American Journal of Respiratory and Critical Care Medicine

Guidelines, protocols, and checklists (together called “protocols”) can be immensely helpful in the clinical arena. However, clinicians and health care systems relying on protocols must assess whether benefits are being accrued, whether such “benefits” are real, and whether harm can be detected. These concerns are important because protocols (as opposed to drugs or other interventions) are seldom subjected to rigorous testing; instead, they are often implemented on the basis of belief or the results of simplistic “before and after” studies. We describe two concepts, “protocol misalignment” and “protocol misattribution,” and discuss how understanding these concepts might help improve outcome and prevent unanticipated harm. Ultimately, applying protocols to the same standards of proof as other interventions might increase insight and help ensure “true” patient benefit.

Practice norms in medicine have evolved from a “standard of care” (as per one’s peers) toward guidelines, checklists, and protocols. Guidelines are often considered to be a distillate of the best available evidence assembled by experts in the field and then made available to the practitioner. Guideline writers often suggest that the guidelines are a framework only, and the individual practitioner is free to tailor the guidelines to the particular patient (1). In contrast, checklists are reminders of itemized elements necessary for proper preparation, procedures, or provision of care. All items on a checklist should be confirmed as the clinician reviews the list.

Protocols are dynamic guides based on a series of “if–then” statements: given the responses, the protocol provides explicit instruction about what is to be done in each circumstance. Protocols complement guidelines and checklists and have important advantages and disadvantages in clinical practice (2). Perhaps surprisingly, a large prospective study (∼6,000 patients in 57 intensive care units [ICUs]) reported no net benefit associated with increased protocol use in ICUs (3). Also in acute care, a systematic review of nine studies of safety checklists was unable to draw conclusions about effectiveness (4).

It would seem that protocols must improve care in at least some circumstances. If this assumption is true (and we believe it to be true) then the absence of even a slight overall benefit across large numbers of patients (3) suggests that such benefit might be accompanied (but countered) by protocol-induced harm.

In this Perspective, we discuss two important phenomena that are empirically recognized but have not previously been well defined: “protocol misalignment” and “impact misattribution.” These phenomena, although seldom explicitly discussed, may help explain how protocols might not always achieve the sought-after benefit. These concepts may be especially useful where faith or strong belief are the dominant determinants of protocol use. Because guidelines, protocols, and checklists all strive to direct or modify behavior according to a set of recommendations, they are considered together—for simplicity—as “protocols” in this Perspective.

Protocol misalignment is a mismatch between the context in which a protocol is developed and the context in which it is implemented. Impact misattribution is a mismatch between the proposed—versus actual—reasons offered to explain how a protocol resulted in improved outcome. Such “misalignment” or “misattribution” can occur equally with guidelines, checklists, or protocols.

In the remainder of this Perspective, we review these two concepts and explain how, in certain contexts, misalignment and/or misattribution can result in missed benefit or even adverse outcome. This is important because in many cases, benefit has not necessarily resulted from the implementation of protocols (35) or guidelines (6) in acute care; in some cases, guidelines directed toward pneumonia (7, 8) or mechanical ventilation (9, 10) may have resulted in harm. Indeed, there is evolving concern that the current approach to developing guidelines is inherently flawed (11), and there is solid evidence that where expert subspecialty care is provided in an outpatient setting for at least one specific disease (asthma), the provision of written action plans confers no benefit across a range of possible outcomes (12).

Protocols designed for one set of practitioners might not work well for another. An illustrative example comes from a study of a fluid management protocol in severe sepsis. In a study of more than 3,000 children with severe sepsis in sub-Saharan Africa (13), bolus fluid administration, a standard element of protocols routinely recommended in developed countries, was hypothesized to decrease mortality, where “no bolus” was the usual care in that region. Unfortunately, bolus administration increased mortality (13). A comparable result occurred in a smaller, similar study 3 years later (14).

Whereas the mechanisms of harm in this example (13) are uncertain, it is a fact that many lifesaving aspects of care routinely available in developed countries (e.g., blood transfusion, mechanical ventilation, invasive monitoring) were unavailable to the clinicians in this resource-poor setting. In addition, the disease patterns in the region were different from those where the protocols were developed; for example, children with severe sepsis in sub-Saharan Africa are often malnourished, suffer from malaria, and are profoundly anemic.

