis rather than a general comment about ‘we thought it would be interesting to look at . . .’ Adequate demographic information is needed whether it is about patients or tissue samples, especially in a clinical study. If the study is complex, then be prepared to write at length about the methods, but put this in the online data archive as only the reviewers and those who are really interested in what you have done will look at this. This online supplement is becoming an increasingly important resource which raises the standard of the resulting manuscript. Almost all of the high IF journals discussed above offer the use of an online repository for presentation of supplemental data. The discussion should critically review what you have done and not offer a critique of what others have done. It is essential to cite relevant previous research, but this needs to be placed in the context of what you have found—you should explain what has been achieved with your study that has not been done by others.
Given the low acceptance rate of many of these journals, you need to be tough enough to accept rejection letters—which will undoubtedly come your way. This is always a depressing occurrence no matter how senior you are, especially if the reviewers have missed the point of your paper. However, sometimes this is because you have not made the point clearly enough rather than because they are particularly stupid. You may wish to challenge the rejection and resubmit the manuscript with a detailed rebuttal of the criticisms, but the chances of success are low especially when the covering letter states that they ‘do not want a resubmission and the decision is final’. The urge to immediately send the manuscript to the next journal on your target list is great, but do not do this. At least put it to one side for a day or so and then think again about what the reviewers have said and see if it is possible to improve the paper. Most papers can be improved in the light of sensible critical comments.
HOW JOURNALS ASSESS MANUSCRIPTS
Most journals provide clear guidelines to reviewers on how to assess manuscripts submitted for publication. Respirology, the official journal of the Asian Pacific Society of Respirology, asks reviewers to evaluate the quality of the enclosed manuscript according to pre-defined criteria as outlined in Table 1. Using international standards for respiratory publications as the benchmark, reviewers are then asked how they would rate the manuscript, both as presented and if revised in accordance with the reviewers' comments (Table 1). Reviewers are then asked to make a recommendation in terms of acceptance, and to provide confidential comments to the Associate Editor and comments that will be sent to the author. Consideration of the method of assessment by referees provides guidance to authors on the key features of the manuscript that need to be reviewed prior to submission.
Respirology—Official Journal of the Asian Pacific Society of Respirology—Reviewer Assessment Form
THE MOST COMMON REASONS EDITORS REJECT A MANUSCRIPT
Rejection of a scientific paper by a journal should not be the end of the world. Two major treatment discoveries, that of scurvy with vitamin C25 and peptic ulcer disease with antibiotics, were characterized by the triad of initial scientific rejection, then ridicule followed by acceptance (to paraphrase Michel Eyquem de Montnigne). Although these examples are uncommon, an author's passion and belief in a scientific result, idea or observation will drive one to seek publication. Rejection of a manuscript is experienced by all accomplished researchers at some stage of their career. Each will have a tale to tell of the (in)justice of the manuscript selection process for publication in high quality scientific journals. With this in mind, however, the common reasons that journals reject scientific manuscripts relate to poor scientific quality, incorrect submission format or absence of common ‘overlap’ of manuscript and journal interest.
High scientific quality can be achieved through various means. Choose a field of research which is of genuine interest to you. Read quality articles around this area. Have discussions with several trusted and experienced mentors. Identify an original idea then consider a tight study design with adequate controls. Affiliate yourself with a statistician early on and, if required, register the trial with a clinical trials registry prior to initiation of the project.
Journals do not like submitted articles that are not formatted correctly. Attention should be given to the ‘authors instructions’ available on all journal web sites, such that the submitted manuscript design should equate with that of the journal. Avoid plagiarism, ensure grammatical fluency, include a statement of study limitations, provide a contemporary reference list (and include some related to the journal) and make sure your data are accurate and are reproduced accurately in abstract, text, tables and figures. A well written manuscript is brief and has every word accounted for. A classic example of this is the manuscript describing DNA by Watson and Crick—only two pages in length with six references.26 More effort is required to write a brief and concise article than a long manuscript. Have an external person read the manuscript to see whether it flows and is understandable.
Manuscripts with excessive numbers of authors can attract greater scrutiny. Each author should have participated sufficiently in the work, in the data analysis and in the preparation of the manuscript, and have reviewed and approved the manuscript as submitted to take public responsibility for it. The contribution of co-workers not involved in all of these steps can be recognized in the acknowledgements section—this is often more acceptable than long authorship lists.
The letter of introduction to the journal editors must be brief and succinct and indicate an overlap of interest between your study and that of the journal. Most journals will send the manuscript to a subeditor knowledgeable in the field of your research. If deemed appropriate, the manuscript will be sent to several external reviewers. Journals have trouble locating good reviewers.27 When submitting an article, it is helpful to provide a list of three to four potential reviewers. It is likely that they will be used, plus one to two extra reviewers. Suggest yourself to the journal editorial team as a person interested in reviewing and learn from the process. Editors do not look favourably upon submitting authors who in the past have refused to review for the journal.
