N.P. Juffermans [1,2], L. Heunks , P. Pickkers , A.J.C. Slooter , P. Hemelaar , M. van der Jagt , M. Brackel7,8, L. Vloet [8,9], D. Gommers 
1, Department of Intensive Care Medicine, OLVG Hospital, Amsterdam, the Netherlands
2, Laboratory of Experimental Intensive Care and Anaesthesiology, Amsterdam UMC, University of Amsterdam, Amsterdam, the Netherlands
3, Department of Intensive Care Medicine, Amsterdam UMC, University of Amsterdam, Amsterdam, the Netherlands
4, Department of Intensive Care Medicine, Radboud University Medical Centre, Nijmegen, the Netherlands
5, Department of Intensive Care Medicine, University Medical Center Utrecht, Utrecht, the Netherlands
6, Department of Intensive Care Medicine, Erasmus University Medical Center Rotterdam, Rotterdam, the Netherlands
7, IC Connect, patient organisation, Alkmaar, the Netherlands
8, The Family and Patient Centered Intensive Care Foundation, Alkmaar, the Netherlands,
9, Department of Emergency and Critical Care, HAN University of Applied Science, Nijmegen, the Netherlands
N.P. Juffermans - email@example.com
Development of a research agenda for critical care medicine in the Netherlands: the time is now
The development of a research agenda is important for critical care medicine in the Netherlands in order to improve care for critically ill patients. Also, defining the most important research questions informs policy makers and funding societies. This manuscript describes the initiation of a research agenda for critical care medicine by identifying relevant knowledge gaps in the following domains: haemodynamic management, mechanical ventilation, infection, inflammation, neurology, post intensive care syndrome, ethics and artificial intelligence. A survey sent out to all NVIC members indicated that Dutch intensivists deemed research questions in all domains relevant, without a clear priority for a specific area of research. Consultation with patient representatives confirmed the importance of this list of research topics, adding themes related to outcome of critical illness, aftercare and relatives. In line with this, knowledge gaps were identified in the Dutch guidelines in every domain of critical care medicine. Taken together, a research agenda for critical care is needed and can cover research questions related to all domains of care.
The COVID-19 pandemic has highlighted the crucial role of critical care medicine in our society. Financial support for research initiatives aimed at understanding COVID-19 and improving the care of critically ill COVID-19 patients was rapidly made available by the Dutch government. However, also after this pandemic is contained, there will be a persistent need to invest in research on optimising treatment for all critically ill patients to improve their outcome, regardless of the reason for ICU admission. In fact, characteristics of COVID-19 disease overlap with critical care syndromes due to other causes, including a dysregulated host response, severe hypoxaemia, shock, delirium and coagulation abnormalities. It is equally urgent to improve the care of COVID-19 patients as well as other patients with critical illnesses. COVID-19 has also helped to broaden awareness of the long-term consequences of ICU treatment that hamper quality of life, collectively referred to as the post intensive care syndrome (PICS). These long-term effects necessitate aftercare and rehabilitation programs. Again, PICS is not specific for COVID-19 but pertains to all patients with a prolonged trajectory of ICU stay.
Critical care medicine has made important advances in the past years, justifying an independent scientific society in the Netherlands. Currently, Dutch critical care medicine does not have its own research agenda, which hampers further development. This development is much needed. Although advances have been established in recent years, many fundamental and clinically relevant questions remain to be investigated in every domain of critical care medicine. These research domains cover the classical themes of the critically ill, including haemodynamics, ventilation, infection, inflammation and neurology. Overarching domains include ethics, long-term consequences of critical illness such as PICS and critical care methodology such as applications of artificial intelligence (table 1).
Table 1. ICU research questions per domain
The development of a research agenda for critical care medicine is important to inform policy makers and funding societies such as the Netherlands Organisation of Research and Development (ZonMw), as well as other healthcare professionals. Recently, the Research in Critical Care Network of the Netherlands (RCCNet) was founded with the aim to improve collaboration and impact of scientific research by engaging all ICUs in the Netherlands. RCCNet is endorsed by the Dutch Society of Critical Care (NVIC), which requested RCCNet to start developing a much needed research agenda for critical care medicine in the Netherlands. This manuscript describes the start of this effort, in which close collaboration was sought with all stakeholders, including Dutch intensivists, the NVIC, the Family and Patient Centered Intensive Care (FCIC) Foundation and the patient organisation IC Connect.
