Full Text View
Protocol and statistical analysis plan for the phase 3 randomised controlled Treatment of Invasively Ventilated Adults with Early Activity and Mobilisation (TEAM III) trial
Jeffrey J Presneill , Rinaldo Bellomo , Kathy Brickell, Heidi Buhr , Belinda J Gabbe , Doug W Gould , Meg Harrold, Alisa M Higgins , Sally Hurford , Theodore Iwashyna , Ary Serpa Neto , Alistair Nichol , Stefan J Schaller, Janani Sivasuthan, Claire Tipping , Steven Webb, Paul Young , Carol L Hodgson , The TEAM Study Investigators and the Australian and New Zealand Intensive Care Society Clinical Trials Group
Crit Care Resusc 2021; 23 (3): 262-272
- Jeffrey J Presneill 1, 2
- Rinaldo Bellomo 1, 3
- Kathy Brickell 4
- Heidi Buhr 5
- Belinda J Gabbe 6
- Doug W Gould 7
- Meg Harrold 8, 9
- Alisa M Higgins 1
- Sally Hurford 10
- Theodore Iwashyna 11, 12
- Ary Serpa Neto 1, 13
- Alistair Nichol 1, 14
- Stefan J Schaller 15
- Janani Sivasuthan 1
- Claire Tipping 16
- Steven Webb 1, 17
- Paul Young 12, 17, 18
- Carol L Hodgson 1, 16
- The TEAM Study Investigators and the Australian and New Zealand Intensive Care Society Clinical Trials Group
OBJECTIVE: To describe the protocol and statistical analysis plan for the Treatment of Invasively Ventilated Adults with Early Activity and Mobilisation (TEAM III) trial.
DESIGN: An international, multicentre, parallel-group, randomised controlled phase 3 trial.
SETTING: Intensive care units (ICUs) in Australia, New Zealand, Germany, Ireland, the United Kingdom and Brazil.
PATIENTS: 750 adult patients expected to receive mechanical ventilation for more than 48 hours.
INTERVENTIONS: Early activity and mobilisation delivered to critically ill patients in an ICU for up to 28 days compared with standard care.
MAIN OUTCOME MEASURES: The primary outcome is the number of days alive and out of hospital at 180 days after randomisation. Secondary outcomes include ICU-free days, ventilator-free days, delirium-free days, all-cause mortality at 28 and 180 days after randomisation, and functional outcome at 180 days after randomisation.
RESULTS: Recruitment at 46 research sites passed 576 patients in March 2021. Final collection of all 180-day outcome data for the target of 750 patients is anticipated by May 2022.
CONCLUSIONS: Consistent with international guidelines, a detailed protocol and prospective analysis plan has been developed for the TEAM III trial. This plan specifies the statistical models for evaluating primary and secondary outcomes, defines covariates for adjusted analyses, and defines methods for exploratory analyses. Application of this protocol and statistical analysis plan to the forthcoming TEAM III trial will facilitate unbiased analyses of the clinical data collected.
TRIAL REGISTRATION: ClinicalTrials.gov identifier NCT03133377.
