Investigator-led clinical research consortia: The Canadian Critical Care Trials Group

John C. Marshall, MD, FRCSC, Deborah J. Cook, MD, MSc, FRCPC, for the Canadian Critical Care Trials Group

Advances in the care of critically ill patients are dependent diagnosis and management of infection in the , upon rigorous clinical research undertaken to characterize natu- and end-of-life care. In the process, the CCCTG has developed a ral history and risk factors, and determine optimal approaches to highly effective culture of scientific mentoring, and has served as the management of the diseases of the critically ill patient. The a model for investigator-led critical care research groups around Canadian Critical Care Trials Group (CCCTG) was formed in 1989 the world. This review summarizes the history, activities, ap- to foster such research. It has grown to become a national, proaches, and challenges of the CCCTG, in the conviction that multidisciplinary organization with more than 100 members, and investigator-led groups such as ours represent the future of more than 3 dozen active research programs. Its members have intensive care unit-based research. (Crit Care Med 2009; been highly successful in obtaining funding for, completing, and 37[Suppl.]:S165–S172) publishing well-designed studies that have informed international KEY WORDS: critical care trials; research collaborations; inves- practice in areas such as transfusion, prophylaxis, tigator-led research long term outcomes from acute respiratory distress syndrome,

he Canadian Critical Care Tri- tered clinical research. This situation is program, therefore, is strongly influ- als Group (CCCTG) is the old- changing, and there is a growing need to enced by the requirements of regulatory est investigator-driven clinical optimize the structure and enhance the agencies, and limited by the requirement research consortium in the influence of investigator-led groups. This that a commercially viable product Tdiscipline of critical care. It has also been overview of the history and tribulations of emerge at the end. one of the most productive. Founded by a the CCCTG provides some subjective, but The questions posed in investigator- handful of academic intensivists in 1989, hopefully useful insights into the experi- led studies typically arise from either cu- it has grown to a national organization of ence of one such group over the past two riosity, on the one hand, or confusion and more than one hundred members – both decades, synthesizing these into a series controversy on the other. Curiosity drives physicians and nonphysicians – with an of lessons that may prove useful to oth- basic research to determine the structure active research portfolio of more than 35 ers. of DNA, or to understand the processes research programs. It has served as a through which healthcare workers come model for similar groups in Australia and The Research Question to make decisions regarding end-of-life New Zealand, Sweden, and Scotland, and care in critical illness (2). Confusion – Research begins with a question. How for new initiatives in Ireland, South Af- the awareness that several plausible ther- the question is formulated shapes the rica, the Middle East, and the United apeutic options exist, and that clinicians States. subsequent research enterprise. Even more importantly, who poses it has a are divided over which to use – spawns Although investigator-led research studies to determine optimal approaches consortia are well-established in oncol- profound impact on how the enterprise evolves, where it will ultimately go, and to the of the septic patient ogy (1), they are a relative novelty in (3), to assess the relative merits of col- critical care, where industry-run trials whether it will reach that end. loids and crystalloids (4), or to determine still dominate the landscape of multicen- Clinical research undertaken by in- dustry is undertaken with the objective of whether maintenance of normoglycemia bringing a new technology to market, will benefit the critically ill patient (5). and having done that, ensuring and ex- Controversy for the clinician is reformat- From the Department of , and the Interde- panding its commercial niche. Innova- ted as clinical equipoise for the investiga- partmental Division of Critical Care , Univer- tion arises in the technology under study, tor (6), and spurs a structured effort to sity of Toronto, and the Research Laboratory, Li resolve the uncertainty that breeds prac- Ka Shing Knowledge Institute, St. Michael’s Hospital, rather than in the methods by which it is Toronto, Ontario, Canada, and the CLARITY Research studied, or the rigor with which they are tice variability. For such studies, the ul- Group, McMaster University, Hamilton, Ontario, Canada. applied to ensure the most reliable an- timate conclusion is often less important The authors have not disclosed any potential conflicts swer. The question is more than whether than the reliability of the answer. of interest. The research question addressed by For information regarding this article, Email: the technology works; rather the sponsor [email protected] seeks to achieve regulatory approval, and investigator-led trials may be articulated Copyright © 2008 by the Society of Critical Care to maximize the financial return associ- collectively by the consortium or by the Medicine and Lippincott Williams & Wilkins ated with subsequent sales of the prod- individual investigator. A collectively ar- DOI: 10.1097/CCM.0b013e3181921079 uct. The design of the clinical research ticulated research agenda enables a sys-