Misalignment could also occur for precisely the opposite reasons, where sophisticated experts working in highly resourced and equipped settings might be constrained to work with a protocol that simply does not permit key interventions of which the experts are capable. For example, in a German cardiovascular ICU patients fared worse when care options were restricted by a protocol (15). In this study, patients who had undergone cardiac surgery were randomized to care by expert intensivists using either a rudimentary algorithm that incorporated basic parameters only, or to care by the same experts using a sophisticated algorithm incorporating complex parameters (15). Here it appeared that an overly simplistic protocol “dumbed down” the abilities and scope of the experts.

Another illustration comes from the obstetric literature. The Term Breech Trial (16) concluded that cesarean section is the safest mode of delivery for breech babies at term, and this broadly changed medical practice (17). However, differences in expertise and standards among centers became apparent (18), and these factors appeared to determine the incidence of complications. Indeed, the outcomes from centers that could not provide the highest standard of care were not applicable to much of the developed world (18). Subsequent data (19) failed to replicate the original conclusions (16) and current guidelines, reflecting these insights, indicate that the method of delivery “should depend on the experience of the health provider” (20).

Finally, protocol misalignment is not restricted to mismatch of resources or expertise. One of the most famous “behavior changing” studies in modern critical care was the study of tight glucose control reported from Leuven in 2001 (21). In that study, normalization of glucose in critically ill patients (mostly cardiac surgical) reduced hospital mortality by more than 30% (relative risk; absolute risk decreased by 3.4%). This result was taken up by many medical organizations in the form of protocolized normalization of glucose in critically ill patients (22). A subsequent very large (6,100 patients) multicenter randomized controlled trial reported that tight glucose control did not reduce mortality in broad populations of critically ill patients (23); instead, it increased mortality by more than 10% (relative risk; absolute risk increased by 2.6%). If the multinational data (23) are indeed broadly applicable to mechanically ventilated ICU patients, then implementation of protocols directing glucose normalization might potentially have been responsible for approximately 26,000 additional deaths per year in the United States alone (assuming approximately 1 million ventilated ICU patients per year in the United States). Ironically, in some health care systems, reimbursement had been made contingent on the use of such protocols.

How Misalignment May Cause Harm

Failure to appreciate the differences between the context for which a protocol is developed and where it is deployed may cause harm. Problems include expectations of the clinicians’ expertise, assumptions about the presenting diseases, assessments of the capacity of the health care system, or appreciating the limits of the protocol. If a protocol that is enforced overestimates the abilities of the provider, it will presume expertise or facilities beyond those that exist, and could thereby cause harm by requiring clinicians to try to function above their level of expertise and perform tasks they are unable to do safely.

In contrast, where the enforced protocol is more rudimentary than the capability of the provider (or the system), the resulting level of care may be lower than that of which the clinician is capable; thus, an unduly simplistic protocol could compromise the best possible care available from an expert.

Misalignment may occur because the protocol presumes a certain case mix but is applied to patients with different diseases or syndromes; this could cause unanticipated harm. For example, hyperventilation may be a lifesaving measure in incipient brainstem herniation; however, applying a hyperventilation protocol to patients with traumatic head injury who are not at immediate risk of brainstem herniation worsens outcome (24). This is because hyperventilation causes intense cerebral vasoconstriction, which, while lowering intracranial pressure, also causes cerebral ischemia (25). Thus, if a patient is at imminent risk of death from brainstem herniation, a protocol stipulating this maneuver might be lifesaving, and the risk of cerebral ischemia would be worthwhile. However, where a patient is not at risk from herniation, then a protocol to prevent herniation (by cerebral vasoconstriction) will not help, and the risk of cerebral ischemia is not worthwhile; such a protocol in that context affords the probability of harm without the possibility of benefit.

The mechanism of harm from misalignment of tight glucose control protocols is uncertain (26). However, it seems likely that the higher calorie intake during the Leuven study (vs. many international norms), as well as cointerventions inherent in multiple additional patient assessments, may explain why the protocolized approach reduced mortality in one center (i.e., Leuven) (21) but increased mortality in 42 other centers (i.e., Canada and Australasia) (23).