To determine the most relevant research questions that need to be answered in critical care medicine in the Netherlands, several approaches were followed. The first approach aimed to determine which ICU research questions were judged to be most relevant by Dutch intensivists. For this, a survey was sent to all members of the NVIC. The survey contained a list of research questions in every domain of critical care medicine (table 1). These research questions were formulated in consensus between the four members of the RCCNet executive committee. All NVIC members were approached by email and asked to rank these research questions from 1 to 5 (1=waste of time and effort, 2=interesting but no priority, 3=important but not as important as other questions, 4=reasonably important, 5=very important). In addition, NVIC members were asked to add questions, if deemed necessary. Two reminder emails were sent with several weeks in between.
The second approach aimed to determine the specific research areas in which Dutch ICU researchers are perceived to have a strong track record. This was done in recognition that research is performed internationally and not all relevant research questions need to be answered by Dutch research efforts. For this aim, we approached past and/or present editors-in-chief of the five critical care medicine journals with the highest impact factors, covering clinical or preclinical science. None of the authors of this manuscript were included. These editors-in-chief were asked to select 5-10 of the following areas of expertise that they considered to have the highest priority: pathophysiology of shock, treatment of shock, diagnosis and monitoring of shock, mechanical ventilation, weaning from mechanical ventilation, pathophysiology of acute respiratory distress syndrome (ARDS) and ventilator-induced lung injury (VILI), extracorporeal membrane oxygenation (ECMO), antibiotic treatment, prevention of infectious disease/selective decontamination of the digestive tract (SDD), pathophysiology and biomarkers of inflammatory syndromes, immunomodulation, delirium/acute encephalopathy, ICU-acquired weakness, acute kidney injury, cardiac arrest, trauma, acute brain injury, thrombosis and haemostasis, post intensive care syndrome (PICS), quality indicators, artificial intelligence and ethics. They were also allowed to mention other topics, if not covered by these.
Table 2. Opinions of editors-in-chief on domains in which Dutch ICU researchers have a track record
The final approach was that the guideline committee of the NVIC, the Dutch FCIC foundation and the patient organisation IC Connect were asked to indicate 5-10 knowledge gaps in the existing critical care medicine guidelines and practice with the highest priority.
Survey to NVIC members
Of the 899 critical care physicians to whom the survey was sent, 51 returned the survey, resulting in a response rate of 6%. All research questions were rated with a mean between 3 and 4, with the domain ‘inflammation’ receiving the highest score, but with only slight differences between domains (figure 1). All research domains were deemed relevant but no domain in particular received a clear priority. With regard to a specific question, only the question about optimal modulation of the inflammatory response of the Inflammation domain received a score above 4 (figure 2). There were six questions that received a score below 3, which were related to the infection, haemodynamics or PICS domains. Again, most questions were rated with a mean between 3 and 4, indicating that all questions were deemed more or less relevant without a clear priority.
Qualitative questions to editors-in-chief of critical care journals
The response rate of the editors-in-chief was 100% (5/5). The areas in which, in their view, the Netherlands has a strong research reputation are shown in table 2. Answers suggest that Dutch ICU researchers excel in aspects related to inflammation, shock, respiratory failure, brain injury and ethics. However, it is important to realise that the editors-in-chief indicated that they found the question difficult to answer because of the lack of a systematic approach and hence there is a risk of bias.
Knowledge gaps as identified in Dutch ICU guidelines
All members of the guideline committee of the NVIC (n=7) scored all domains as having knowledge gaps. The domain receiving the highest research priority score was mechanical ventilation, in particular the pathophysiology of ARDS and optimal monitoring of weaning. The second most important domain was delirium. Other domains that also received a high score included infection prevention, PICS and ECMO. The committee stressed that ranking was found to be somewhat arbitrary as knowledge gaps are present in all domains of critical care.