Crit Care Med 2009 Vol. 37, No. 1 (Suppl.) S165 tematic and iterative approach to advanc- It has proven particularly important in thresholds for transfusing stable, anemic ing knowledge about the clinical developing a generic approach to clinical critically ill patients. The design of an 838 management of a disease process. For research that we term programmatic re- patient multicenter RCT (19) was in- example, the National Surgical Breast Ad- search, and that views research as an formed by preliminary studies that de- juvant Project has undertaken a series of integrated approach to a clinical ques- fined variability in transfusion practice studies over the past three decades to tion, rather than as one, or even a series, using the methods of a scenario-based determine the optimal management of of clinical trials. questionnaire (20) and an observational women with breast cancer, addressing cohort study (21), and a 69-patient pilot such questions as the extent of surgical Programmatic Research: The study to evaluate recruitment feasibility therapy (7), and the need for adjuvant CCCTG Model and clinician compliance with the trans- radiotherapy or chemotherapy. Specific fusion threshold (22). research questions arise from the find- Having decided to focus on the risk of As the CCCTG has matured, this pro- ings of previous studies, and are priori- associated with the use of H2 grammatic research model has evolved to tized by the research collective. A similar blockers for stress ulcer prophylaxis for become the foundation of our approach approach has informed the research its initial venture into collaborative mul- to clinical research (2, 23). Programmatic agenda of the National Institutes of ticenter research, the CCCTG faced the research considers clinical research to be Health-funded Acute Respiratory Distress challenge of designing a protocol to ad- an integrated series of investigations, us- Syndrome Network collaborative that dress the question. That we would under- ing a variety of complementary method- studies the management of critically ill take a multicenter, blinded, randomized ologies, to determine best practice in patients with acute respiratory distress controlled trial was self-evident; however, critical care, and to implement that syndrome (8, 9). key study design issues were immediately knowledge to improve patient outcomes The CCCTG has adopted the alternate apparent. How would we define clinically (Fig. 2). model, undertaking research programs important bleeding and VAP as the two A research program starts with a for- that are brought forward by individual primary outcome measures for the trial? mal structured review of what is known members, and that reflect their unique And as importantly, there was a shared about the research question through the interests and passions. Our inaugural impression that rates of clinically impor- performance of one or more systematic meeting in the spring of 1989 brought tant stress bleeding were decreasing: we reviews or meta-analyses. Since our stud- together about 10 academic intensivists lacked the basic epidemiologic informa- ies typically address the comparative effi- to discuss the possibility of establishing a tion on incidence and risk factors that cacy of two or more available clinical national network for investigator-initi- would enable us to identify an at-risk strategies, it is of fundamental impor- ated clinical research. Participants were population to study and to estimate the tance to quantify practice variability to invited to bring proposals for research size of study population necessary to an- establish that clinical equipoise exists, projects. Three of the six proposals pre- swer our primary question. Thus we de- and to identify plausible approaches to sented addressed the question of the risk cided to undertake preliminary pilot compare in an RCT. Clinician attitudes of -associated pneumonia (VAP) studies to better inform the design of a resulting from the use of acid-reducing definitive trial, including meta-analyses can be probed through scenario-based strategies for stress ulcer prophylaxis, on the role of gastric pH in the etiology of questionnaires, while practice variability and so we decided to undertake a trial, led VAP (11) and on the impact of stress ulcer in the real world is best assessed through by Dr. Deborah Cook, to determine prophylaxis on rates of bleeding (12), and observational studies. For the TRICC whether stress ulcer prophylaxis with su- a large observational study to determine trial, for example, transfusion thresholds cralfate would reduce the risk of pneu- the incidence of, and risk factors for, of 70 g/L in the restrictive arm, and 100 monia associated with the use of the his- stress bleeding in contemporary Cana- g/L in the liberal arm were established on tamine H2 blocker, (10). dian intensive care units (ICUs) (13). A the basis of a scenario-based question- The advantages of an investigator- 1200 patient randomized controlled trial naire and an observational study, and driven research agenda to a clinical re- (RCT) comparing with raniti- represented acceptable, but divergent search consortium are several. It is re- dine (10) generated a series of secondary transfusion thresholds with which clini- sponsive to the needs and interests of the analyses of risk factors for clinically im- cians would feel comfortable. For the Ap- members of the research group and portant gastrointestinal bleeding in me- propriate Antimicrobial Therapy in Criti- shaped over time by the most active chanically ventilated patients (14), of the cal Care (AATICC) program evaluating members. The questions addressed re- burden of illness of upper gastrointestinal the utility of empirical for pa- flect issues of contemporary interest to a bleeding (15) and VAP (16), and of the tients thought to be at intermediate risk community of practitioners, and the re- utility of invasive diagnostic techniques of having a nosocomial infection, a sce- search portfolio changes over time as the for VAP (17), that in turn, provided im- nario-based study revealed striking diver- dominant clinical issues change. It also portant pilot data for the design of a large gence in the use of empirical antibiotics tends to promote research questions that RCT addressing the diagnosis and treat- for three hypothetical patients with pos- go beyond clinical management of a dis- ment of VAP (18) (Fig. 1). sible nosocomial infection (24), and so ease process to focus on the process of We developed this model further demonstrated that although individual research itself. Finally, it is ideally suited within the context of a research program clinicians have strongly held opinions, a to organizations such as the CCCTG that on transfusion of the critically ill. Led by state of community equipoise exists. Ob- do not receive sustained core funding Dr. Paul He´bert, the Transfusion Re- servational studies play an further role in over time, but whose work is supported quirements in Critical Care (TRICC) pro- characterizing the potential study popu- by grants for individual research projects. gram sought to determine optimal lation – determining the prevalence of