Knowing whether an intervention causes a given outcome is a cornerstone of medical practice, and is the basis on which any intervention is recommended. Although studies of standardized intervention in critical care are often large, some highly cited studies lack contemporaneous control groups and are not randomized (2729). Given the mounting doubts about the validity of conclusions from even tightly controlled research (30), we must be especially cautious about drawing inference where the studies have obvious gaps in design, however unavoidable. Caution is especially warranted where the interventions are being recommended on a wholesale basis to large populations—particularly where patients are acutely ill and the mortality is high.

Misattribution in Sepsis Therapy

Sepsis is an important cause of mortality in the critically ill, and as part of an ongoing attempt to improve outcomes in patients with sepsis the Surviving Sepsis Campaign advocated a series of guidelines designed to standardize and improve management (31). These guidelines had broad support across multiple professional societies, the Institute for Healthcare Improvement, and others, and have been widely promulgated.

An important claim of benefit was based on the implementation of one such guideline (a “bundle” of interventions) in 2,500 patients (59 institutions); this was followed by a reduction in sepsis-associated mortality from 44 to 39.7%, corresponding to a relative reduction in mortality of 9.8% (29). Because sepsis is extremely common, a “true” reduction of this magnitude resulting from the intervention would have constituted a major breakthrough; indeed, the accompanying editorial suggested that this was the case (32). However, because the study was a “before versus after” design and was not blinded, it is unclear whether the reduced mortality was a result of the implemented guideline, or was due to secular trends or a Hawthorne effect.

Fortunately, the guideline consisted of two elements, “resuscitation” (i.e., fluids, hemodynamic support, antibiotics) and “intervention,” and this enabled interrogation of cause and effect. Each of the “intervention” elements has been individually proved nonbeneficial (corticosteroids, activated protein C) or harmful (tight glucose control) by subsequent high-quality randomized controlled trials, and could individually not have accounted for the reduced mortality. The “resuscitation” elements were implemented with greater frequency during the study, and this was associated with reduced mortality. However, although the lower mortality rate was maintained 1 year later, the implementation of the “resuscitation” elements had reverted to prestudy levels. Thus, implementation of the “resuscitation” elements could not explain the reduced mortality and the explanation must lie elsewhere (e.g., secular trends, Hawthorne effect).

Finally, the magnitude of the effect and the mechanisms seem implausible. The difference in levels of complete implementation (after vs. before) was modest (overall, 10.0 vs. 5.3%) (29); if such an increment in implementation were truly responsible for a relative reduction in mortality of 9.8%, this would suggest that a modest additional increase in implementation might virtually eliminate mortality in severe sepsis.

Misattribution in Perioperative Care

Adverse surgical events are often serious and sometimes lethal, and many may be preventable. The Surgical Safety Checklist is sponsored by the World Health Organization and has been instituted in many countries with the aim of lessening such adverse outcomes. The checklist consists of a series of routine pre- and interoperative checks focusing on universally accepted (and basic) issues including airway assessment, intravenous access, surgical site verification, and so on (28). Its impact was initially assessed in a study conducted across eight centers worldwide (28); the results were impressive: overall, complications (including death) were reduced by almost one-third.

As with the sepsis guideline, it is important to know whether the checklist implementation was responsible for this major reduction in complication rate. If the intervention was responsible for the reduced complication rate, then the sites in which outcome had improved should also have been the sites in which checklist compliance had improved. However, the relationship between improved checklist compliance and lowered complication rates at the individual sites reveals almost complete discordance between cause (checklist compliance) and effect (reduced complication rate) (Table 1; κ statistic, –0.4). Such dissociation between cause and effect indicates that factors other than the checklist elements must account for the lower complication rates.

Table 1. Surgical Safety Checklist: Relationship between Implementation and Outcome

SiteComplications ReducedImplementation IncreasedConcordanceDiscordance

κ = –0.412 (SE, 0.318; 95% confidence interval, –1.000 to 0.212). Based on data from Reference 28.

