[MUSIC] Hello, this is Janet Holbrook. I am a professor of epidemiology at the Johns Hopkins Center for Clinical Trials and today I'm going to be talking about reporting results from randomized clinical trials. I'm going to be talking about two types of reports of results from randomized clinical trials. One is peer reviewed journals, something that most of us are probably quite familiar with when the report from a trial is detailed in a journal, the new England Journal of Medicine. But secondly, I'm going to be talking about data monitoring reports, and data monitoring reports are very important in clinical trials. But they're rarely seen because they're confidential for the most part and not talked about very often. I want to remind everyone that there is an ethical imperative for reporting the results from a clinical trial. We've done an experiment in humans many times with public funding and the people who joined that experiment to be participants have the expectation that we're going to use the data that we collected on them to advance medical care. So, reporting the results of the trials is an important piece of ensuring that the knowledge is shared. Secondly, there is interim monitoring, I told you I was going to talk about data monitoring reports. When we conduct experiments in humans, we can't just flip the coin and walk away, even though we may randomize someone and treat them for only six months. It may take four or five years to actually complete that trial and there is a responsibility to ensure that we're not placing future participants at greater harm than we expected when we designed the trial. So, we have to be looking at the data as it accumulates, and I'll talk about that more in the section on data monitoring reports. I'm not going to talk about regulatory reports, reports that you might submit to the FDA either with the results of the trial or with safety events, serious adverse events that occur during a trial. I'm also not going to be talking about secondary analyses of clinical trials. I'm going to focus on the analysis and reporting of the primary results. I'm not going to be reporting about systematic reviews which gathered together many clinical trials and have a consolidated reporting of the results. And finally, this is not a primer on the use of English for scientific writing. I have some recommendations that I have found useful over the years, but I am not an English instructor. So my perspective is one of someone who has been doing multicenter clinical trials. So these are usually large trials that can range anywhere from a hundred to thousands of people. The interventions that I have experience with and that I'm relying on to talk to you were interventions for health for individuals with eye disease or airway disease. Mostly, youitis for eye disease for airway disease, trials of treatments for asthma or COPD or chronic cough. I'm going to be talking based on my experience, writing journal articles, preparing data monitoring reports, and I'm going to use several examples from studies that I've worked on. So I'm going to start off with peer reviewed journal articles. So what should go into a peer reviewed journal article? The four basic elements that go into a peer reviewed journal article are the background. Why did you decide to do this trial, what was known, why was another study needed the methods? What did you do exactly detailing who was enrolled, what they received and what you measured and how you measured it? The results, what did you find? And of course the discussion kind of integrating it into the scientific evidence and what do these results add, and what do they mean? One of the first decisions that you need to make when embarking on writing up the results from a randomized clinical trial is selecting the journal that may seem counterintuitive that you select where you're going to publish it right up front. But that decision can be very influential on the manuscript because different journals have different formats about how they want their abstracts, how many words are allowed? The number of tables and figures, so all of that is important when you're putting your manuscript together. You also need to be thinking about what your target audience is. Is this a paper that you think will be of widespread interest because it's a very common disease or it's a commonly used intervention or you've had some surprising results? Or do you think it's a paper that may be more applicable to a specialty journal for ophthalmologist treating eye disease or specifically for pulmonologist. And you can get a sense of that by looking at where other studies like yours have been published. I would encourage you to aim high because that will give you the best largest audience for your paper. However the process of submitting an article can be quite long, of course, there's a long process of writing it but then there can be a long process for reviewing it and even when the article ends up being rejected. So, some things to think about when thinking about which journal is, what's the impact factor for that journal? Is there an editorial pre-submission review process, can you write to the editor, tell them the basic outline of your study and they may give you a yea or nay rate at the start. And this is more common in some of the highest rated journals like New England Journal of Medicine. Also don't be discouraged when you receive a rejection notice. One thing you get a lot of good feedback from the reviewers that did review the paper and you'll be able to improve your paper. And it's normal, most papers get rejected on the first go around and it often takes three submissions to get a paper published. Nowadays, there are two models for journals. There's a subscription journal which is the traditional model which you either have to have a personal subscription to access. Or you can access it through an institutional subscription or there are open access models. The open access models are open to anyone who wants to read them. They're generally published on the internet, but there are some cost involved with those. You usually have to pay maybe 1 to $3,000 to get something published in an open access journal. And finally you need to be careful to avoid predatory journals. These are journals that are taking advantage of the open access mode of publications. Generally they have very low or inconsistent publication standards, they very aggressively solicit papers because this is a money making enterprise. Often they have false or misleading information about their journal about the editors of their journals if you have any suspicion or you're really unfamiliar with the journal. It is good to check out their websites, check out to see if their editors are real people and they can be quite costly to publish. Also, you should read articles in the journal and see what the quality of the publications are. And that will give you a good sense of whether this is a predatory journal or a journal whose mission is to publish evidence. So I'm going to start with talking about CONSORT statements. CONSORT stands for consolidated standards of reporting trials. And it encompasses a number of initiatives that were developed by the consort group which is a group of clinical trialist epidemiologist methodology. That first and journal editors that first organized in 1996. In response to the low quality of clinical trial publications, they released their first set of guidelines in 2001. And have been releasing new guidelines, either updates on those or expansions of those guidelines ever since. And it is part of a broader effort to improve reporting of all types of health research. So they give you a checklist of things that should