S166 Crit Care Med 2009 Vol. 37, No. 1 (Suppl.) Structure and Operations meta-analysis of RCTs bleeding case Membership in the CCCTG is open to healthcare workers with an interest in pilot prognosis control study the care of the critically ill patient. We study bleeding risk currently have approximately 100 mem- factor study bers, including physicians, nurses, phar- macists, respiratory therapists, and train- bleeding ees, and representing both adult and Stress VAP case prognosis/risk factor pediatric critical care. Members are self- control study study Ulcer funded, and membership fees for nonphy- RCT sicians and trainees are reduced. The activities of the CCCTG are coor- VAP risk dinated by an Executive Committee con- bleeding case factor study control study sisting of a Chair, past-Chair, Secretary, Treasurer, 4 Councilors or members-at- VAP diagnosis large (2 each representing adult and pe- VAP study diatric members), and the chairs of the adjudication subcommittees of the organization. study There are 5 such subcommittees. A Grants and Manuscript Review Commit- Pneumonia tee undertakes peer review of all grants or RCT manuscripts being submitted under the auspices of the CCCTG. A Guidelines Com- Figure 1. The Canadian Critical Care Trials Group Research Program in Stress Ulcer Prophylaxis. mittee coordinates activities in the area of Having decided to address the question of whether ablation of gastric acidity during stress ulcer guideline development and knowledge prophylaxis increases the risk of ventilator-associated pneumonia (VAP) as its first clinical study, the translation. A Website Committee main- Canadian Critical Care Trials Group recognized that answering that question required an integrated tains the CCCTG website (www.ccctg.ca), series of studies to synthesize what was known, to establish the epidemiology of stress bleeding, and to achieve consensus on metrics and definitions. Our completed 1200-patient study spawned further while an Ethics Committee coordinates ac- work to look at the attributable costs of stress ulceration, and the diagnosis of VAP, and so set the stage tivities addressing ethical issues in critical for a second large randomized controlled trial (RCT) addressing the specific questions of the diagnosis care research. Finally, an Education Com- and treatment of VAP. mittee runs educational activities for train- ees. In addition, a member of the Clinical Research Coordinators Group - initiated to and risk factors for a particular condition, and the relatively small effect size that address the educational needs and profes- and so facilitating estimates of recruit- any given intervention can plausibly be sional development of critical care research ment rates, and ascertaining anticipated expected to achieve, dictate that defini- coordinators – and a member of the Cana- outcomes in the population of interest (23). tive trials must be large, sometimes re- dian Critical Care Translational Biology Pilot studies permit the investigator to cruiting thousands of patients (4). These Group (CCCTBG) – founded to promote evaluate the feasibility, acceptability, and trials are expensive and complex: prelim- collaborative basic science research using compliance rates of a study protocol, and inary studies such as those described can CCCTG programs as platforms (vide infra) so increase confidence that a larger trial facilitate their successful conduct. Inclu- – also sit on the Executive Committee. powered to assess clinical impact can be sion and exclusion criteria can be more Scientific meetings of the CCCTG are completed successfully. Clinical out- readily defined, and the feasibility of the held three times a year. Two of these comes are of secondary importance. One planned intervention understood. Esti- meetings take place in locales remote from cities to promote camaraderie and can evaluate recruitment rates, assess the mates of sample size and recruitment feasibility of an intervention, determine collegiality, and are held over 2 to 3 days rates can be more reliable, and study compliance with a protocol, evaluate with time set aside for leisure activities; teams can be created. Finally, granting whether blinding is effective, and evalu- the third meeting, held in the fall in as- agencies look favorably upon evidence ate tools for data collection and study sociation with the Critical Care Canada . Pilot studies are also excel- that a study is well thought out and fea- Forum in Toronto, is an intensive day- lent platforms for determining clinician sible. long meeting. Members wishing to pro- perspectives through the administration The completion of a trial does not sig- pose a new research program are re- of questionnaires, for conducting biological nal the end of a research program. To quired to submit an abstract and draft studies to evaluate diagnostic biomarkers, improve patient care, new knowledge protocol in advance of the meeting. The or to use the systematic intervention of a must be adopted into practice, and so proposal is then presented at a meeting clinical trial to address a biological ques- guidelines development and knowledge for discussion and critique, and a poll tion. translation strategies represent the final taken to assess the level of interest of the The randomized controlled trial is the elements of the programmatic research membership in hearing further about the centerpiece of any clinical research pro- model. Inevitably the process of studying project, and in becoming active partici- gram. The inherent heterogeneity en- a question generates important new pants of the research team. The process is countered among critically ill patients, questions, and the cycle continues. quite informal. We have not found a need

