Dear Ron and colleagues, A useful analysis, and I guess many would agree with it.
Secular changes are a major issue in comparing clinical trials over time. The UK's ProMISe study is nearing completion and should report towards the end of this year. Regards Julian Professor Julian Bion FRCP FRCA FFICM MD Professor of Intensive Care Medicine, University of Birmingham PA: Mrs Claire Price: [email protected]<mailto:[email protected]%20%20%20%20> Tel: +44 (0)121 371 6816 Chair, Novel Therapies Committee Queen Elizabeth Hospital Birmingham [email protected]<mailto:[email protected]> Associate Non-Executive Director, Worcs Acute Healthcare Trust Chair, Quality Governance Committee [email protected] <mailto:[email protected]%CA> Royal Airforce Civilian Advisor in Intensive Care Medicine [email protected]<mailto:[email protected]> Chair, NICE Acute Medical Emergencies Guideline Development Group National Clinical Guideline Centre Project Manager: [email protected]<mailto:[email protected]> Chair, UK Critical Care Leadership Forum Administrator: [email protected]<mailto:[email protected]> Tel: +44 (0)20 7092 1653 From: Ron Elkin <[email protected]<mailto:[email protected]>> Date: Thursday, 17 July 2014 23:12 To: Sue Beswick <[email protected]<mailto:[email protected]>> Cc: "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>> Subject: Re: [Sepsis Groups] Impact of ProCESS study on your protocols Hi Sue, The study has certainly generated a buzz. For objective, serious students of this disease, however, the study should raise serious concerns about protocol, data, and conclusions. I'm sure these will be addressed in medical and other nursing forums in the months to come. A few comments or questions as examples: 1) "Usual care" has been irrevocably changed since publication of the EGDT study in 2001, as well as guidelines from SSC supported by many of our professional societies. Indeed sepsis management protocols existed in many of the ProCESS hospitals, so the control groups, protocol-based (PB) standard care and usual care, were treated by physicians well versed in EGDT protocols. 2) The mortality rates in each study group were unexpectedly and remarkably low, around 20%, and probably not representative of the mortality rates for similar patients in most US hospitals. As a result of the low mortality rate, some question whether the study was adequately powered to examine differences between study groups, and whether the study is generalizable to 5000 US hospitals. Moreover, why abandon measures that contributed to such impressive mortality reductions? Are we immune to regressive behavior if practice guidelines are relaxed or removed? 3) The protocol instruction for the first 6 hours was to avoid central line placement, CVP measurement, and ScvO2 in both control groups, PB standard care and usual care, unless peripheral access was inadequate. Yet, over 55% of patients in these groups received them for unstated reasons. One might reasonably speculate they were placed for hypotension and administration of vasopressors. Not stated, however, is how often these lines were utilized for CVP measurements that confirmed or guided resuscitation. In the transcript of a recent NQF conference call, available to the public, an author of the study stated CVP measurements were documented in about 1/3 of the control patients but were not used to guide therapy as evidenced by the lack of followup measurements. However, almost any experienced clinician will act similarly on some single measurements - a patient with a CVP of 3 on vasopressors will almost always receive volume. Also not reported are the number of control patients with lines who had ScvO2 measurements, except for the few who received continuous oximetry lines. It also remains possible that blood sample measurements of ScvO2 were utilized in control patients, but this is not addressed in the manuscript. We don't know how often CVP and ScvO2 measurements were made in control patients with central lines before randomization. We don't know how often clinicians acted on CVPs estimated by bedside neck exam, vertical column height of blood in the lines that were inserted, or IVC dimensions and change with respiration. We don't know how many lines, CVPs, and ScvO2s were added in the control groups after the protocol instructions expired at 6 hours. It is still possible and beneficial to rescue an inadequate resuscitation beyond 6 hours. In short, we don't know enough about management of the control groups. 4) Protocol non-adherence was reported in 11.9% but information in the appendix suggests higher. MAP goals were achieved in only 83%. Overall bundle compliance is not reported. In short, we don't know enough about the quality of management in the EGDT group. 5) Not reported are statistical comparisons between all study patients with lines versus without, control patients with lines versus without, and each of the 2 control group. So in summary, care in any of the study groups is not adequately described, and care in the control groups appears to be significantly contaminated by EGDT. I for one do not favor protocol changes on the basis of this study at this time, and I know for a fact that I have a lot of company. Thanks Ron Elkin MD California Pacific Medical Center San Francisco, CA On Thu, Jul 17, 2014 at 6:23 AM, Sue Beswick <[email protected]<mailto:[email protected]>> wrote: Is anyone adapting their protocols with the findings that came out this year with the ProCESS study? We are looking at making some changes. Sue Sue Beswick APRN, MS, CCNS, CCRN CNS Critical Care Greenville Health System 701 Grove Road l Greenville, SC 29605 Office: 864-455-4884<tel:864-455-4884> _______________________________________________ Sepsisgroups mailing list [email protected]<mailto:[email protected]> http://lists.sepsisgroups.org/listinfo.cgi/sepsisgroups-sepsisgroups.org _______________________________________________ Sepsisgroups mailing list [email protected] http://lists.sepsisgroups.org/listinfo.cgi/sepsisgroups-sepsisgroups.org