- Iwashyna TJ. Survivorship will be the defining challenge of critical care in the 21st century. Ann Intern Med 2010; 153: 204-5
- Herridge MS, Tansey CM, Matte A, et al. Functional disability 5 years after acute respiratory distress syndrome. N Engl J Med 2011; 364: 1293-304
- Iwashyna TJ, Ely EW, Smith DM, Langa KM. Long-term cognitive impairment and functional disability among survivors of severe sepsis. JAMA 2010; 304: 1787-94
- Adler J, Malone D. Early mobilization in the intensive care unit: a systematic review. Cardiopulm Phys Ther J 2012; 23: 5-13
- Zhang L, Hu W, Cai Z, et al. Early mobilization of critically ill patients in the intensive care unit: a systematic review and meta-analysis. PLoS One 2019; 14: e0223185
- Hodgson CL, Stiller K, Needham DM, et al. Expert consensus and recommendations on safety criteria for active mobilization of mechanically ventilated critically ill adults. Crit Care 2014; 18: 658
- Hodgson CL, Bailey M, Bellomo R, et al. A binational multicenter pilot feasibility randomized controlled trial of early goal-directed mobilization in the ICU. Crit Care Med 2016; 44: 1145-52
- Craig P, Dieppe P, Macintyre S, et al. Developing and evaluating complex interventions: the new Medical Research Council guidance. BMJ 2008; 337: a1655
- Hoffmann TC, Glasziou PP, Boutron I, et al. Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ 2014; 348: g1687
- Tipping CJ, Bailey MJ, Bellomo R, et al. The ICU Mobility Scale has construct and predictive validity and is responsive. A multicenter observational study. Ann Am Thorac Soc 2016; 13: 887-93
- Hodgson C, Needham D, Haines K, et al. Feasibility and inter-rater reliability of the ICU Mobility Scale. Heart Lung 2014; 43: 19-24
- Team Study Investigators; Hodgson C, Bellomo R, et al. Early mobilization and recovery in mechanically ventilated patients in the ICU: a bi-national, multi-centre, prospective cohort study. Crit Care 2015; 19: 81
- Australian Clinical Trials Alliance. Guidance on implementability. Melbourne: ACTA, 2019
- Myles PS, Shulman MA, Heritier S, et al. Validation of days at home as an outcome measure after surgery: a prospective cohort study in Australia. BMJ Open 2017; 7: e015828
- Jerath A, Austin PC, Wijeysundera DN. Days alive and out of hospital: validation of a patient-centered outcome for perioperative medicine. Anesthesiology 2019; 131: 84-93
- Tipping CJ, Harrold M, Holland A, et al. The effects of active mobilisation and rehabilitation in ICU on mortality and function: a systematic review. Intensive Care Med 2017; 43: 171-83
- Cuthbertson BH, Elders A, Hall S, et al. Mortality and quality of life in the five years after severe sepsis. Crit Care 2013; 17: R70
- Dowdy DW, Eid MP, Dennison CR, et al. Quality of life after acute respiratory distress syndrome: a meta-analysis. Intensive Care Med 2006; 32: 1115-24
- Hodgson CL, Udy AA, Bailey M, et al. The impact of disability in survivors of critical illness. Intensive Care Med 2017; 43: 992-1001
- Ustun TB, Chatterji S, Kostanjsek N, et al. Developing the World Health Organization Disability Assessment Schedule 2.0. Bull World Health Organ 2010; 88: 815-23
- Hopkins RO, Suchyta MR, Kamdar BB, et al. Instrumental activities of daily living after critical illness: a systematic review. Ann Am Thorac Soc 2017; 14: 1332-43
- Lawton MP, Brody EM. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist 1969; 9: 179-86
- Pandharipande PP, Girard TD, Jackson JC, et al. Long-term cognitive impairment after critical illness. N Engl J Med 2013; 369: 1306-16
- Rabiee A, Nikayin S, Hashem MD, et al. Depressive symptoms after critical illness: a systematic review and meta-analysis. Crit Care Med 2016; 44: 1744-53
- Weiss DS. The Impact of Event Scale–Revised. In: Wilson JP, Keane TM, editors. Assessing psychological trauma and PTSD: a practitioner’s handbook. 2nd ed. New York: Guilford Press, 2004
- US Food and Drug Administration. Multiple endpoints in clinical trials: guidance for industry. Silver Spring: FDA, 2017. https://www.fda.gov/media/102657/download (viewed July 2021)