Crit Care Med 2009 Vol. 37, No. 1 (Suppl.) S167 questions is broad, and may include nu- Articulate Study Question and ances of the study question, inclusion and Research Objectives exclusion criteria, outcome measures, funding, strategies to increase recruitment, or future research directions. Minutes are taken at these discussions and the investi- gator is provided with a summary of the key What is Currently Known? What is is Current Current Practice? Practice? points that emerge.

Systematic reviews • Scenario-based questionnaire Trials and Tribulations • Pre-clinical • Observational studies • Clinical • Focus groups Since its inception almost two decades ago, the CCCTG has published more than 60 peer-reviewed papers, including nine in the New England Journal of Medicine. In addition, we have published a large Definitions, Metrics and Study Prevalence and Risk Factors Study Definitions, Metrics number of abstracts, as well as peer- andOutcomes Outcomes reviewed manuscripts relating to CCCTG • Systematic reviews programs, although not formally au- • Observational studies • Focus groups thored by the group. We have enjoyed • Consensus processes considerable success with peer review • Pilot observational studies funding agencies, having been successful with 8 of the last 11 grant applications submitted to the Canadian Institutes for Can a Definitive RCT be Health Research. Our current research Undertaken? portfolio includes more than 30 pro- grams, led by more than two dozen dif- Randomized, Controlled Trial ferent investigators; some of these are • Observational prevalence summarized in Table 1. studies A recent survey of member sites pro- • Pilot studies: • Primary trial • Feasibility • Ancillary studies vided insight into current resources, and • Acceptability • Primary analysis the immediate challenges faced by the • Intervention, data • Secondary reports CCCTG. We surveyed 26 centers recruit- forms, metrics ing patients to CCCTG trials (20 adult and 6 pediatric ICUs). These ICUs staffed 609 critical care beds (median 22), and admitted 2698 patients (median 92; range Knowledge Translation 45 to 250) during the study month of April 2005. They recruited patients to an • Knowledge implementation average of 1.7 (range 0–4) industry- • Synthesis and guidelines sponsored trials, 1.4 (range 0–5) CCCTG • Emerging questions for further trials, and 1.7 (range 0–7), investigator- study initiated, non-CCCTG trials. Sites re- Figure 2. Programmatic research: The Canadian Critical Care Trials Group model of clinical research cruited an average of five critically ill entails the answering of a question. While an adequately powered randomized controlled trial (RCT) patients to a research study each month, is a component of this process, the Canadian Critical Care Trials Group has evolved the philosophy that although there was considerable variabil- the process involves multiple steps to systematically evaluate what is known, to characterize clinician ity in recruitment success among units behavior, to determine whether the question can be addressed, and ultimately, to translate the results (Fig. 3). We have set an objective of re- of clinical research so that clinical practice is changed. These elements comprise a programmatic cruiting 12% of patients admitted to a model of clinical research. research endeavor. Only four centers (15%) achieved this target during the study month; it is clear that there is an to reject potential projects that have only in an abbreviated form, and to lead unmet opportunity for greater productiv- achieved a sufficient level of development more intensive discussions with the mem- ity in CCCTG centers. to be presented to the group, but rather bership on a regular basis. These protocol Of the 26 responding centers, 21 have observed that those that fail to discussions are both the scientific highlight (81%) indicated that they employed a re- evolve disappear by attrition, an occur- of our meetings, and for the investigator, search coordinator for ICU-based clinical rence which is relatively uncommon. the most valuable aspect of CCCTG mem- research. Sources of salary support for The principal investigator of any bership. After providing a brief summary of this individual varied. Industry-funded CCCTG-affiliated research program is ex- the project and its progress since the last clinical trials were the most important pected to provide updates on the progress meeting, the investigator poses specific source of salary support: 76% of ICUs of the research at each meeting, even if questions for discussion. The scope of these employing a research coordinator de-

