Critical Care Checklists, the Keystone Project, And
Total Page:16
File Type:pdf, Size:1020Kb
Brief Reports Critical care checklists, the Keystone Project, and the Office for Human Research Protections: A case for streamlining the approval process in quality-improvement research* Richard H. Savel, MD, FCCM; Evan B. Goldstein, DO; Michael A. Gropper, MD, PhD, FCCP Checklists have been recently promulgated as a method to ipating hospitals in Michigan as well as for not having obtained enhance patient safety and improve outcomes for critically ill informed consent from each patient and clinician involved in the patients. Specifically, recent work performed by researchers from project. This article documents the recent events surrounding the the Johns Hopkins Medical Institutions has demonstrated that the Keystone Project and the response to the actions taken by the addition of checklists to usual care in the intensive care unit is Office for Human Research Protections in the lay press and the associated with a decrease in the incidence of catheter-related new media (Internet and blogs), articulates how a determination bloodstream infections. Initially evaluated at the institutional can be made if a project is quality-improvement, human-subjects level, this effort has been successfully expanded to the state level research, or both, and proposes some solutions to create a as part of the Michigan Keystone Project. Although this work has structured approach to this kind of research in the future. (Crit recently received significant positive attention in the lay press, Care Med 2009; 37:725–728) the Office for Human Research Protections—as they felt that this KEY WORDS: informed consent; quality improvement; human was a research project requiring Institutional Review Board ap- subjects research; Keystone Project; catheter-related blood- proval and informed consent—put the data collection on hold for stream infection; checklists; new media; blogs lack of approval by the Institutional Review Board at the partic- hecklists—important remind- els (3, 4); currently, these initiatives are fections dropped significantly and re- ers providing much-needed beginning to gain attention in the lay mained low (3). These checklists were an structure in complex environ- press (5, 6). attempt to enhance compliance with an ments—are one of the tools The purpose of this article is to review, approach recommended by both the Cen- Crecommended to improve patient safety discuss, and explain the distinction be- ters for Disease Control and Prevention outcomes in the intensive care unit (ICU) tween quality improvement (QI) and QI as well as the Institute for Healthcare (1). Although checklists have been used research, the controversies surrounding Improvement and included the following: to enhance safety in other fields for some the manner in which QI research is cur- hand washing, using full-barrier infec- time (most notably in the aerospace in- rently conducted, and the role of the Of- tion precautions during catheter inser- dustry and the military [2]), their pres- fice for Human Research Protections tion, cleaning the patient’s skin with ence in health care has only recently be- (OHRP) in the structure of these trials— chlorhexidine, avoiding (whenever possi- come prominent. Implementation of ICU specifically presenting the recent case of ble) the femoral site for line placement, checklists has been successfully docu- the Keystone Project in Michigan, and and removing unnecessary catheters in a mented at the institutional and state lev- the temporary cessation of data collection timely fashion. This project was described because of a perceived lack of regulatory as a prospective cohort study with a con- evaluation and proper informed consent current control ICU, and the Institutional *See also p. 791. (IC). In addition, we describe the strong Review Board (IRB) at the Johns Hopkins From the Department of Critical Care Medicine reaction on the Internet to the termina- Medical Institutions approved the study, (RHS), Montefiore Medical Center, Bronx, NY; Depart- tion of this project and propose several with a waiver for IC being granted (3). ment of Surgery (EBG), Maimonides Medical Center, possible solutions to this research conun- In an analogous—albeit much larger— Brooklyn, NY; and Department of Anesthesia/Critical drum. Care Medicine (MAG), University of California San Fran- study, this same group worked with the cisco School of Medicine (UCSF), San Francisco, CA. Michigan Health and Hospital Associa- Dr. Gropper is the Director of Critical Care Medi- The Keystone Project tion to implement checklists in ICUs cine at UCSF, Vice Chairman of the Department of Anesthesia for Clinical Quality, and a Professor of Dr. Peter Pronovost and his group at throughout that entire state (the Key- Anesthesia and Physiology at UCSF. Johns Hopkins have been at the forefront stone Project). Again, Pronovost was able The authors have not disclosed any potential con- in the academic analysis of the use of to demonstrate that the Institute for flicts of interest. Healthcare