Negative Results in Neonatal Trials: Clinical Lessons and Future Directions – A Narrative Review

Community engagement is crucial to ongoing development and testing of sorely needed new biomedical HIV prevention technologies. Yet, negative trial results raise significant challenges for community engagement in HIV prevention trials, including the early termination of the Cellulose Sulfate microbicide trial and two Phase IIb HIV vaccine trials (STEP and Phambili). The present study aimed to explore the perspectives and experiences of civil society organization (CSO) representatives regarding negative HIV prevention trial results and perceived implications for future trials. We conducted in-depth interviews with 14 respondents from a broad range of South African and international CSOs, and analyzed data using thematic analysis. CSO representatives reported disappointment in response to negative trial results, but acknowledged such outcomes as inherent to clinical research. Respondents indicated that in theory negative trial results seem likely to impact on willingness to participate in future trials, but that in practice people in South Africa have continued to volunteer. Negative trial results were described as having contributed to improving ethical standards, and to a re-evaluation of the scientific agenda. Such negative results were identified as potentially impacting on funding for trials and engagement activities. Our findings indicate that trial closures may be used constructively to support opportunities for reflection and renewed vigilance in strategies for stakeholder engagement, communicating trial outcomes, and building research literacy among communities; however, these strategies require sustained resources for community engagement and capacity-building.