S168 Crit Care Med 2009 Vol. 37, No. 1 (Suppl.) Table 1. The Canadian Critical Care Trials Group research portfolio

Study Lead Investigator Research Focus Status

Appropriate Antimicrobial Therapy Mary-Anne Aarts, John Marshall Empiric therapy for suspected Survey, observational study, and in Critical Care ICU-acquired infection meta-analysis published; pilot RCT complete and under review ABATE Ventilator-Associated John Muscedere, Tasnim Sinuff Knowledge translation of guidelines for Knowledge translation study Pneumonia the prevention of ventilator- funded and in progress associated pneumonia ABLE: Age of Blood Evaluation Paul Hébert, Jacques Lacroix Age of transfused red blood cells Systematic review and pilot RCT completed Definitive 2510 patient trial funded Adrenal Insufficiency in the Kusum Menon Prevalence of adrenal insufficiency and Data collection ongoing Pediatric ICU (PICU) use of steroids in PICU A Fib Sal Kanji Management of new-onset atrial Systematic review complete; fibrillation pilot study recruiting CONSENT Karen Burns Barriers to informed consent for ICU Observational study funded and research to be initiated Early Determination of Neurologic Alexis Turgeon Predictors of outcome in TBI Systematic review completed; Prognosis in Brain Injury observational studies and survey in progress Fluids after Cardiac Surgery Sheldon Magder Colloids vs. crystalloids after cardiac Pilot study completed surgery Hyp-HIT Jamie Hutchison Hypothermia for pediatric head injury 240-patient clinical trial complete; manuscript published, NEJM OSCILLATE Niall Ferguson, Maureen Meade High frequency oscillation for acute Systematic review complete; lung injury pilot study complete PRECISE Lauralyn McIntyre Crystalloids vs colloids in septic Pilot study completed; practice surveys completed; RCT proposal submitted for funding PROTECT Deborah Cook Prophylaxis of venous Multiple observational studies thromboembolism complete Evaluation of bioaccumulation of low molecular weight , accepted for publication Pilot studies of protocol published 3600-patient RCT recruiting REDOXs Daren Heyland Anti-oxidant-enhanced enteral nutrition Pilot study completed; 1200 patient RCT funded and recruiting Toward RECOVER Margaret Herridge Quality of life, caregiver burden, and Pilot studies ongoing neuromuscular sequelae in ICU survivors Resident Work Hours Study Chris Parshuram Models of on-call during critical care Three surveys completed training Survey completed Trial funded and recruiting SLEAP Sangeeta Mehta Sedation Pilot RCT completed; 425-patient RCT funded and recruiting SUGAR: Survival Using Glucose Dean Chittock Tight glucose control Collaboration with Australia and Algorithm New Zealand Intensive Care Society Clinical Trials Group NICE trial; 6100-patient study completed Validation of Severity of Illness David Wensley Validation of PRISM and PIM-2 Data from more than 8000 Scoring Systems in PICUs admissions collected Vasopressin in Pediatric Shock Karen Choong Vasopressin in septic children 69-patient pilot study completed WEAN: Wean Early and Karen Burns, François Lellouche Automated weaning protocol Pilot study recruiting Automatically using New Algorithm

ICU, intensive care unit; RCT, randomized controlled trial; TBI, traumatic brain injury; NEJM, New England Journal of Medicine; PRISM, Pediatric Risk of Mortality; PIM, Pediatric Index of Mortality.