- International Council on Harmonisation. ICH harmonised guideline: addendum on estimands and sensitivity analysis in clinical trials to the guideline on statistical principles for clinical trials (ICH E9(R1)). https://database.ich.org/sites/default/files/E9-R1_Step4_Guideline_2019_1203.pdf (viewed July 2021)
- Schulz KF, Altman DG, Moher D; CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ 2010; 340: c332
- Hosmer DW, Lemeshow S, Sturdivant RX. Applied logistic regression. Hoboken: Wiley, 2013: 1-500
- Vittinghoff E, Glidden DV, Shiboski SC, McCulloch CE. Regression methods in biostatistics: linear, logistic, survival, and repeated measures models. Boston: Springer, 2012
- Grambsch PM, Therneau TM. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 1994; 81: 515-26
- Koenker R. Quantile regression. Cambridge: Cambridge University Press, 2005
- R Core Team. R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing, 2021
- Koenker R. quantreg: quantile regression. R package version 5.86. https://CRAN.R-project.org/package=quantreg (viewed July 2021)
- Hagemann A. Cluster-robust bootstrap inference in quantile regression models. J Am Stat Assoc 2017; 112: 446-56
- European Medicines Agency Committee for Medicinal Products for Human Use. Guideline on adjustment for baseline covariates (draft). 2013. https://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2013/06/WC500144946.pdf (viewed July 2021)
- Jakobsen JC, Gluud C, Wetterslev J, Winkel P. When and how should multiple imputation be used for handling missing data in randomised clinical trials — a practical guide with flowcharts. BMC Med Res Methodol 2017; 17: 162
- Little RJ, D’Agostino R, Cohen ML, et al. The prevention and treatment of missing data in clinical trials. N Engl J Med 2012; 367: 1355-60
- Marshall A, Altman DG, Holder RL, Royston P. Combining estimates of interest in prognostic modelling studies after multiple imputation: current practice and guidelines. BMC Med Res Methodol 2009; 9: 57
- Geraci M. Estimation of regression quantiles in complex surveys with data missing at random: an application to birthweight determinants. Stat Methods Med Res 2016; 25: 1393-421
- Vidoni ML, Reininger BM, Lee M. A comparison of mean-based and quantile regression methods for analyzing self-report dietary intake data. J Probab Stat 2019; 2019: 9750538
- Angus DC, Derde L, Al-Beidh F, et al; Writing Committee for the REMAP-CAP Investigators. Effect of hydrocortisone on mortality and organ support in patients with severe COVID-19: the REMAP-CAP COVID-19 corticosteroid domain randomized clinical trial. JAMA 2020; 324: 1317-29
- Ananth CV, Kleinbaum DG. Regression models for ordinal responses: a review of methods and applications. Int J Epidemiol 1997; 26: 1323-33
- Armstrong BG, Sloan M. Ordinal regression models for epidemiologic data. Am J Epidemiol 1989; 129: 191-204
- Agresti A. Modelling ordered categorical data: recent advances and future challenges. Stat Med 1999; 18: 2191-207
- Brock GN, Barnes C, Ramirez JA, Myers J. How to handle mortality when investigating length of hospital stay and time to clinical stability. BMC Med Res Methodol 2011; 11: 144
- Wang R, Lagakos SW, Ware JH, et al. Statistics in medicine — reporting of subgroup analyses in clinical trials. N Engl J Med 2007; 357: 2189-94
- Rockwood K, Song X, MacKnight C, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ 2005; 173: 489-95
- RECOVERY Collaborative Group; Horby P, Lim WS, Emberson JR, et al. Dexamethasone in hospitalized patients with Covid-19. N Engl J Med 2021; 384: 693-704
- Ustun TB, Kostanjesek N, Chatterji S, Rehm J; World Health Organization. Measuring health and disability: manual for WHO Disability Assessment Schedule (WHODAS 2.0). Geneva: WHO, 2010
- Knott RJ, Harris A, Higgins A, et al. Cost-effectiveness of erythropoietin in traumatic brain injury: a multinational trial-based economic analysis. J Neurotrauma 2019; 36: 2541-8
- Moher D, Hopewell S, Schulz KF, et al. CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials. BMJ 2010; 340: c869
- Ye C, Beyene J, Browne G, Thabane L. Estimating treatment effects in randomised controlled trials with non-compliance: a simulation study. BMJ Open 2014; 4: e005362
