Effective commercial/industry-led research delivery in UK critical care

The UK Critical Care Research Group and Clinical Research Network (CRN) Critical Care Specialty Group, with important inputs from Faculty of Intensive Care Medicine, and Intensive Care Society have made significant progress in designing and delivering translational, health services and basic science research. ¹ The annual UK Critical Care Research forums, supported by both ICNARC and the Intensive Care Forum, have provided an important driver for this progress. ² These advances were further strengthened by a successful James Lind Alliance research prioritization exercise in 2014. ³ There is though an acknowledged need to better engage commercial partners and undertake industry-led research involving equipment manufacturers, providers of diagnostics or Pharma. This would widen the potential research portfolio, and provide additional sources of income to research delivery teams in critical care units across the UK, compensating for some of the limitations to non-commercial CRN funding (e.g. short-term nurse contracts, latency in funding, and disconnect between front-line National Health Service (NHS) systems and the National Institutes of Health Research (NIHR)/CRNs). Indeed, industry-led research can provide vital traction for research delivery teams working in busy UK critical care units with multiple competing demands on resources.

The relatively protected resourcing of critical care, granularity of information capture; and the existence of complete, refined sources of validated, epidemiological, and case-mix adjusted outcome data (via Intensive Care National Audit & Research Centre, ICNARC and Scottish Intensive Care Society Audit Group, SICSAG) should make UK critical care very attractive to the commercial sector. This was confirmed by preliminary discussions at a Research Colloquium in 2015 with major funders including industry. ⁴ The NIHR CRN structures have also developed an international reputation for effective robust research delivery and the UK Critical Care Research Group is receiving increasing numbers of requests for support for international investigators.

Despite all these drivers behind commercial research studies in UK critical care, uptake has been slow and patchy. Here, we would like to highlight some of the barriers to delivering commercially funded research studies in the context of UK critical care, to offer an explanation for this disconnect between opportunity and delivery, and offer some potential solutions.

Barriers

An overarching problem with industrial studies in critical care is that commercial research processes used successfully in other clinical contexts such as ‘cancer’ or ‘cardiovascular medicine’ or ‘stroke’ are directly applied. However, this leads to fundamental structural problems when research questions are asked or protocols are applied in the critical care context. Critical illness almost inevitably involves multi-organ dysfunction and patients may have critical illness syndromes (‘sepsis’, ‘adult respiratory distress syndrome’, ‘traumatic brain injury’) rather than clearly defined disease processes. Further, there may be underlying chronic (even irreversible) disease underlying the critical illness including significant effects from the trajectory of decline prior to presentation with organ failure. The personalized medicine approach, so successful in cancer for example, does not therefore always translate in the critical care setting, leading to significant infeasibility in relation to target populations, consent processes, intervention timing and data collection. Extubation on an intensive care unit is not like chemotherapy, and yet it may be treated as such within protocols: it is not usually feasible for patients to complete a questionnaire on quality of life 1 h post extubation. Finally, protocols may not be informed by up to date epidemiology (including outcomes); knowledge or critical care definitions (e.g. sepsis or adult respiratory distress syndrome).

The third party clinical research organizations (CROs) or ‘monitoring’ companies selected by sponsors (commercial organizations) to deliver the studies have different practices, standards, documentation and resources. Further, it appears to be common to change the CRO during the research study recruitment period, leading to unnecessary duplication and replication of research processes, loss of continuity, and time-consuming reconstruction of site files. The clinical research associates (CRAs) who deliver CRO services often have no background in critical care, particularly in the UK context and indeed may know very little about the scientific basis of the study hypothesis. This leads to a protracted indirect conversation with sponsors and sponsor physicians/scientists, with CRAs acting as a highly inefficient ‘go between’, slowing the whole process down.

Site selection processes and electronic/survey screening tools are ad hoc, and poorly designed to establish suitability. Further, they do not permit any strategic matching (at a national level) of resources/case mix to the commercial sponsors main research question; sharing of resources; or strategic support to promote growth across this sector across the whole country.

The study documents (main protocol, summary documents, laboratory documents) are constructed with technical language and acronyms that make rapid assessment of study hypotheses and feasibility, difficult. This matters, because inevitably local critical care research delivery teams, particular research leads and senior research nurses have limited time and bandwidth. They often have to review multiple applications and weigh the relative feasibility and merits of each and consider potential interactions, co-enrolment issues or overt conflicts between studies (either commercial or non-commercial).