Crit Care Med 2009 Vol. 37, No. 1 (Suppl.) S169 Centre

Mean

0 5 10 15 20 25 30 Percent of Adm itted Patients Recruited to a Study Figure 3. Recruitment rates to clinical studies in Canadian Critical Care Trials Group-participating centers. Across the 25 centers evaluated, the mean number of patients recruited to a clinical study during the month of April 2005 was 5. pended on this source of salary funding, more than one clinical trial. The ethical are seeking support to develop a decen- either in whole or in part. Other sources and statistical issues associated with co- tralized network of methods centers and of coordinator salary support included enrollment of patients in more than one to establish common research data plat- per-patient payments from peer review- clinical trial are complex. They are also forms and monitoring methods for future funded CCCTG trials, institutional sup- the focus of a current research initia- CCCTG-funded trials. port, other academic funds, and clinical tive undertaken by the CCCTG in col- Yet another area of challenge identi- earnings of the ICU staff. Sustainable and laboration with members of the Austra- fied by our needs assessment survey secure funding for research coordinator lia and New Zealand Intensive Care arises through interactions with institu- salary was identified as the single greatest Society (ANZICS) Clinical Trials Group tional research ethics boards (REBs). In- need, and the greatest threat to ongoing (CTG). On the one hand, co-enrollment terventions in the intensive care unit are participation in clinical research. In Can- improves research efficiency and expe- typically urgent, and decisions must be ada, salary funding from peer review dites the answering of questions that can made without delay; study of the manage- granting agencies for funded clinical tri- ultimately improve patient care. On the ment of these urgent clinical problems als is modest, and per-patient reimburse- other hand, co-enrollment may pose an requires that consent be obtained rapidly, ment for CCCTG studies is typically in extra burden on family members who are and patients be randomized expedi- the range of $1000 to $1500 CAN$ per approached for third party consent, and if tiously. Yet by virtue of the problem that patient. In the face of rising pharmacy there is a potential interaction between has brought the patient to the ICU, crit- and other study costs, reimbursement interventions, may alter anticipated rates ically ill patients are rarely able to provide from CCCTG trials is insufficient to meet of the expected outcome (although ran- informed first person consent to partici- salary expenses. Industry-funded studies domization should minimize systematic pate in clinical trials, and consent must typically provide generous reimburse- bias resulting from one or other interven- be obtained from a substitute decision- ment schemes, enabling sites to use over- tion). maker. In upholding a mandate to pro- age generated from these studies to close Other themes emerged as future needs tect vulnerable research subjects, Re- the salary gap. However industry-funded for the CCCTG, and these are shaping our search ethics boards often place studies are an inconsistent source of efforts to secure infrastructural support. restrictions on the obtaining of consent funding, compete for eligible patients, Methodologic support for CCCTG studies for ICU research. Even in the absence of and consume considerable amounts of has been typically provided by the home approval of alternate models of consent coordinator time. Thus finding alternate institution of the principal investigator. such as deferred consent, REB policies means of providing secure funding over While a number of our centers house often prohibit telephone consent or faxed time for research coordinators has well-developed, cutting edge clinical re- consent forms, with the result that po- emerged as the primary need of the search units, it is apparent that research tentially eligible study subjects are ex- CCCTG, and the primary threat to our in the critically ill poses unique method- cluded from participation. Thus another future success. ologic challenges, and benefits from the area of current research interest within the Half of our recruiting sites indicated availability of dedicated methodologists CCCTG is the process of informed consent that they sometimes recruited patients to to optimize research conduct. Thus we for critical care research – another theme