be included in your title, introduction methods, results, discussions, the different sections of the paper. And often when you submit a paper you will have to submit a CONSORT checklist and actually specify on what page did you report the information. There are additional CONSORT statements for different designs such as cluster design trials or factorial trials. And an integral part of the CONSORT guidelines is to have a flow diagram commonly referred to as a CONSORT diagram that I will go over in a few slides. So this gives you an idea of some of the different extensions to the CONSORT statement. On the right-hand side, you can see a box that lists all of the different types of extensions to consort. They have ones based on different types of design of studies. As I mentioned, cluster trials to adaptive designs for different interventions because as we've expanded the use of clinical trials. There have been different types of things we've been trying to test and so they may require a different perspective on reporting or different items to be reported. And they also have some on data reporting. CONSORT really is part of a larger group which is called the Equator Network which has the goal of enhancing the quality and transparency of health research. And if you visit the equator site, you'll see guidelines for several different types of studies. Including, observational studies, systematic reviews, how to write a study protocol, there's several is worthwhile visiting. So I promised I mentioned what impact factors were and they are basically a measure of the frequency of citations for an average article published in a particular journal. So in 1975, the journal citation reports started keeping track of how often different articles were sighted in subsequent publications. Indicating that your research was important because it was cited by others. It's important to remember that this is a descriptor for the journal and not for any particular one paper and that may be clear by seeing how it's calculated as a moving average. So what they do is count the number of citations of a particular article, let's say one that was published in 2021, so they will count the citations of that article in 2022. And 2023 and then they'll take that as the numerator and the denominator of this ratio is the total number of all site double articles published In that journal in 2021. So you can see it's not specific to an article, it's specific to how many times articles in a journal has been published. Now I think impact factors are calculated by another group and now they are called site scores and they use four-year windows. So for an article published In 2021, they would look at citations through 2026 of that to calculate the ratio. Higher scores are better, but the scores are not very high in general, the top 10% of medical journals have scores that are just above one. Now, some of the most well known journals have pretty high scores like land set has 91. New England journal is about 81 and JAMA is about 25. But there are lots of good journals with factors of around five or less. And you have to remember that part of this is how widely disseminated the journal is. So if it's a journal that targets a specific group or country or language, it may have a lower impact score just because it is very targeted. Another thing that comes up quite often when you talk to students about writing papers is whether you should write in the passive voice. There's a myth that science should always be written in the passive voice. So first let's understand what's the difference between active and passive. So active is when the subject of the sentence does the action. So here we enrolled individuals 18 to 65. So we, the researchers, enrolled the individuals. In the passive voice, something is done to the subject of the sentence. So in this case, the individuals of this age group were enrolled. In general, the passive voice emphasizes the action rather than who was performing the action. And sometimes that seems like a good perspective for scientists because we want to get out of the way of our information and let the information tell the story. It's also sometimes a good technique to reduce the number of words. So when you're really trying to get the abstract down to that 250 words, you might consider sometimes using the passive voice. For results, the active voice is generally preferable, but it depends and you have to think about what is the subject of the sentence, in that case as well. So here's three versions of the same result. The treatment effect was similar in men and women, so we're emphasizing the treatment effect. Gender did not significantly influence the treatment effect. Or we did not find an effect of gender on the treatment effect where the active subject is the researcher. Well, in my estimation, the most clear sentence is the treatment effect was similar in men and women. And I would definitely not want to use the last one. However, I might use the middle one if I was emphasizing the effects of gender. However I might use the middle one if I was interested in the influence of gender on treatment effects. Some tips for writing articles. Keep it clear and simple. It is a common pitfall that people want to sound scientific, they want to sound smart. Use multi syllable words and use too many words. Saying things like a majority of people instead of saying most people, at the present time instead of saying now. So avoid run on sentences and remember to keep it simple. And sometimes the hardest thing about writing a paper is getting started. So you have to allot yourself sometime and give yourself a couple hours every week, it takes time to put things together. Prepare the drafts of the tables and figures first. It's good to have the tables all laid out and the figures at least a draft version of them prepared. The methods usually are pretty easy to write. You have a protocol, what you need to do is to condense that information and there's always online supplements and often you are required to post the protocol there anyway. The results, again, you don't have to be creative. You have to talk about what's in those tables and figures. So all that is pretty straightforward. Where it gets a little bit more complicated is the discussion, what is the meaning of your results? And the introduction as well. You can start with the protocol, with the reasoning for doing the protocol. But you may need to update it with new information that has come out while you were completing your study. And at the end is when you should think about what's the best title for this and writing the abstract. Because you really need to know all of the content of the paper and what are the most important points before you write the abstract. Don't burden your text with a lot of numbers, refer to your tables. Their tools, they are the best way to present your data, so rely on them. Keep your references organized as you go along. Editing is really very important, that's why you have co authors. And you really have to be merciless with yourself. And even though you love that sentence, you thought it was crafted perfectly, when it starts to not feel right is when you should cut it out. Remember that the tables and figures that you prepare are pretty much going to look like what you submitted. There are no magic editors making your tables look beautiful, so you have to put the effort in up front. And also, as I mentioned, online supplements are being used more frequently. And they are quite useful for getting information about the methods or maybe even some of the analysis that you just can't fit in the paper. Good luck. The hardest part about writing a paper is getting started. [MUSIC]