- Lehmann EL. Nonparametrics: statistical methods based on ranks. New York: Springer-Verlag, 2006
- ARISE Investigators; ANZICS Clinical Trials Group; Peake SL, Delaney A, Bailey M, et al. Goal-directed resuscitation for patients with early septic shock. N Engl J Med 2014; 371: 1496-506
- Jennison C, Turnbull BW. Group sequential methods with applications to clinical trials. Boca Raton: Chapman and Hall/CRC, 2000: pp 1-390
- Cytel. East 6.5. Statistical software for the design, simulation and monitoring clinical trials. Cambridge: Cytel, 2018
- SAS Institute. SAS 9.4. Cary: SAS Institute, 2017
- Stata Corporation. Stata statistical software: release 16. 16.1 ed. College Station: StataCorp, 2019
- International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use. Integrated addendum to ICH E6(R1): guideline for good clinical practice E6(R2). 2016. https://database.ich.org/sites/default/files/E6_R2_Addendum.pdf (viewed July 2021)
Sensitivity analyses of the primary outcome
Missing data. If more than 5% of the primary outcome data are missing, complete case analyses will be accompanied by further sensitivity analyses using two SD best–worst case, two SD worst–best case, and multiple imputation under the assumption of data missing at random. 37, 38, 39, 40
Grid of quantiles. The quantile regression TEAM intervention effect estimate, adjusted according to the covariate(s) included in the primary outcome model, will be computed on a selected discrete set of quantiles within the 0.01 ≤ τ ≤ 0.99 range to summarise any non-constant TEAM intervention effect within the DAOH180upper and lower tails. These results will be reported graphically and/or in tables that also show, for comparison, the mean TEAM intervention effect estimate across all quantiles returned from a similarly constructed linear mixed-effect model. 35, 40, 41
Results at each research site. The differential primary outcome according to randomised treatment group will be tabulated according to research site.
Ordinal categorical analysis. The differential effect of treatment will be estimated within a proportional odds cumulative logistic model treating DAOH180as an ordinal categorical random variable. 42
Time-to-event analyses. The time to events of interest, including mortality to 180 days, will be assessed using Kaplan–Meier plots and log-rank tests. These will be supplemented by Cox proportional hazard regression models including variables as for the primary outcome model, with the proportional hazards assumption assessed using Schoenfeld residuals,31 as described above. The time to discharge according to treatment group will be assessed using a competing risks model with death before discharge treated as a competing risk. 46
Other binary outcomes. Other binary variables, including adverse events, will be analysed with logistic regression models to estimate odds ratios with 95% CIs.
Pre-specified subgroup analyses of the primary outcome.Possible heterogeneity of treatment effect in the following pre-specified subgroups will be evaluated by tests of interaction between each subgroup and the study treatment in the models described above: 47
- baseline WHODAS 2.0 category (no and mild disability versus moderate, severe and complete disability);
- age (dichotomised at the full analysis set median age);
- diagnosis (sepsis versus trauma versus other);
- severity of illness (dichotomised at the full analysis set median APACHE [Acute Physiology and Chronic Health Evaluation] III score); and
- frailty (a binary reduction of the seven-point Clinical Frailty Scale at 1–4 [well and vulnerable] versus 5–7 [frail]).
Rockwood K, Song X, MacKnight C, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ 2005; 173: 489-95
Outcomes at 28 days. Outcomes at 28 days are listed in Table 2. The scores of ICU-free, ventilator-free and delirium-free days evaluated at 28 days after randomisation will be modelled as continuous random variables in the same way as for the primary analysis. These scores will be zero for patients who die before Day 28.