Standardizing Safety Assessment and Reporting for Neonatal Clinical Trials

This May, Guest Editors Marc Humbert, MD, PhD, Director of the French Pulmonary Hypertension Reference Center, and Professor of Medicine at Paris-Saclay University, Le Kremlin-Bicêtre, France, and Mark Nicolls, MD, Professor of Medicine, Pulmonary, Allergy, and Critical Care Medicine at Stanford University School of Medicine in Stanford, California, held a discussion with Norman Stockbridge, MD, PhD, Director of the Division of Cardiology and Nephrology in the Office of Cardiology, Hematology, Endocrinology, and Nephrology at the Center for Drug Evaluation and Research, US Food and Drug Administration in Silver Spring, Maryland, and Roham Zamanian, MD, Associate Professor of Medicine and Director, Adult Pulmonary Hypertension Program at Stanford University in Stanford, California, on the topic of clinical trials, drug development, and publishing negative trial results.Dr Nicolls: Thank you all for being here today. Let me open the discussion. There have been several negative clinical trials that have been performed in the pulmonary arterial hypertension (PAH) space. I would like Roham to comment on how important he thinks it is to publish the results of negative trials. What can we learn from negative trials, and do you have any ideas about how we might develop leverage for getting the data out?Dr Zamanian: I think we should start by recognizing that we are discussing a rare disease field with a number of stakeholders: patients, patient advocates, clinicians, scientists, academicians, industry, regulators, clinical trialists, and even policy makers. Obviously, the concern is that there is a general bias against “negative” clinical trials, and those biases may be different across stakeholders and also impact them differently. I think maybe this issue is much more pronounced in a rare disease field where there is a limited number of subjects to study and tremendous competition for funding studies which are often costly. The bias not to publish negative trials is a tremendous loss for the community, especially with pulmonary hypertension (PH), where there are so few subjects and they're very expensive to study in an efficient and appropriate manner.The other thing I want to say, and probably this is not groundbreaking, is that a study can be deemed a negative trial when it either has constraints in its design or its implementation, or it does not appear to show what the investigators had hoped and had hypothesized that it would. Therefore, it's brushed under the table, especially if reporting of the trial negatively impacts the financial wellbeing of a sponsor or the reputation of investigators.Dr Nicolls: Can you give some examples of things that we can learn from negative trials beyond the fact that the drug in question didn't reach its primary endpoint?Dr Zamanian: I think we could learn either that a mechanism is not relevant to the disease, or that the disease doesn't respond in the way that we thought it would. More importantly, we can learn a lot about trial design, from selection of endpoints, use of specific enrichment strategies, to trial conduct and challenges with subpopulation recruitment. All trials, positive or negative, can be tremendously informative. To be able to perform a “postmortem” as to why a study is negative is crucial, and it could move the field forward, so we need to overcome the misconception that a negative trial is a “failed” experience.Dr Nicolls: Dr Stockbridge, maybe I can ask you to jump in on this point. I don't know if you have an opinion on publishing negative trials from your vantage point; it's not really what you're most involved with, but do you have any comments on anything Roham said?Dr Stockbridge: We're talking here about nontrivial trials. I don't think that anybody cares quite so much about the first in-man and Phase 2 exploratory studies, but trials that actually had a reasonable shot at establishing effectiveness. It's important to note that a healthy clinical trial therapeutic area development ecosystem has to have trials that fail. If you win on every trial, if that's all that ever happened, then you're not doing enough trials. It's an important aspect that things do fail, and people need to understand something about what happens there.One aspect of this is whether the investigators going into a trial have negotiated the freedom to publish regardless of the outcome. I urge clinical trialists to make sure that they retain the right to publish. You don't want the company constraining that; even if they have some input on it, they should not constrain it.The other aspect of this is that I think patients who invest their time and energy and risk in participating in these trials don't generally think they're doing it to support some company. They generally think they're doing it to support medical knowledge in some area. If we don't publish those studies and make use of what they've invested, I don't think we're doing what we're supposed to with respect to them.Dr Nicolls: That's a great point. Building on that question, can you use your imagination to think of any kind of leverage that might be exerted in the future to compel companies to release the information? Roham and I are actually involved in a situation like this, so we would love to publish the results of our negative trial and dig into the data. I think Roham was interested in the subject even prior to that experience, so it's not just because of that. What kind of leverage or language would be helpful? I think sometimes the problem might be that the company is not releasing the information because they believe that it will in some way be bad for their stockholders to have a negative trial published.Dr Stockbridge: I think it's too late when the trial's over. After you sign the contract, it's too late. The place to be working on this kind of problem is setting up groups that have a strong enough presence in the field that companies are likely to come to them to get a trial done. Then you're in the driver's seat. You have some ability, at that point, to negotiate how the contract is written, to ensure that you've got access to the data, and