It is often clear that patient information sheets (PISs) are designed as a ‘catch all’ and remove risk of liability across international legal systems. However, the result is that PISs are often very long, text heavy, full of technical language (medical and legal) and often a key reason for non-feasibility through poor consent rates. Indeed, within our institution we have obtained an indication that families will provide consent, only for this to be ultimately refused when they read the PIS. The PISs are also a barrier to studies that require multiple steps or differential consent. A distressed family have to consent to ‘all or nothing’ rather than having a staged consent to facilitate screening tests, then randomization/intervention and the question of optional samples (e.g. for genetics).

Internet trial systems, including electronic patient screening tools, case report forms and event reporting systems are often poorly designed and do not fit the UK context. Further, there is huge heterogeneity in these systems across the commercial sector.

The true costs of delivering research in the UK are often not appropriately estimated and resourced. For example, there may be no/inadequate resource allocated to set-up, screening, site initiation and monitoring. The lack of resource around screening is often further exacerbated by requests for submission of extended screening data on populations of patients that do not match inclusion criteria. Finally, requests for additional data after study completion may have no funding.

Solutions

We would like to suggest some simple measures that would help to start to address these problems:

The UK critical care research group should set up its own Commercial Research Organisation (CRO), employing a pool of CRAs with expertise in the UK setting. The CRAs should have UK critical care experience, but also understand how allied resources (imaging, laboratories, wider NHS institutions) operate, including limitations. Although, this company may not initially be selected for international studies, it would provide a resource for studies solely/mainly based in the UK and provide a potential source of income for the UK critical care research group. This research delivery company would grow as commercial research becomes more embedded in UK critical care.

The current ad-hoc approach to individual investigators and institutions from commercial companies needs to be stopped. It is inefficient and disruptive. The pre-screening assessments should be informed by the UK Critical Care Research Group to ensure they fully explore feasibility and suitability of sites and that they can be pre-populated with relevant data (e.g. local resources, epidemiological, case mix). This would also ensure appropriate and efficient distribution and also link this screening process to delivery to help drive up research activity. Potential collaborations between geographically or operationally linked hospitals could be exploited and resources (laboratories/staff) shared.

The UK Critical Care Research Group should create a research design service, including potential support from ICNARC and SICSAG (e.g. in relation to statistical/health economic/long-term outcomes analysis). This could provide potential commercial companies with a resource to answer phase 2 and phase 3 questions in UK critical care units. This would lead to pragmatic, feasible protocols with additional benefits of linking the main trial databases to other health care databases in the UK (e.g. in relation to long-term follow-up).

The UK Critical Care Research Group should provide allocated research leads to assess and summarize the research protocols in ‘UK Critical Care speak’. The UK Children's Cancer Research Group (UKCCRG) summary documents would rapidly become worth their weight in gold. Obviously the summarizer could be the assessor for CRN adoption. This would allow individual teams with stretched resources to make rapid assessments and accelerate decision-making.

The UK Critical Care Research Group could provide a single web-based system to support CRFs and the submission of all study data, ultimately linked to ICNARC/SICSAG with appropriate controls.

The pre-site/site initiation visits need to be re-structured to not only enable the sponsoring company and supporting clinical research support service (including individual CRAs) to assess centres, but allow these centres to efficiently gauge feasibility and efficiently answer early questions without a clumsy (and often duplicated) 2-stage process whereby the sponsor is contacted by the CRA/monitoring representative for each query from sites.

A centralized approach would make uniform standards and governance in relation to research conduct easier, including efficient shared learning from adverse event reporting. The frequent application of additional selection processes pre- and post randomization; protocol amendments; and lack of rules regarding publication of negative results could be better explored and challenged.

In summary, we believe that the UK Critical Care Research Group, in tandem with the NIHR Specialty Group, can offer an unrivalled engine for delivering commercially sponsored research across the whole industry sector related to critical care. Further, it is clear that a UK perspective is relevant for delivering research in the UK. ⁵ We believe that a small initial coordinated investment would rapidly pay-back many fold. Finally, a more organized approach would lead to the asking of better research questions and drive closer cooperation between the clinical critical care service and basic laboratories/phase I investigators.