Knowledge gaps as identified by patient organisations
The FCIC foundation and IC Connect are currently in the process of developing a knowledge agenda from the perspective of patients and relatives about the psychosocial consequences of critical illness and ICU treatment. From preliminary data on consultation rounds with patients and relatives, as well as derived from both personal and management experiences, FCIC and IC Connect recognise all the topics mentioned above to be important for research in critical care. From the perspective of ICU patients and their relatives, the most important knowledge gaps are prevention, incidence and treatment of the longterm consequences of an ICU admission in aftercare and rehabilitation programs, which should also include economic and social aspects and should involve both patients and relatives.
Results show that research questions in all domains of critical care research are deemed relevant by intensivists, editors-inchief, members of the NVIC guideline committee and patient organisations. This is an important finding, suggesting gaps in knowledge and evidence in every domain of critical care medicine.
Of note, the research questions that were evaluated were mostly generic, covering basic aspects of everyday clinical practice on the ICU that are relevant in a broad group of critically ill patients, irrespective of the specific cause of their illness. No clear priority could be given to a specific area of research. If anything, inflammation received the highest score, which may reflect the broad area in which a dysregulated inflammatory response may play a role. Of importance, no research questions were deemed irrelevant. Thereby, we argue that this result highlights a need for increased funding resources with the aim to fill knowledge gaps in all aspects of critical care medicine. Determining which specific research questions need to be prioritised was not the aim of this current investigation. Rather, we aimed to highlight the domains with obvious knowledge gaps.
The response rate of intensivists was very low. Of note, the survey was sent in an ongoing COVID-19 pandemic, which could have contributed to the modest response rate. Most probably, respondents were more likely to be personally engaged in performing critical care research compared with those not responding, although we do not feel that this caused systematic bias in the results. Responses are likely to represent opinions of critical care physicians practising in the Netherlands.
Obviously, not all fields of critical care research need necessarily to be performed in the Netherlands. For optimal gains from funding, it is important to consider in which domains Dutch critical care researchers already have an established and wellrecognised research line Therefore, we asked the opinion of past or present editors-in-chief of the five most influential critical care journals about domains in ICU research in which Dutch researchers have a strong track record. Their general view seems to be that ICU research in the Netherlands covers a broad area of research domains. Of note, the risk of bias in these results may be high, as respondents noted that their answers may be driven by personal connections and recent meetings rather than by an objective evaluation. Nevertheless, to a considerable extent, their suggestions were in agreement. Although their responses are limited in size and quantitative character, we think we can conclude that Dutch ICU researchers successfully cover a broad range of topics in critical care research.
The NVIC guideline committee indicated that knowledge gaps are present in guidelines covering all domains of critical care, most importantly in optimisation of mechanical ventilation and weaning.
Important topics of research from the perspective of former ICU patients and relatives are PICS, integrated multidisciplinary aftercare and rehabilitation programs, delirium, sepsis, weaning and care of relatives. Although it is likely that improvements in the acute phase of critical illness will lead to a decrease in the incidence and severity of PICS, it is equally important to acknowledge that long-term sequelae of critical illness are not likely to be avoided altogether. From the patient perspective, use of patient-reported outcomes in studies can improve trial design. Also, continuation of the collection of data on recovery in patients after ICU discharge can help to inform ICU care; artificial intelligence can play an important role in this respect.
Taking an overall perspective, a final comment relates to lessons learned from COVID-19 in terms of organisation of ICU care. Research should also focus on innovative solutions that can support ICU care during times of shortage in ICU nurses. The COVID-19 pandemic has resulted in novel collaborations between critical care researchers and researchers in other fields, which were successful in obtaining funding to set up a broad network for data sharing and interpretation. We aim to further stimulate such collaborations, also after containment of the pandemic. A research agenda for critical care medicine is a crucial step in that direction.
Relevant knowledge gaps exist in all ICU domains, highlighting the need for a research agenda for critical care medicine in the Netherlands.
All authors declare no conflicts of interest. No funding or financial support was received.