some privileges with regard to publishing no matter how it comes out. If you wait until it's over, you're going to have a hard time.Dr Nicolls: Because not everyone is in the luxurious position of having that kind of international reputation, and sometimes these companies are one-off companies that may not come back with another PH drug, I think something that anybody could do, any investigator-initiated proposal, would be to start at the contracting process before the trial. I think that could be a good takeaway point from this discussion.Dr Zamanian: At the World Symposium on PH in 2018 in Nice, France, Dr Stockbridge and I and others, including Lewis Rubin, were on the same task force, and from that section came the recommendation that you develop a process for reporting negative clinical trials, you make it a contractual obligation up front and utilize the leverage of international experts who carry a lot of weight in that process. The second part to that is that you may also need to develop an independent and self-funded core, whether that resides as part of the data safety monitoring board or the steering/adjudication committee or an analytic core committee that would carry out and report on analyses, so that you're not depending on the industry for providing statistical support. Tie that in with some sort of a mandatory publication timeline which balances the need to report on a study with the industry need to absorb the findings—I would suggest something like a publication embargo period of maybe 6 months so that the impact of the data on the company may be minimized. Really, the stockholders will know by 6 months after the study whether the study was negative or positive.Dr Nicolls: That's great. I think maybe we should move onto another subject. I'll invite Marc Humbert to introduce any topics he'd like to discuss.Dr Humbert: I wanted to focus a little bit on Phase 2 studies in PAH because we usually have different endpoints in Phase 2 and Phase 3 clinical trials, and the negative trials give us different messages. I was thinking that the relatively small Phase 2 studies, in which primary endpoints are usually pulmonary vascular resistance (PVR) measurements obtained by invasive right-heart catheterization, should always be presented in great detail with careful presentation of individual data, not only because we want to have a complete analysis of the active compound results, but also because of the importance of the results observed in the placebo arm, or the control group. I wonder whether either Dr Zamanian or Dr Stockbridge would like to comment on the possibility of having a single control group for several active drugs tested in order to expose fewer patients to a placebo, especially when data obtained by right-heart catheterization are needed.Dr Zamanian: I don't know if I have an answer, and I would love to hear Dr Stockbridge's opinion on this. Several of us have discussed the idea of using a master protocol in early phase development as a platform for what you've just said, Marc. The idea is to test multiple drugs more efficiently and allow for a single placebo arm that would be used for a single control arm and others which are the intervention/therapeutic arms. This would allow a more efficient drug development program and allow for some degree of adaptation, utilizing biomarkers or maybe some phenotype information.The difficulty with the idea of the master protocol is not, I think, a problem for academicians. I think when it comes to the biopharma companies, they may view it as competing with each other. I would love to hear Dr Stock-bridge's thoughts about utilizing that in the setting of a master protocol.Dr Stockbridge: We're certainly on record as having advocated for people to do that as a way to get efficient Phase 2 studies done. I think the reason why it doesn't happen is because nobody knows how, or nobody knows what to do to organize such a study. I think the problem has been getting companies that are not anxious to share and not anxious to even be perceived as helping to move the field, move their colleagues along, move their competitors along. It's hard to bring them to the table. They need to be more or less at the same stage because the first company doesn't have a lot of incentive to help drag the rest of the community along. This does need to be organized, I think, at the academic consortia level or some international funding agency or something, and then you're going to have to sell it to the companies.Dr Humbert: It's really interesting. I think that patients should be really strong advocates for that in close partnership with other stakeholders. Maybe the academicians, as Roham said, should try to develop such master protocols and persuade the National Institutes of Health (NIH) or some European funding agencies to support this. Of course, it's difficult to organize, but I think it would be very good to do it as soon as possible, or least try to, because a big disappointment for me is to see multiple, very small Phase 2 studies with invasive measures first presented as negative on press release, but with no data whatsoever presented to the community. I feel that it's a big waste, and I think it's not really ethical to keep these data hidden somewhere.Dr Nicolls: I agree. We're about halfway through our time here, so Marc, I have three different ideas for the next discussion. One is, I'll always remember being at a meeting with Norman Stockbridge and talking about the utility of using the 6-minute walk distance.Another subject is the challenges of add-on (ie, adjuvant) therapy in the 21st century, when we add in a new potentially disease-modifying therapy on top of maximal vasodilation. It can be very, very challenging to get a readout; even though it might be an important therapy, the ability to detect change in a rare disease is challenging.The third topic is, is there a value in developing endotypes for a rare disease, like looking for responder subsets in an already rare disease?What do you think would be the most interesting one for us to discuss?Dr Humbert: I think the patients would like to hear from experts about the typical classical endpoint like 6-minute walk