Although acceptance rates are often quite low in the top respiratory journals (<25%), acceptance will eventually occur if the paper is of sufficiently high quality. If two of three reviewers' comments are supportive and a third is unfairly negative, and your manuscript is rejected, you can ask the journal editors for a reassessment. Perhaps, suggest to the journal that it consider an editorial on the topic in dispute to provide the counter argument.
Following a rejection from Journal A, if one is to resubmit to Journal B (usually of a lower ranking based upon IF), consider including Journal A's reviewers' comments and your responses (literally ‘Comment 1, 2, 3 . . . and Response 1, 2, 3 . . .’) in the resubmitted covering letter to Journal B. Incorporate the review comments, if reasonable, into the new manuscript for resubmission.
HOW TO DO RESEARCH IN RESOURCE-LIMITED COUNTRIES
The capacity to undertake research in resource-limited countries such as China, India and developing regions in Africa, are restricted more by fiscal input and availability of sophisticated facilities, rather than by resources such as disease presence, clinical data and human intelligence. In many countries, these latter resources exist in abundance at the local level and, if utilized wisely, can provide unique advantages to undertaking scientific research relative to resource rich countries. For example, Chinese clinicians are uniquely placed to design studies with an emphasis on translational medicine. In regard to undertaking translational research in ‘resource-limited’ countries, the routes of research translation could be considered to be either: from bench to bedside; from empirical to evidence based; or from bedside to bench and back to bedside.
From bench to bedside
A vital role of laboratory-based, clinical scientists is to translate basic research findings into simpler, cheaper and more effective treatments and/or measures with less adverse effects for purposes of clinical diagnosis and management. A good example of such a translational story can be found in the early studies on thiol compounds (e.g. carbocisteine), which had been shown in a few studies to be selective scavengers of reactive oxygen species28 with useful antioxidative properties.29,30 Such a finding suggested that carbocisteine could prove to be significant for the management of COPD in which oxidative stress contributes substantially to airway inflammation.31 These preliminary results lead to a multi-centre randomized controlled trial (the PEACE Study) which showed that, compared with placebo, carbocisteine was associated with a 24.5% reduction in exacerbation of COPD, improved quality of life and no adverse effects.32 The reduction in COPD exacerbation with carbocisteine compared favourably with inhaled corticosteroids plus long-acting β-agonists or Tiotropine, and was more affordable (only one sixth the costs for standard treatments) and simple to use (oral administration).32 The study represented a ‘cost-effective and well-tolerated way of lessening the burden of exacerbations for the growing number of patients in developing countries affected by COPD’.33 As such, these series of studies represent a useful paradigm for effective translational scientific research in resource limited countries.
From empirical to evidence based
Local traditional medicine, such as Chinese, Indian-Ayurvedic, Mongolian and Tibetan, represents a potentially enormous resource for improving human health. Practice of traditional medicine has been refined over many centuries and is largely empirical in nature. This poses a challenge to Western medicine which demands pooled data from large-scale clinical studies in order to provide convincing evidence of bedside effectiveness. Shifting local traditional medicine from empirically based to evidence based is a potentially important role of translational medicine for clinical scientists in resource-poor countries. Examples of this can be found in several successful clinical studies undertaken to evaluate the therapeutic efficacy of Chinese herbs.
For example, influenza was commonly treated with certain herbal formulas, such as weeping forsythia, heartleaf houttuynia herb and honeysuckle flower. During the 2009–2010 outbreak of swine flu, a randomized controlled trial of some of these natural herbal medicines (in the form of a commercially available product, Lianhua Qingwen capsule (LHC); Shijiazhuang, Hebei Province, China) was conducted in 244 patients with oseltamivir as the active control.34 Compared with the control, LHC yielded similar outcomes in symptom relief and reducing viral loads, demonstrating itself to be a potent alternative to oseltamivir against H1N1 infection. It is highly likely that the translation of empirical-based local traditional medicine to evidence-based platform has much to offer future human health.
From bedside to bench to bedside
From a clinical perspective, laboratory research in resource-limited countries needs to be focused on the most topically relevant health issues with the greatest clinical relevance. Such an approach ensures that subsequent findings can be used immediately in order to make a difference in clinical practice. Two examples of such a strategy proving to be effective include the use of Artemisinin in the management of malaria and the use of Arsenic in the management of leukaemia.