S170 Crit Care Med 2009 Vol. 37, No. 1 (Suppl.) that has opened further collaborative op- sive Care Evaluation-Survival Using Glu- creasingly mature processes of research portunities with the ANZICS CTG. cose Algorithm Regulation study, initi- oversight and mentoring ensure that ated by the ANZICS CTG, and separately CCCTG protocols are well thought out, Clinical Research as a Platform funded by the Canadian Institutes for and scientifically rigorous. These formal for Additional Research Health Research which is evaluating the processes also provide our membership benefits of tight glucose control in an with an ongoing process of informal The process of research in ICUs gen- heterogeneous population of critically ill mentoring and education through lively erates multiple questions that are appro- patients. Enrollment concluded in Au- discussions at our scientific meetings. We priate and important themes for re- gust 2008, after more than 6100 patients hold regular educational sessions for re- search, and investigator-led research had been randomized, making the trial search coordinators, and recently have consortia are ideally positioned to address one of the largest ICU studies ever con- initiated a program for critical care train- these. The CCCTG has used its research ducted, and holding the promise of pro- ees, with the objective of educating the programs as platforms for a number of viding a definitive answer to an important next generation of Canadian critical care such ancillary studies; in a number of area of contemporary controversy. researchers. cases, the initiative was led by a research Preliminary discussions have been Finally, we have successfully nurtured coordinator. held with investigator-led clinical re- a culture of genuine respect and collegi- For example, we analyzed data from a search groups in North America, South ality. This culture has been facilitated by screening log maintained to monitor re- America, Europe, Africa, and Australasia, holding our meetings in out-of-town rec- cruitment into multiple trials in order to under the auspices of the World Federa- reational sites, rather than in airport ho- develop metrics to understand variability tion of Societies of Intensive and Critical tels or convention centers, by building in recruitment rates, and to devise strat- Care Medicine, with the goal of holding recreational time into the scientific egies to maximize enrollment into future regular discussions on themes of com- agenda, and by ensuring ample time in studies (25). Another coordinator-initi- mon interest, and ultimately, of promot- the evening for good food and wine, to ated study evaluated the time spent by a ing global, investigator-led collaborative accompany impassioned conversation. research coordinator in tasks other than critical care research. The remarkable success of the CCCTG data collection during the initiation and argues persuasively for new models of maintenance of a clinical trial (2, 26). Conclusions: The CCCTG and critical care research. Just as investiga- More recently we have initiated research the Future of Investigator-Led tor-led research has become the norm in programs addressing informed consent in Clinical Research In Critical oncology and , so we believe the critically ill, and the need for im- Illness that a shift from commercial to investi- proved outcome measures for use in clin- gator-led clinical research holds the ical trials. Two decades after our first meeting, greatest promise for systematic study of Approximately 7 years ago, a group of the CCCTG finds itself embracing suc- the risks and benefits of what we do, and CCCTG members with active basic sci- cesses that far exceed our fondest early even for high-quality study of novel diag- ence research interests established the hopes. Canadian investigators lead the nostic and therapeutic strategies. The Canadian Critical Care Translational Bi- world in productivity in critical care re- possibilities are truly extraordinary. ology Group to promote collaborative ba- search, whether measured as a function sic and translational research in associa- of population or gross domestic product REFERENCES tion with the CCCTG, and to take (28), and the CCCTG has played an im- advantage of opportunities such as clini- portant role in establishing this preemi- 1. Keating P, Cambrosio A: From screening to cal data collection or the use of a system- nence. The commitment of a talented clinical research: The cure of leukemia and atic study intervention to study the biol- group of experts has been a fundamental the early development of the cooperative on- ogy of critical illness. They have element of our success. However three cology groups, 1955–1966. Bull Hist Med published their first manuscript (27) and additional factors have been of immense 2002; 76:299–334 are initiating further collaborative stud- importance. 2. Cook D, Rocker G, Marshall J, et al: Levels of ies. First, we have chosen to study ques- care in the intensive care unit: A research tions that reflect the daily concerns of program. Am J Crit Care 2006; 15:269–279 3. Rivers E, Nguyen B, Havstad S, et al: Early International Collaboration: practicing intensivists. This has re- goal-directed therapy in the treatment of se- Opportunities and Needs sponded to a previously unmet need in a vere sepsis and . N Engl J Med field in which clinical research initiatives 2001; 345:1368–1377 The CCCTG has served as a model for were dominated by commercial studies 4. Finfer S, Bellomo R, Boyce N, et al: A com- other collaborative, investigator-led clin- undertaken for product registration, or parison of albumin and saline for fluid resus- ical research groups. With their rapidly studies that reflected an individual inves- citation in the intensive care unit. N Engl evolving success, the potential for inter- tigator’s desire to better understand the J Med 2004; 350:2247–2256 national collaborative research has be- clinical state of critical illness. This focus 5. Van den Berghe G, Wilmer A, Hermans G, et come a reality. We have, for example, has struck a responsive chord with grant- al: Intensive insulin therapy in the medical conducted a number of studies in collab- ing agencies, journals, and a new gener- ICU. N Engl J Med 2006; 354:449–461 6. Freedman B: Equipoise and the ethics of oration with the ANZICS CTG, and with ation of clinician investigators who present clinical research. N Engl J Med 1987; 317: investigators in the United States, En- new ideas at our scientific meetings. 141–145 gland, France, Sweden and Saudi Arabia. Second, we have created a collabora- 7. Fisher B, Bauer M, Margolese R, et al: Five This collaborative spirit, for example, has tive structure that combines scientific year results of a randomized clinical trial generated the Normoglycemia in Inten- rigor with intense collegiality. Our in- comparing total mastectomy and segmental