Improvement/Centers for Dis- For information regarding this article, E-mail: checklists in the ICU in an attempt to [email protected] improve outcomes. Their landmark 2004 ease Control and Prevention checklists Copyright © 2009 by the Society of Critical Care study demonstrated that by adding could be implemented, and their use was Medicine and Lippincott Williams & Wilkins checklists to standard ICU practice, the associated with a persistent decrease in DOI: 10.1097/CCM.0b013e31819541f8 rate of catheter-related bloodstream in- catheter-related bloodstream infections Crit Care Med 2009 Vol. 37, No. 2 725 (and, presumably, saved lives). The Johns cally, concerns were raised that QI they received a letter of complaint alleg- Hopkins IRB made a determination that projects (both current and future) might ing that research was being performed the study would be exempt from IRB eval- be construed as human-subjects research that potentially violated federal regula- uation, and IC was waived (4). Of note, (HuSR) if patient data were collected and tions; they investigated and discovered the project was performed without ap- actions taken by the result of these data. that research was, in fact, being per- proval from local IRBs (local approval was The wording of a statement from the formed without IC; finally, Johns Hop- not sought) as well as in some small OHRP in response to numerous letters to kins postponed any further data collec- hospitals where an IRB did not exist. Ad- the organization did little to alleviate tion until a full investigation could be ministrators in participating Michigan these concerns: completed. There are significant differ- hospitals, however, were aware that this “As stated above, the regulations do ences, however, between this project and project was happening, as it was ap- not apply when institutions are only im- more traditional HuSR, including: 1) the proved by (and actually being per- plementing practices to improve the risks to the patient of being in the project formed with assistance from) the Mich- quality of care. At the same time, if insti- (minimal to zero); 2) the risks of not igan Health and Hospital Association. tutions are planning research activities improving compliance with patient safety The decisions made by the Hopkins IRB examining the effectiveness of interven- guidelines (significant); 3) the impracti- to have the study be exempt from re- tions to improve the quality of care, then cability of performing IC (both for the view and not require local IRB approval the regulatory protections are important patients and potentially the clinicians) were (in retrospect) controversial and to protect the rights and welfare of hu- given the number of hospitals involved, will be discussed further below. man research subjects ...”(11, 16). the necessity of these interventions, and Such vague statements continue to the difficulties of gaining IC in critically Checklists in the Lay Press fuel the strongly worded, somewhat emo- ill patients or their surrogates, in general; tional, academic debate and discourse re- and 4) the fact that the interventions Approximately 1 year after the Key- garding what, in fact, constitutes QI/ being performed could have been (some stone Project results were published, Atul HuSR and what sorts of regulatory would argue should have been) intro- Gawande, MD (a noted Harvard surgeon), protections will be required. duced into clinical practice without ei- wrote a piece entitled “The Checklist” in In mid-February 2008, after much ther IRB approval or IC. These distinc- the December 10, 2007, edition of The pressure from various medical societies, tions are important, and are the primary New Yorker, focusing on the important, including a letter signed by five medical rationale as to why this kind of QI re- often unrecognized, value of the multi- and nursing organizations (17), the search should be handled in a different disciplinary critical care team; specifi- OHRP came out with another statement, regulatory fashion. cally, he expounded on the manner in in which their stand on the issue appears which the aforementioned checklists en- to have changed: hance patient safety in the ICU (5). This “We do not want to stand in the way of Distinguishing QI from HuSR article garnered significant discussion in quality improvement activities that pose both the traditional and the new media minimal risks to subjects. [OHRP] Fundamentally, QI is felt to be an in- (Internet and blogs) (7–11). regulations provide great flexibility and tegral part of the ongoing management of A mere 3 weeks later, in the December should not have inhibited this activity the system for delivering clinical care, 30, 2007, edition of the New York Times, .... Such research would likely have when compared with HuSR, which can be Dr. Gawande broke yet another story re- been eligible for both expedited IRB re- defined as a knowledge-seeking enter- volving around checklists and patient view and a waiver of the informed con- prise that is independent of routine med- safety in the ICU: the OHRP had an- sent requirement.