Abstract
Clinical trial registration regulations have improved a lot over the past decades. However, not every country has made the same progress. Fraud is also large problem, it is estimated that half of all clinical trials are never registered. Data manipulation has been detected in a small percentage of clinical trials but researchers report a much higher number fraud by their colleagues. Clinical trial data shows great potential to help with research even after its original trial has ended. This potential can’t be reached however, due to problems with data sharing. Clinical trial data sharing platforms only exist on a small scale which hinders secondary use of clinical trial data.
Introduction
Clinical trials, a subset of clinical studies, account for a large amount of medical data available today. Approximately 100 clinical trials are registered every day on the biggest clinical trial registry website, which has over 280,000 registered trials currently. [1] Sadly the data from these clinical trials is not as easily available. Furthermore, only about half of all clinical trials are registered at registry websites. This article will look into these and more problems related to clinical trials as well as positive development and usage of clinical trial data.
What are clinical trials?
Clinical trials are a form of clinical studies. This refers to a study which performs research with volunteers to answer medical questions or increase medical knowledge. There are two types of clinical studies, observational and interventional. Interventional studies are usually called clinical trials and are characterized by participants receiving interventions from the researchers. The interventions can be a wide range of things such as drugs, lifestyle changes or medical devices. In observational studies the researchers observe the participants without imposing interventions. Clinical trials are divided into five phases. The development of a new drug starts in pre-clinical research. The drug is tested in-vitro and in-vivo but not on humans. Information about efficacy (degree of ability to produce a desired effect) and toxicity is gathered. Many new experimental drugs don’t pass this phase because the negative effects are too great or because there would be no significant benefit in developing the drug. If the drug shows potential and passes safety tests it will move to phase 1. In this phase 20 to 100 healthy volunteers are given the drug or product. The goal of this phase is to test the safety of the product and, in the case of a drug, to find the maximum safe dosage. If the product is considered safe, it will move on to phase 2. During this phase the efficacy and side effects are examined. This is done with several hundred volunteers, mostly patients in the target group. The best dosage is also further determined. Phase 3 consists of large scale (300 – 3,000 patients) research to further test efficacy and effectiveness. After phase 3 all the results will be submitted to governmental instances such as the United States Food and Drug Administration (FDA) and the product will be accepted or rejected. The products that are approved for sale enter phase 4: post marketing surveillance. The drug enters the market but is still being monitored. This safety surveillance makes sure that long-term effects are also discovered. The clinical trial durations are not decided by law but are determined on a trial by trial basis by the researchers. The estimated duration and other information are gathered and formed into a protocol which is sent to the FDA for approval. Once the FDA approves of this new drug testing protocol the study can enter phase 1. In the end, only 1 out of every 5,000 drugs that enter pre-clinical research eventually passes all the phases and becomes an approved drug. [2] The development of a drug from its discovery to production can take up to 20 years and in some cases even longer. In some cases the total costs can exceed $1 billion. [3]
Clinical trial regulations
Clinical trials are not subject to the same rules everywhere in the world. In the US for example, research done with human subjects has to be approved by a local Institutional Review Board (IRB) and the FDA. Studies that enter phase 2 must register their study on clinicaltrials.gov which is part of the National Institutes of Health (NIH). When the study is registered the pre-clinical and phase 1 information also has to be entered. Alternatively, studies can choose to register before they complete phase 1. This can help them find volunteers for their research. Finally, drugs must be approved by the FDA before they are allowed on the market. These are only a few examples of regulations in the US. Clinical trials in the EU are governed by the Clinical Trials Regulation and the EU has their own database at clinicaltrialregister.eu. Similarly to the US, clinical trials must be registered to enter phase 2. Not every country has such strict rules however. When clinical trials became subject to many laws and rules in the US and EU, companies moved their trials over to India on a big scale. The laws in India were outdated and much more favorable for the companies. This caused a huge surge in clinical trials based in India. Because of many incidents of which some got a lot of publicity, India was forced to update its laws which in turn brought the number of clinical trials to an all time low. In 2013 the approved number of clinical trials by Drug Controller General of India (DCGI) was as low as 17 [4]. Ever since this all-time low India has been changing the laws and regulations to find a middle ground and the number of clinical trials is slowly rising again. Even though progress is being made in India, some countries don’t require clinical trials to be registered at all.
Clinical trial usage
Clinical trial data is generated from the clinical trial which is used to answer a medical question or gain specific medical information. The data either serves to answer the question of the study or to support the patient in the medical problems that might arise. After the clinical trial is finished this data is hopefully uploaded and used in the article or paper which addresses the study. Ideally, the clinical trial data will continue to be useful beyond this point though. This is where secondary use of clinical trial data comes in. This is the usage of clinical trial data from previous studies for your own study or research. Even though a lot of data is available, this still doesn’t mean the right data is always available for further research. If your research question is too specific the available data might be too general. This also goes the other way around, if a previous study had a too specific research question its data might not contain all the information you need. The problem is amplified by the fact that searching for data is still very hard, so availability of sources is limited. The sharing of clinical trial data has a long way to go but examples of its potential exist too; The Placebo & Standard of Care Data Sharing (PSoC) Initiative focusses on sharing and reusing data from clinical studies. Data from 90 clinical trials are stored in its repository as of 2017. A clinical trial using this data revealed that the data sped up the study time by 30%.[5]
What is holding back clinical trial data usage?
There is a lot of clinical trial data available nowadays. All this data could lead to great breakthroughs and discoveries but there is one large obstacle preventing this from happening. This is the sharing of clinical trial data. A few quotes from Dr. John Marshall, Georgetown University illustrate the problem: “(…) there is so much variability in the input data (pictures, text, etc.) which means that they haven’t really figured out yet how to simplify this data so it can be used collaboratively. (…) Data sharing is an incredible obstacle to our success right now. The data is there but it isn’t necessarily in the same place. They don’t talk or even speak the same language. Some privacy laws in the US prevent them from connecting that patient data.”[6] To summarize, data sharing is very difficult and sometimes even impossible because there are no normalization standards. Multiple data sharing platforms such as Georgetown’s G-DOC [7] are available or are undergoing development with the goal of providing this service, but doing so is not trivial. One article describing this issue even called it In search of the Holy Grail [8], where the Holy Grail would be a data management system that allows for the integration of many sources of clinical data. Such systems would also allow for more secondary use of clinical trial data. Another issue is fraud. It is estimated that 50% of all clinical trails are never registered. The numbers range from 19% to 98% depending on the study and criteria so there is no clear percentage of unpublished trials.[10] The clinical trial data can also be subject to fraud. A study showed that 0.01% to 0.40% of investigated datasets contained fraudulent data. Other studies showed that 27% to 50% of researchers said they knew of or suspected data manipulation by colleagues.[11]
Conclusion
Clinical trial data is generated in very large quantities all over the world. All this data could prove immensely useful for new research or to accelerate current studies. Clinical trial registration regulation has improved a lot over the past decades but not every country has made the same progress yet. Clinical trial data sharing however is still in its early stages. Because there is no global platform with a normalised data structure, gathering useful data for secondary uses is very difficult to do. Small programs such as G-DOC are addressing this issue by creating data sharing platforms. These platforms are either very specialised or locally focused however so there is still a long way to go.
Feature image credit: Yoshua Coleman Licensed via Creative Commons 2.0.