Health-related quality of life. Health-related quality of life (HRQOL), health problems in each domain of the EQ-5D-5L and functional outcomes assessed at 180 days will be summarised as means (with SDs), medians (with interquartile range) or as proportions. Unadjusted differences in HRQOL between groups will be assessed initially using the two-sample ttest or Wilcoxon rank-sum test, or the Fisher exact test for proportions, as appropriate. HRQOL measures will further be evaluated using multivariable linear regression (or quantile regression where normality assumptions are violated) with adjustment for clustering by site and potential baseline imbalance (P < 0.05). These outcomes include the EQ-5D-5L utility score, EuroQol visual analogue scale (EQ-VAS) score (range, 0–100), ADL score (range, 0–20), IADL score (range, 0–8), WHODAS 2.0 12L score (percentage), MOCA-Blind score (range, 0–22), HADS score (range, 0–21 for either anxiety or depression), and the IES-R score (range, 0–88). EQ-5D-5L utility scores will be calculated using an Australian or UK value set if available at the time of analysis, or a crosswalk value set where an appropriate EQ-5D-5L value set is not available. The WHODAS 2.0 12L percentage scores will be used to categorise patients into five mutually exclusive disability categories at 180 days: no disability (0–4%); mild disability (5–24%); moderate disability (25–49%); severe disability (50–95%); and complete disability (96–100%). 50
Cost-effectiveness analyses. Cost-effectiveness analyses will be conducted from a health care payer perspective using a within-trial time horizon. Incremental cost-effectiveness ratios will be calculated as a cost per additional quality-adjusted life-year and cost per additional day alive and out of hospital gained at Day 180 for early activity and mobilisation compared with standard care. The EQ-5D-5L, administered 6 months after randomisation, will enable utilities to be determined and subsequent calculation of quality-adjusted life-years (QALYs). Costs will be determined based on resource use during the intensive care, acute care and post-acute care periods up to Day 180 and will include costs of the intervention where appropriate. To address the issue of transferability in multinational trials, we will adjust for heterogeneity across geographic regions (which may arise, for example, owing to differences in treatment patterns) and obtain estimates for costs and effects in specific regions of the study.
All costs will be converted to a common currency using purchasing power parity statistics from the Organisation for Economic Co-operation and Development. Health care costs will be compared between regions using the same methods as used in previous economic evaluations of multinational clinical trials by our group. 51
Per-protocol analyses. It is well understood that trial protocols may not be followed fully for some trial participants. 52
Study monitors from the ANZIC Research Centre, the Medical Research Institute of New Zealand and Intensive Care National Audit and Research Centre in the UK will monitor all research sites in person, with the assistance of a German-speaking trial monitor at sites in Germany, a Portuguese-speaking monitor at sites in Brazil and a trial monitor from the Irish Critical Care Clinical Trials Network at sites in Ireland. During monitoring visits, multiple aspects of data validity are checked, including consent documents, inclusion and exclusion criteria, adverse events, and all admission and discharge dates required to calculate primary outcome data and important daily data. An independent data and safety monitoring committee (DSMC) continues to oversee the quality of the trial and has access to trial outcome and accumulated safety data, including the differential proportions of total mortality. Further details are available in the Online Appendix, C.
Sample size, power and interim analysis schedule
- differential all-cause mortality at hospital discharge, censored at Day 28;
- the differential number of days alive and not in hospital to Day 28, with a score of zero allocated to any survivors who remained in hospital at Day 28; and
- the differential number of days alive and not in hospital to Day 180 for those with available data.
Safety and adverse event analyses
Ethics and informed consent
Compliance with Good Clinical Practice requirements
The potential benefits of a bundle of activities comprising early activity and mobilisation during prolonged invasive mechanical ventilation are supported by scientific rationale. This intervention has the potential to shorten ventilator dependence, decrease mortality, and reduce physical and cognitive functional decline associated with illnesses needing prolonged mechanical ventilation in an ICU.
The TEAM III trial is designed to detect an important beneficial effect of early activity and mobilisation, if one exists, while minimising potential risks. Application of the statistical analysis plan within this TEAM III trial protocol will facilitate evaluation of these important clinical data and support confidence in the subsequent generalisation of its findings. The aim of the TEAM III trial is to provide definitive guidance for clinicians regarding the true efficacy and safety of a bundle of activities comprising early activity and mobilisation in the management of critical illness in adults.
Acknowledgements: This investigator-initiated research is funded in Australia by the National Health and Medical Research Council (GTN 1120319) and in New Zealand by the Health Research Council of New Zealand (GTN 19/021). Trial design, data collection and analyses are being undertaken independent of the funding bodies. Trial coordination is being led by the Australian and New Zealand Intensive Care Research Centre, part of Monash University in Melbourne, in collaboration with the Medical Research Institute of New Zealand and the UK’s Intensive Care National Audit and Research Centre. The trial is endorsed by the Australian and New Zealand Intensive Care Society Clinical Trials Group. The members of the TEAM management committee and the TEAM trial sites and investigators are listed in the Online Appendix, A and B respectively).