Crit Care Med 2009 Vol. 37, No. 1 (Suppl.) S171 mastectomy with or without radiation in the Care Trials Group. Crit Care Med 1999; sive care unit: a multicentre cohort study. treatment of breast cancer. N Engl J Med 27:2812–2817 Crit Care 1999; 3:57–63 1985; 312:665–673 15. Cook DJ, Griffith LE, Walter SD, et al: The 22. Hebert PC, Wells G, Marshall J, et al: Trans- 8. Brower RG, Matthay MA, Morris A, et al: attributable mortality and length of intensive fusion requirements in critical care. A pilot Ventilation with lower tidal volumes as com- care unit stay of clinically important gastro- study. JAMA 1995; 273:1439–1444 pared with traditional tidal volumes for acute intestinal bleeding in critically ill patients. 23. Cook D, Heyland D, Marshall J, et al: On lung injury and the acute respiratory distress Crit Care 2001; 5(6):368–375 the need for observational studies to design syndrome. N Engl J Med 2000; 342: 16. Heyland DK, Cook DJ, Griffith L, et al: The and interpret randomized trials in ICU pa- 1301–1308 attributable morbidity and mortality of ven- tients: A case study in stress ulcer prophy- 9. National Heart LaBIARDSACTN, Wiedemann tilator-associated pneumonia in the critically laxis. Intensive Care Med 2001; 27:347–354 HP, Wheeler AP, et al: Comparison of two fluid- ill patient. Am J Respir Crit Care Med 1999; 24. Aarts MA, Granton J, Cook DJ, et al: Empiric management strategies in acute lung injury. 159:1249–1256 Antimicrobial Therapy in Critical Illness: Re- N Engl J Med 2006; 354:2564–2575 17. Heyland DK, Cook DJ, Marshall JC, et al: The sults of a Surgical Infection Society Survey. 10. Cook DJ, Guyatt GH, Marshall JC, et al: A clinical utility of invasive diagnostic tech- Surg Infect 2007; 8:329–336 randomized trial of sucralfate versus raniti- niques in the setting of ventilator-associated 25. Foster D, Cook D, Granton J, et al: Use of a dine for stress ulcer prophylaxis in critically pneumonia. Chest 1999; 115:1076–1084 screen log to audit patient recruitment into ill patients. N Engl J Med 1998; 338:791–797 18. Canadian Critical Care Trials Group, Heyland multiple randomized trials in the intensive 11. Cook DJ, Laine LA, Guyatt GH, Raffin TA: D, Dodek P, et al: A randomized trial of care unit. Crit Care Med 2000; 28:867–871 Nosocomial pneumonia and the role of gas- tric pH: A meta-analysis. Chest 1991; 100: diagnostic techniques for ventilator-associ- 26. McDonald E, Clarke F, Dale C, et al: ICU 7–13 ated pneumonia. N Engl J Med 2006; 355: research coordinator activities: A multi- 12. Cook DJ, Witt LG, Cook RJ, et al: Stress 2619–2630 center time in-motion study. J Crit Care ulcer prophylaxis in the critically ill: a 19. Hebert PC, Wells G, Blajchman MA, et al: A 2005; 9:P267 meta-analysis. Am J Med 1991; 91:519–527 multicentre randomized controlled clinical 27. Haqqani AS, Hutchison JS, Ward R, et al: 13. Cook DJ, Fuller H, Guyatt GH, et al: Risk trial of transfusion requirements in critical Biomarkers and Diagnosis; Protein Biomar- factors for gastrointestinal bleeding in crit- care. N Engl J Med 1999; 340:409–417 kers in Serum of Pediatric Patients with Se- ically ill patients. N Engl J Med 1994; 330: 20. Hebert PC, Wells G, Martin C, et al: A Canadian vere Traumatic Brain Injury Identified by 377–381 survey of transfusion practices in critically ICAT-LC-MS/MS. J Neurotrauma 2007; 14. Cook D, Heyland D, Griffith L, et al: Risk ill patients. Crit Care Med 1998; 26: 24(1):54–74 factors for clinically important upper gas- 482–487 28. Michalopoulos A, Bliziotis IA, Rizos M, et al: trointestinal bleeding in patients requiring 21. Hebert PC, Wells G, Martin C, et al: Variation Worldwide research productivity in critical . Canadian Critical in red cell transfusion practice in the inten- care medicine. Crit Care 2005; 9:R258–R265

S172 Crit Care Med 2009 Vol. 37, No. 1 (Suppl.)