In the case of Artemisinin, a crude extract of Artemisia annua has been used in China over the past millennium for the treatment of malaria. In 1971, these anti-malarial effects led to the laboratory-based discovery of Artemisinin, a purer extract with strong actions against plasmodium,35 and in 1973, the discovery of a reduced type of Artemisinin, 10 times more powerful than the original compound.36 The chemotherapeutic potential of Artemisinin as an anti-malarial combination therapy has been well-documented,37 and its use has recently shown to significantly contribute to the rapid elimination of malaria on Moheli Island of Comoros.38
In the case of arsenic, in 1971, based on a local empiric remedy, Chinese pharmacists developed a preparation containing arsenic trioxide and mercury chloride to treat acute myeloid leukaemia. Subsequent studies using this preparation have shown improvements in remission rates,39 and that, when used together with trans-retinoic acid, can produce a >90% 5-year survival in acute promyelocytic leukaemia.40 This is a powerful demonstration of how translation from bedside to bench and then back to bedside has changed acute promyelocytic leukaemia from being highly fatal to highly curable.41
Research in resource limited countries can result in world-class findings, particularly when clinical scientists keenly take advantage of local circumstances and carefully identify existing unmet clinical needs.
TIPS ON SUCCESSFUL GRANT WRITING
There are two key questions that any experienced reviewer and/or grant review panel member will ask about your grant:
- 1Is the work important?
- 2Can they deliver (have they delivered before and do they have a good approach)?
Because consideration and discussion of a research grant application is focused around these two questions, it is essential that it is written with these questions foremost in your mind.
Specific tips for writing a grant
- 1Be clear. Be clear in your own mind first. If you cannot explain your grant to someone in one or two sentences, then the assessors will also be guessing.
- 2Make it easy on the assessors. At a minimum, the grant should be clear, not too ‘text-dense’, well broken up with clear headings and, if possible, it should even be enjoyable to read. One way to make it easy is to have a ‘punchy’ opening sentence e.g. ‘z is a key question. We have found x and y with your money so far, have written 50 high impact papers and we are now poised to find z’.
- 3How to start. Introduce the project based on what is known already, what is not known and what is the GAP in knowledge that your work will answer? If the assessors don't perceive that you are answering a gap in knowledge, they won't fund your study.
- 4Why has it not already been done? What do you bring to the table that will enable you to succeed where others cannot or have not for example, special patient populations, new reagents, new models, etc.
- 5Use cartoons if possible. This is especially important if the project is complicated. Find someone who can draw a simple yet clear cartoon of what you are proposing and how each aim fits into it. These cartoons are similar to the ones you see in review articles.
- 6State each of your aims clearly. It is essential that each aim is logical, well defined, reasonable and attainable.
- 7Your methods must address the aims and hypotheses. Too many applicants describe general methods but don't articulate clearly how their method will address each specific aim. When assessing the results of these methods, it is important to use outcomes that are easily measurable. Also, make sure your statistics section is strong.
- 8State preliminary data clearly. In doing this, it is important to focus on key experiments rather than an avalanche of figures.
- 9Read the instructions for applicants. This is especially important when considering how your application will be assessed. For example, if the instructions state that 25% of the total score will be awarded for significance, training of others, strength of collaboration, translatability, etc., it is important that you write a clear statement of these aspects of your work, rather than a few throw-away sentences. It is also important to understand how funding bodies' stated priority funding areas relate to your grant.
- 10Don't hold back on comments about innovativeness. You will need to tell the assessors about this—do not assume that this is always obvious to the assessors, even though you think everyone knows that you were the first to publish a novel finding.
- 11Be clear and honest about limitations and anticipated difficulties. These are not always obvious to assessors. Tell them that you recognize the potential pitfalls and that you have a way around them. This approach also avoids getting the ‘obvious’ questions from the assessor. In other words, do their thinking for them and anticipate their questions.
- 12Involve collaborators early, not at the last minute. When one has agreed to collaborate, it is annoying to be handed the grant the night before the deadline. In this circumstance it is too late to provide useful input.
Potential mistakes in grant writing
- 1Don't waste your time on impossible grants. You need to count the cost of your time input versus what your chances are of being awarded the grant. You might be better off performing another set of experiments than wasting your time on a ‘shotgun approach’ (applying for every grant that is remotely possible to get).
- 2Don't falsely reduce your budget because you think you stand a better chance of getting funded, unless you definitely have the rest of the money somewhere else. If you try to do a full grant on half the money, it will get too hard.
- 3Don't write your grant in the form of ‘I need money’ and don't ‘tug at the heartstrings’. As hard as your life as a scientist might be, the assessors are morally obliged to strictly follow the assessment guidelines as they relate to the quality of the work, your track record, etc.
- 4Don't claim you will find ‘the silver bullet’. Research is hard; the assessors know that, so if you claim that you will cure cancer they are unlikely to fund you.
- 5Don't take knock-backs personally (easier said than done)—and don't give up.