In each of these cases (28, 29) it is likely that the improved outcomes were real; however, the reasons for benefit were almost certainly not a direct effect of the respective guideline or protocol. Instead, the improved outcomes may have been related to an improving baseline outcome (i.e., “secular” trends) or the complex social and environmental effects of bringing collective attention to a recognized clinical challenge (i.e., a Hawthorne effect) (33, 34). Additional evidence for this comes from a follow-up study of the surgical checklist (>200,000 surgeries, 101 institutions in a developed country with excellent overall standards) that reported no net impact of the checklist on outcomes (35). Indeed, whereas there was no net positive or negative impact of checklist introduction on overall risk of surgical complications across all the hospitals (35), there was a suggestion of benefit in a subset of hospitals (i.e., desired) that was countered by a suggestion of harm in a different subset of hospitals (i.e., potential misalignment).

How Misattribution May Cause Harm

Misattribution per se might not intuitively be expected to cause harm; after all, if a “rule” is implemented and outcomes improve, that hardly amounts to harm. Indeed, in the computer sales industry the so-called screen effect, summarized as “I don’t care how it works, as long as it does work,” may often be considered pragmatic, especially for short-term management in an individual case. However, caution is necessary for two reasons, especially when this thinking is applied to populations of sick patients.

First, what is the true explanation for the observed benefit (e.g., reduced mortality or rate of complication)? This is key because the “true” explanation must be identified so that it can be understood and enhanced. For example, if it was determined that increases in attentiveness or professionalism were the drivers of better outcomes in the surgical checklist study, then maintaining and improving the beneficial outcomes could hinge on “human” factors (education, attention, participation, ownership) more than simply a focus on checking the list.

Second, failure to identify the “active” component of an intervention can result in undue focus on elements that do not work, leading at best to wasted effort and at worst, harm. For example, in the earlier Surviving Sepsis bundles (which are composite protocols with several elements), recognition that many of the bundle elements simply do not work eventually allowed issuance of revised guidelines that excluded such elements. Understanding the active (vs. inactive) component is important for focusing valuable resources and effort. However, it is especially important where performance may be rewarded for compliance with protocols instead of being linked to education, physician presence or effort, or patient outcome. For example, increased compliance with pneumonia treatment protocols has resulted in an unintended collateral rise in the incidence of (and death from) resistant organisms (8) or antibiotic-related Clostridium difficile colitis (7). Similarly, reduced target time to first antibiotic dose has been associated with increased incidence of sepsis misdiagnosis (without reducing actual time to drug administration) (36). Finally, there is an increased risk of inappropriate (false-positive) cardiac catheterization in settings where emergency department personnel activate the cardiac catheter laboratory solely on a protocolized basis (37). In the worse cases, clinicians could be paid for compliance with bundle elements that do not work or cause harm.

Legislating medical care is rare, but seems reasonable where the effectiveness of therapy—and cost of its omission—are especially high, particularly where there are important public health concerns. Severe sepsis can certainly be lethal, and timely therapy is always desirable. However, diagnosis of sepsis is often difficult and beyond resuscitation, source control, and antibiotics, all tested therapies have failed. Nonetheless, in New York State, hospitals are now required to have in place guidelines that mandate protocolized responses to patients with criteria for suspected severe sepsis (38), despite the fact that available screening tools substantially overdiagnose sepsis, and widespread application of resuscitation (e.g., intravenous fluids) or therapy (e.g., antibiotics) can have considerable adverse consequences at the individual and population levels.

Health care leaders and professional societies should dissuade regulators from assuming that legislation will correct gaps in care caused by inadequate education or insufficient professionalism (39). Medical care that is substandard should be dealt with as a matter of competence and not according to rule-based standards.

In general guidelines, protocols and checklists are aimed at improving or maintaining performance, and are an integral part of clinical research designed to discover new advances in performance.

Incorporating Protocols in Research

In research studies, protocolization of any intervention being studied—whether delivered by experts or otherwise—is necessary. This is necessary not because a given protocol is known to be the best care (the fact that it is being studied means that this is not known); instead it is because without explicit information, the reader of the research will not know what was done in the study, and will therefore be unable to draw conclusions (40). Needless to say, an expert clinician using the insights derived from a research study is expected to perform better than when relying solely on the study protocol (40) (otherwise their expertise is noncontributory) and better than they would have before learning the results of the study (otherwise the study insight is noncontributory).