distance, and then we can touch on the other subjects. I would like to discuss a little more the 6-minute walk distance as a Phase 3 endpoint, either on its own or as part of a composite endpoint.Dr Nicolls: Can I ask it in the form of a controversial question? Basically, some believe that the 6-minute walk distance is an overly crude measure of determining whether a drug is useful. The Food and Drug Administration (FDA) has historically been very much in favor of using this benchmark, and it's time-honored and useful. It's certainly usable. Why don't I ask Roham to go first?Roham, what is your view of the 33-m idea, and what do you think about where we are as a field with regard to endpoints?Dr Zamanian: I'm not sure if I specifically should comment on the 6-minute walk, but generally, I think one of the challenges we currently have with endpoints is that it can be difficult to show a treatment effect. Let's assume that's about 6-minute walk distance. We now have the largest number of patients in therapies that are on at least dual therapy from a vasodilator perspective. Most of the current trials enroll new patients on these therapies with New York Heart Class 2 or early Class 3 symptoms.I think, from a 6-minute walk distance perspective, it may be valid to say that we're challenged, because of existing success with vasodilator therapies, to show a treatment effect going forward. On the other end of it, I guess, if you accept that argument and you move toward clinical worsening, whether just purely mortality or a composite endpoint, those studies become extremely difficult and expensive to perform in our rare disease population. I think that challenge that I see is not entirely just the 6-minute walk.I would love to hear Dr Stockbridge's opinion on what the future would be. Mark, I think you and I have been involved with a number of studies, both academic and some early-phase industry, where I think that the fact that we have these patients, even those on intravenous therapies, who are maximally vasodilated, makes it increasingly difficult to pick up a signal of a 6-minute walk distance change with the number of patients we keep enrolling into these Phase 2 studies, which is about 100 to 150 patients.Dr Nicolls: Before I turn it over to Dr Stockbridge, do you want to just quickly comment on this 33-m cutoff? Because you'll recall that there are some trials that have moved forward to Phase 3 with a 10-m walk difference, which turned out to be—I think Phase 3 was terminated. Do you think that's reasonable, or do you think that's unreasonable?Dr Zamanian: I think some of those decisions are based around the appetite for risk and what investigators or decision makers of those studies will accept as a potential “signal” without meeting this sort of a bright-line rule of a P value of .05. I believe Steve Mathai's publication is a study that shows the minimum important difference for a 6-minute walk distance in a connective tissue disease population is around 33 m.I know Marc and others have published also on what's appropriate and meaningful change in 6-minute walk distance. I think those decisions might be both a scientific and a business interpretation of the data in early phase studies.Dr Nicolls: Great. Maybe I'll turn it over to you, Norm. Any Stockbridge: of I need to respond to this that the 6-minute We have the came in and if it was and we said, Because it was other people At no point we that this was the or the or anything more an way to the is another thing that's kind of interesting here, and I don't think anything about it from a of this place other me was a you were to come in with a more disease and say, want to do a 6-minute we would you that it's but you need to show us an effect that a patient has some to We that with We didn't do that with the first few and we didn't start it until we were to get about the on the order of I think the 6-minute walk that have been with vasodilator therapy have been have been I sure of is because we didn't ask for big we got a of drug development here, an to for things that were more of endpoints, I think we some that will now the field more toward disease-modifying of We probably be where we are if we had the of that an effect be that was actually Nicolls: That's a really discussion. The way that I was it was that the at the 6-minute walk distance as the for trials. was That's what always Stockbridge: Nicolls: I'm just whether or not Roham or Marc was under that as if you thought that the thought that the 6-minute walk distance was the for the Humbert: I think we have that the that the 6-minute walk distance was quite a way to favor drug development in a in rare like and we that it has been extremely That's something we very problem we right now is, what can we as an that will support drug development and we develop clinical trials, we often discuss composite endpoints such as to clinical or to treatment To be it has a negative aspect in the walk distance is as which is more I think that we all to develop a more positive endpoint of but it is quite and we often end up with a of which usually walk distance and biomarkers which have no for the I say that I think a lot about endpoints, and I try to the way to develop a drug efficiently for I in the the walk distance has us quite and that other endpoints are more to think the walk distance was a good endpoint for early drug development in PAH and PH future drug development, we will have to using maybe I know that the agencies are quite open to these of We are very from Phase 3 clinical trials with a positive endpoint on in but I it will come one I wanted to because Roham it is the of vasodilator I would say because I not sure it should be as I would say that it's the current of we have of patients on or therapy who with very pulmonary very and a lot of for in of and other clinical that they are on vasodilator I to that they or the to our current but we will have more to or we have some quite interesting results that you may be able to pulmonary on top of what we currently as vasodilator therapy, to or of Maybe we should think the a little bit and that we will be able to add a treatment other on top of the current of to meaningful