WHAT GRANT PANELS LOOK FOR IN SUCCESSFUL PROPOSALS
The most important first step in drafting a grant application is to carefully review the criteria and guidelines of the granting organization. The grant application must be structured and written in exact accordance with the guideline recommendations. Failure to do so will usually result in an unsuccessful application.
The next priority is to ensure that the research being proposed has a strong hypothesis and the potential to lead to a substantive advance in knowledge, clinical practice and/or improvements in health outcomes. This requires a wide knowledge of the field and preferably previous work, which has provided data to support the hypothesis, to demonstrate feasibility and thereby to form the basis of the proposal. The National Institutes of Health (NIH) has published practical guidelines on how to write a research project grant application.42 It is worthwhile reading this document carefully before drafting the application, as it provides excellent practical tips on the requirements of a successful grant. It proposes that the reviewers are likely to ask eight basic questions (Table 2).
Table 2. Eight basic questions reviewers ask†
|1. How high are the intellectual quality and merit of the study?|
|2. What is its potential impact?|
|3. How novel is the proposal? If not novel, to what extent does potential impact overcome this lack? Is the research likely to produce new data and concepts or confirm existing hypotheses?|
|4. Is the hypothesis valid and have you presented evidence supporting it?|
|5. Are the aims logical?|
|6. Are the procedures appropriate, adequate and feasible for the research?|
|7. Are the investigators qualified? Have they shown competence, credentials and experience?|
|8. Are the facilities adequate and the environment conducive to the research?|
For research ranging from fundamental basic science to randomized controlled clinical trials to public health, the structure of the grant application is likely to be similar. There are usually four basic sections, namely the research background and rationale, the design and methods, the budget, and the track record of the research team.
The rationale provides the research background and statement of the purpose or aims for scientific enquiry, the hypothesis and the potential for new knowledge. It is necessary to persuade the reviewers that the research is a priority, is innovative and has the potential to fill a gap in current knowledge. The anticipated outcomes should demonstrate the potential impact on knowledge, clinical practice, patient outcomes, commercialization or policy.
The research design and methods need to provide sufficient details for technical assessment of the scientific protocol, its feasibility and likely validity of the data. In the case of randomized controlled trials, it is crucial that information for all items in the methods section of the Consort Statement are provided.43 It is essential that a biostatistician is consulted early in the planning of the study. Consideration should be given to obtaining ethical approval prior to the grant's submission as it will enhance the assessment of feasibility.
Drafting the budget requires attention to detail to ensure that it is accurate and justified. It is important that the budget is not underestimated as insufficient funds will place the research at risk; budgets that are overestimated may be considered poor value for money.
The track record provides evidence that the team has the experience, qualifications and infrastructure to deliver the research. Consideration should be given to collaborators who can enhance the project, providing the skills and experience in areas not adequately covered by the research team.
Successful grant applications take time to draft and benefit from external review prior to submission. Before submitting, it is worthwhile considering the most common reasons why projects do not get funded, as advised by the NIH.42 This advice provides a checklist by which the final revision of the grant application can be made (Table 3).
Table 3. Most common reasons for a low score (in priority order)†
|1. Lack of new or original ideas|
|2. Hypothesis ill-defined, superficial, lacking, unfocused or unsupported by preliminary data|
|3. Methods unsuitable or defective and not likely to yield results|
|4. Data collection confused in design, inappropriate instrumentation, poor timing or conditions|
|5. Data management and analysis vague, unsophisticated.|
|6. Inadequate expertise or knowledge of field for principal investigator, or too little time to devote to the work|
|7. Poor resources or facilities; limited access to appropriate patient population|
Following submission, the grant application will be sent out for external review, and the applicants will usually be given the opportunity to respond to the reviewers' comments. In the rebuttal, it is important to address the comments point-by-point. If you do not agree with the issues raised you should explain why, with sound evidence-based arguments. If you do agree, acknowledge this and consider whether the proposal can be modified to address the issues raised. The rebuttal is crucial in persuading the assessing panel of the importance and feasibility of your proposed research. If the research proposal is not funded at the first attempt, careful consideration should be given to whether the application should be revised and resubmitted, or whether it is preferable to begin over with a new idea (Table 4).
Table 4. Common fixable and non-fixable problems†
|Common fixable problems:|
| • Poor writing|
| • Insufficient information, experimental details, or preliminary data|
| • Significance not convincingly stated|
| • Approach not shown to be feasible, but applicant can demonstrate feasibility|
| • Insufficient discussion of obstacles and alternative approaches|
|Not fixable or more difficult problems:|
| • Philosophical issues, e.g. the reviewers believe the work is not significant|
| • Hypothesis is not sound or not supported by data presented|
| • Work has already been done|
| • Methods proposed were not suitable for testing the hypothesis|