现在的位置: 首页研究点评>正文
[Lancet Respir Med发表文章]:组织重症医学国际研究:目前的调整及可能的解决方法
2020年01月18日 研究点评 暂无评论

SPOTLIGHT|ONLINE FIRST

Organising international research in critical care medicine: current challenges and potential solutions

Jan J De Waele, Dylan W de Lange

Lancet Respir Med Published: December 17, 2019 DOI:https://doi.org/10.1016/S2213-2600(19)30475-8

During the first years of intensive care medicine, many of its treatments were introduced because they were thought to be beneficial to patients, despite there being a lack of evidence from high-quality research to support these premises. However, in the last 20 years intensive care medicine research has picked up the pace and countless randomised controlled trials have been performed, supplying clinical care with the much needed scientific foundation, discarding obsolete treatments, and saving lives.

Nowadays, the standard of care in intensive care units (ICUs) has reached such a high level that improvements are generally small and randomised controlled trials are often negative—ie, they do not show a statistical benefit of a new treatment over usual care. For example, when the enormous survival benefit of the first early goal-directed therapy trial could not be confirmed by three trials that were performed more than a decade later. Clearly, better designed, larger scale, multinational studies are required to investigate the real-world value of ICU care. External validity increases when studies include patients from different hospitals, health-care systems, and countries. A larger number of participating sites will ensure fast patient enrolment, which is often necessary in studies with several thousands of patients included. Also, large-scale patient recruitment means some trials last for several years, in which time changes in other aspects of care might affect the impact of the intervention under study. However, large-scale studies are fraught with challenges, both from a study design persepctive as well as a practical, organisational perspective.

Although randomised controlled trials provide us with the best quality evidence for a treatment, there are still some pitfalls that can cause the outcome of the trial to be inadequate. In particular, sample bias, inability to appropriately stratify patients, the selection of inappropriate endpoints, and the inability to mask treatments for patients as well as investigators can lead to poor methodology and negative trials.

In critical care medicine and emergency medicine, a treatment should be given immediately, and postponing appropriate treatment might lead to inferior outcomes. Patients who are admitted to the ICU very often do not have decisional capacity, either because of the illness that caused them to be admitted or because they are sedated. However, the need for written informed consent (or even deferred consent from family or surrogate decision makers) leads to exclusion of up to 50% of eligible patients or delay of the proposed study treatment. The window of opportunity for the maximum treatment effect (the so-called golden hour) might be missed if the treatment can only be applied after informed consent is granted. Therefore, patients included in studies represent a different sample than the patient population in which the treatment is going to be applied in the future (sampling bias). To overcome these limitations of randomised controlled trials, researchers have shifted towards observational research.

To perform larger-scale studies, several research networks have been created that consist of different ICUs contributing to a portfolio of clinical trials. Often these networks are regional or national networks, such as the Canadian Critical Care Trials Group, ANZICS Clinical Trials Group in Oceania, or BRICNet in Brazil. Many of them have been very successful in organising both observational and interventional studies, but performing larger scale, adequately powered studies within a limited amount of time—eg, 12–24 months—remains difficult. Creating multinational research networks to accelerate patient recruitment, however, is a different challenge. Although collaborations have been setup to target the organisation of particular projects, a more long-lasting international or even global initiative has not been attempted for a number of reasons, some of them administrative or legal.

In 2018 the General Data Protection Regulation (GDPR) was issued in Europe, which was intended to protect the privacy of all European citizens. Although clinical research was not the primary objective of this regulation, the consequences for research in the EU are far-reaching and have obstructed observational research throughout Europe. The process of obtaining ethics committee approval for observational studies still requires application with national, regional, or local ethics committees, depending on the country. There is a wide variety in this process from country to country, leading to a huge difference in the duration of the application process that varies from 7 days to up to 10 months. Also, more than half of the 16 countries in this study were required to obtain patient consent for data collection, whereas other countries do not have such a requirement. This disparity leaves many researchers dissatisfied with the procedure and obviously dampens the enthusiasm for international research projects.

In parallel with the above trends, and undoubtedly often triggered by the GDPR legislation, institutions increasingly require formal data use agreements for projects, with individual contracts between the sponsor and every individual participating hospital, requiring review by legal departments; contracts often going back and forth for weeks or months until an agreement is reached.

Finally, insurance for observational studies is also routinely required and should now also be provided by the sponsoring organisation. Insurance premiums can vary considerably from country to country and can further increase the cost as well as the administrative burden of initiation and organisation of international research. All of which lead to a high administrative and financial burden with added complexity in the case of international research projects. Overcoming these hurdles will be challenging particularly for global projects, but at least at the European level there are opportunities for the EU to facilitate these processes.

How can we overcome these hurdles in international research? Firstly, correct interpretation of the GDPR is urgently required, as well as harmonisation at the European level for non-interventional research, similar to what has been set up for interventional studies, including ethics committee application processes and insurance requirements. Development of universal data use agreements between hospitals and researchers might solve this issue; alternatively low-risk observational research could be exempted from the need to have a patient-related insurance. However, authors of medical articles are increasingly considering insurances to protect them against slander lawsuits. Nevertheless, if observational data is to be accessible for research and quality-of-care registries, the informed consent prerequisite will create biased inclusions and limit the external validity of the findings of such observational data.

The methodology of clinical studies in critical care is a quickly evolving area, and the regulations should be flexible enough to allow rapid introduction of these new approaches. One example is adaptive platform trials (APTs). To overcome the methodological flaws of randomised controlled trials, research should imitate reality; researchers should be able to allocate a patient to different treatment options (eg, two different antibiotics rather than one) and should still be able to analyse the difference between two different treatments, provided that all other variables and treatment options (eg, corticosteroids, mechanical ventilation, and vasopressors) are distributed evenly among all study groups. APTs represent the largest departure from traditional randomised design and they allow for the study of multiple interventions in a single disease (or condition) in a perpetual manner. Another benefit of APTs is that they also incorporate within-trial adaptations, such as response-adaptive randomisation rules to preferentially assign interventions that perform most favourably, meaning that a patient is less likely to be assigned to a study group that is (statistically) likely to be inferior to the other study groups. The versatility of APTs enables them to answer a lot of research questions in an efficient manner but makes them much more complex and costly. Another example is the introduction of data science. Evidence for many interventions can be derived from observational data. Data that is already entered into electronic health records could be made accessible for research or quality-of-care registries. Such data should be anonymised and cannot be traced back to individual patients and or health-care institutes. However, if we want observational data to be accessible for these registries, the informed consent prerequisite will create biased inclusions and limit the external validity of the findings of such observational data.

Research in critical care medicine is a rapidly evolving area and evidently there is an increased need for larger scale, multinational studies. Administrative, ethical, and legislative hurdles, however, make meeting this need increasingly difficult, with differences between countries adding further complication. Harmonisation of these aspects at the European level would facilitate research and allow better informed health-care for critically ill patients in Europe and beyond.

给我留言

您必须 [ 登录 ] 才能发表留言!

×
腾讯微博