in pulmonary and have been very in certainly too I remember at the 2018 World Symposium on PH in Nice, there was a to say that all the Phase 2 studies had and that it will be difficult to develop something meaningful it's very difficult to develop new in the field, but when you try to understand the at in maybe which are in at least a of patients, there is certainly for in our patients because are very limited in of and have I think we have to be positive and try hard to develop drugs which should be on top of what we Zamanian: I quite with what you're Marc. I to say that those are I think was to maybe more there are now patients who are on dual or Mark and I are with in an there are some clinical trial subjects who now are on dual or therapy and have that the that were used to the impact of the therapy on as an endpoint are I about this that we need to keep going because there are patients who have very, very disease current we the endpoint can I just are a number of colleagues who have been working on and of and we're very of a meeting by the of the of the patients and using as primary endpoints of clinical trials, I wonder if both Dr Stockbridge and Dr Humbert could to that and how we would as maybe even primary endpoints of Phase 2 studies.Dr Stockbridge: no problem from in having a as the for If patients say they feel on some that's got to be at least as good as on a 6-minute walk, but it has the same potential problem if you've or even 100 people in a study and show a small effect that to be about the same in that effect will be to no got to with that if we're going to move into therapies that have more clinical impact the we have You to want them to show that individual patients are to This an issue with I don't if it's of a of a change in but to be to patients.Dr Humbert: are very interesting I very much in favor of but what Norman just is very The should be of good the effect should be and At the end of the I feel that the we use right now with the drugs we currently develop show This might not be the for all the pulmonary patients with or will show in of That's another field, but I think it that could be a primary endpoint to of but the of the effect has to be as in order to the community, which is not always the with drugs we have but I that it will come one Nicolls: That's great. Maybe we have time for one or more subjects. I see that Dr Stockbridge and just on the first there are where you need to on but in that may have a order more patients we have currently in This us to the challenge of to the of potentially disease-modifying There are different that you could this question, Dr I'm going to you it you would like is maybe a in can with this I'm not it to be when we give we are potentially a and not a of disease, our way forward in the field of actually such as or is extremely difficult because our is based on what we think are and things that are already being by the the that you do you have any ideas for us about how we with Stockbridge: I'll just that in the area of disease and the is if you want to study a new drug, it to be on the of of that the other things you know are useful. That's a little bit in the PAH setting where there are a number of things you could be doing that it's more like You have drugs of different and there may be some about which add to one but also not the that all of the get I think we're going to be out of the of that new therapy be tested on top of other It just though mortality and may be very it's just and studies as the of here and there on the observed and the trial. I think got to be open to the idea of these things in a setting where people that it's not to be on some things you know as as some new Nicolls: is a really interesting I like I don't think it would with a lot of people in the PAH clinical trials community, but from perspective, it interesting. Marc, can you Humbert: I like the thinking the and I have very little to but it Nicolls: Do you think it would be hard to in our Humbert: be because for vasodilator therapy in most PAH patients, and it may be challenging to use of drugs in patients.Dr Nicolls: Dr Stockbridge, do you see any examples of that in Stockbridge: I think people are now just so over the number of therapies that need to be in this that they're now to about but no for it, and of that say, therapy to be may on a different the with other drugs in this field have been the therapeutic area is and people are hard to the therapy that's actually going to be disease a There You had to with If you can the with therapy, you can got a in PAH with disease-modifying is going to be the for from that point Zamanian: Can I ask you you that this idea might not have in our you maybe what some of the challenges would be in this in If you were one of the what would you as these challenges that make it Humbert: you was the idea that would be difficult to Nicolls: What we're is, if we thought we not use the Let's say a new disease-modifying drug that might have a positive vasodilator beyond that of a if there is such a thing as a Let's just it a The of the field is that it's that we do, we have to add it on top of of It's into our by Dr Stockbridge a very I think point that maybe it doesn't have to be that I think if we were to introduce that idea, it would a lot of I think Roham was do you or why do you think there would be would it come What would it Humbert: It would be challenging to ask to vasodilator therapy to be to a new which might be very This will need to be very I quite positive that we are developing new drugs which are really because they with new relevant If some new drugs are working on top of or therapy, strong is that it would be interesting to future development in One way would if this drug is to this new drug and see what's going on in the The other way would be to the that are to be in order to see whether they can be by a new in some of we have new quite and clinical trials have tested either these new drugs and disease-modifying or the of being with these new on their Nicolls: It like we're back up against our time here, so I don't think have time to get to that question, but I think this has been a great discussion. Thank you all for

Publication bias, the failure of small negative trials to be published, has been well documented in all fields of medicine, including rheumatic disease. Evaluating a treatment by summarizing data from...

Analysis of Terminated Hand and Wrist-Related Clinical Trials

This study aims to understand factors contributing to nonpublication and publication bias in clinical trials in Canada. Qualitative interviews were conducted between March 2019 and April 2021 with 34 participants from the Canadian provinces of Alberta, British Columbia and Ontario, including 17 clinical trial investigators, 1 clinical research coordinator, 3 research administrators, 3 research ethics board members and 10 clinical trial participants. We conducted a thematic analysis involving coding of interview transcripts and memo-writing to identify key themes. Several factors contribute to nonpublication and publication bias in clinical trial research. A core theme was that reporting practices are shaped by incentives within the research system taht favour publication of positive over negative trials. Investigators are discouraged from reporting by experiences or perceptions of difficulty in publishing negative findings but rewarded for publishing positive findings in various ways. Trial investigators more strongly associated positive clinical trials than negative trials with opportunities for industry and nonindustry funding and with academic promotion, bonuses and recognition. Research institutions and ethics boards tended to lack well-resourced, proactive policies and practices to ensure trial findings are reported in registries or journals. Clinical trial reporting practices in Canada are shaped by incentives favouring reporting of positive over negative trials, such as funding opportunities and academic promotion, bonuses and recognition. Research institutions could help change incentives by adopting performance metrics that emphasize full reporting of results in journals or registries.

The subjective interpretation of negative trial results during oral plenary presentations

The importance of a Neonatal Intensive Care Unit (NICU) database lies in its critical role in improving the quality of care for very preterm neonates and other high-risk newborns. These databases contain extensive information regarding maternal exposures, pregnancy complications, and neonatal care. They support quality improvement (QI) initiatives, facilitate clinical research, and track health outcomes in order to identify best practices and improve clinical guidelines. The Parkland Memorial Hospital NICU database was originally part of the Maternal and Neonatal Data Acquisition, Transmission and Evaluation project funded by the Robert Wood Johnson Foundation to assess perinatal-neonatal care in Dallas County Texas, 1977-1982. Clinical data points were defined, transcribed and validated in 1977; revalidation has occurred multiple times. Data are prospectively extracted from health records of high-risk neonates among >11,000 births annually. The database contains clinical information on >50,000 neonates, including all initially admitted to the NICU regardless of gestational age or birthweight and since 10/03/2011, all neonates admitted for observation and transferred to the term newborn nursery. The database has provided the basis for QI studies and research designed to assess and improve neonatal care. We discuss the history, evolution, administration, impact on neonatal outcomes, and future directions of our database. IMPACT: A single neonatal intensive care unit (NICU) database was designed for prospective data collection, validated and maintained for 46yrs. This database has supported quality improvement assessment, original clinical research, education and administrative requirements and impacted clinical neonatal care.

Importance: Early termination of clinical trials represents a challenge in clinical research, yet little is known about the frequency and reasons for it. Objective: We aimed to describe reasons for termination, and identify the major predictors of early termination due to low recruitment.Study Selection: We reviewed all cardiovascular clinical trials registered in ClinicalTrials.gov from February 29, 2000 to January 17, 2013 and assessed information about trials that were completed and those that were terminated early.Data Extraction and Synthesis: The dataset was downloaded in XML containing all cardiovascular studies registered. The dataset was exported into R software to extract the values. Logistic regression models were developed to identify independent predictors of early termination due to low recruitment.Main Outcome Measure: Studies classified as completed or terminated prematurely, and reasons for termination. Results: Overall,684 of 6279 trials included (10.9%) were terminated prematurely. The main reason for early termination was lower than expected recruitment rates (278 trials; 53.6%); 24.1% of trials had no clear justification for premature termination. When comparing trials that terminated early due to low recruitment with those that were completed, we found that studies funded by the National Institutes of Health/ United States federal agencies (odds ratio [OR] 0.35, 95% confidence interval [CI] 0.14-0.89), studies of behavior/diet intervention (OR 0.35, 95% CI 0.19-0.65), and single-arm design studies (OR 0.57, 95% CI 0.42-0.78) were associated with a lower risk of early termination. University/hospital (OR 1.52, 95% CI 1.10-2.10) and mixed-source funded studies (OR 2.14, 95% CI 1.52-3.01) were associated with a higher likelihood of early termination due to lower than expected recruitment rates. Conclusions and Relevance: Early termination of interventional cardiovascular trials is common and most frequently due to low recruitment. Funding source, type of intervention, and study design were factors associated with early termination due to low recruitment and might be good targets for the improvement of clinical trial conduct.

Bridging the Translational Research Gap: A Successful Partnership Involving a Physician and a Basic Scientist

To identify and appraise empirical studies on publication and related biases published since 1998; to assess methods to deal with publication and related biases; and to examine, in a random sample of published systematic reviews, measures taken to prevent, reduce and detect dissemination bias. The main literature search, in August 2008, covered the Cochrane Methodology Register Database, MEDLINE, EMBASE, AMED and CINAHL. In May 2009, PubMed, PsycINFO and OpenSIGLE were also searched. Reference lists of retrieved studies were also examined. In Part I, studies were classified as evidence or method studies and data were extracted according to types of dissemination bias or methods for dealing with it. Evidence from empirical studies was summarised narratively. In Part II, 300 systematic reviews were randomly selected from MEDLINE and the methods used to deal with publication and related biases were assessed. Studies with significant or positive results were more likely to be published than those with non-significant or negative results, thereby confirming findings from a previous HTA report. There was convincing evidence that outcome reporting bias exists and has an impact on the pooled summary in systematic reviews. Studies with significant results tended to be published earlier than studies with non-significant results, and empirical evidence suggests that published studies tended to report a greater treatment effect than those from the grey literature. Exclusion of non-English-language studies appeared to result in a high risk of bias in some areas of research such as complementary and alternative medicine. In a few cases, publication and related biases had a potentially detrimental impact on patients or resource use. Publication bias can be prevented before a literature review (e.g. by prospective registration of trials), or detected during a literature review (e.g. by locating unpublished studies, funnel plot and related tests, sensitivity analysis modelling), or its impact can be minimised after a literature review (e.g. by confirmatory large-scale trials, updating the systematic review). The interpretation of funnel plot and related statistical tests, often used to assess publication bias, was often too simplistic and likely misleading. More sophisticated modelling methods have not been widely used. Compared with systematic reviews published in 1996, recent reviews of health-care interventions were more likely to locate and include non-English-language studies and grey literature or unpublished studies, and to test for publication bias. Dissemination of research findings is likely to be a biased process, although the actual impact of such bias depends on specific circumstances. The prospective registration of clinical trials and the endorsement of reporting guidelines may reduce research dissemination bias in clinical research. In systematic reviews, measures can be taken to minimise the impact of dissemination bias by systematically searching for and including relevant studies that are difficult to access. Statistical methods can be useful for sensitivity analyses. Further research is needed to develop methods for qualitatively assessing the risk of publication bias in systematic reviews, and to evaluate the effect of prospective registration of studies, open access policy and improved publication guidelines.

CMV infection and glioma, a highly controversial concept struggling in the clinical arena

Ghostwriters, Data Manipulation and Dollar Diplomacy: How Drug Companies Pull the Strings in Clinical Research

Prevalence, characteristics, and predictors of early termination of cardiovascular clinical trials due to low recruitment: Insights from the ClinicalTrials.gov registry

The 2001 EU Clinical Trials Directive aimed to harmonize the regulation of medical research, but achieved the opposite. Various attempts are underway to update the directive to make it easier to safely conduct medical research in Europe.