Incorporating Protocols to Maintain or Improve Performance

Protocols or guidelines are generally intended to ensure consistency of care, and where elements of care may be lacking it is hoped that the standardization accompanying protocols or guidelines will raise performance of the lower levels and thereby elevate the overall standards. However, in critical care this seems not to have happened (3). Important assumptions regarding implementation may ignore misalignment or misattribution.

For example, the standards may be low because the levels of expertise or staffing are less than would be anticipated (e.g., reasonably full staffing with expert clinicians). In some jurisdictions understaffing may be common (41), and a protocol may be instituted (e.g., for the detection and management of sepsis in emergency departments) where clinicians lack familiarity with the condition. In this situation, a protocol is present because an expert is not; although this may be preferable to having no expert and no protocol, we believe patients or their relatives should clearly understand where guidelines or protocols are introduced to compensate for inadequate expertise or staffing. The only long-term option to counter inadequacies in expertise or staffing is to invest in clinician education and recruitment.

Protocols or guidelines may also be imposed in situations where expertise is already present. It is important here that the protocol complements—and does not replace—the clinicians’ expertise and their presence. Clinicians and institutions should always seek to understand whether introduction of a standard is truly beneficial, and naturally seek to implement strategies that result in benefit.

Finally, guidelines, protocols, and checklists are usually proposed as “safety nets” for competent clinicians, rather than as replacement tools for insufficient staffing or expertise. The best assurance that this is true is demonstration that the clinicians have the ability to function at a high level without the protocol; if this is not the case, then the protocol may be functioning as a “crutch” and not as a “safety net.”

Clinicians and health care systems relying on guidelines, protocols, or checklists must assess whether benefits are being accrued, and whether such “benefits” are real. Continuous assessment of the effects of such standardization must be capable of detecting potential harm. Appreciating the possibilities of misalignment and misattribution will enhance the ability to understand where and how to standardize care to improve outcomes. Ultimately, applying the same standards of proof to the introduction of protocols, guidelines, or checklists as is the norm with the introduction of new drugs might increase insight and result in “true” patient benefit.

The authors thank Drs. Laurent Brochard and Damon Scales for constructive comments.

1. Carter A. Clinical practice guidelines. CMAJ 1992;147:16491650.
2. Tonelli MR, Curtis JR, Guntupalli KK, Rubenfeld GD, Arroliga AC, Brochard L, Douglas IS, Gutterman DD, Hall JR, Kavanagh BP, et al.; ACCP/ATS/SCCM Working Group. An official multi-society statement: the role of clinical research results in the practice of critical care medicine. Am J Respir Crit Care Med 2012;185:11171124.
3. Sevransky JE, Checkley W, Herrera P, Pickering BW, Barr J, Brown SM, Chang SY, Chong D, Kaufman D, Fremont RD, et al.; United States Critical Illness and Injury Trials Group-Critical Illness Outcomes Study Investigators. Protocols and hospital mortality in critically ill patients: the United States Critical Illness and Injury Trials Group Critical Illness Outcomes Study. Crit Care Med 2015;43:20762084.
4. Ko HC, Turner TJ, Finnigan MA. Systematic review of safety checklists for use by medical care teams in acute hospital settings—limited evidence of effectiveness. BMC Health Serv Res 2011;11:211.
5. Morgan L, Hadi M, Pickering S, Robertson E, Griffin D, Collins G, Rivero-Arias O, Catchpole K, McCulloch P, New S. The effect of teamwork training on team performance and clinical outcome in elective orthopaedic surgery: a controlled interrupted time series study. BMJ Open 2015;5:e006216.
6. Doig GS, Simpson F, Finfer S, Delaney A, Davies AR, Mitchell I, Dobb G; Nutrition Guidelines Investigators of the ANZICS Clinical Trials Group. Effect of evidence-based feeding guidelines on mortality of critically ill adults: a cluster randomized controlled trial. JAMA 2008;300:27312741.
7. Polgreen PM, Chen YY, Cavanaugh JE, Ward M, Coffman S, Hornick DB, Diekema DJ, Herwaldt LA. An outbreak of severe Clostridium difficile–associated disease possibly related to inappropriate antimicrobial therapy for community-acquired pneumonia. Infect Control Hosp Epidemiol 2007;28:212214.
8. Kett DH, Cano E, Quartin AA, Mangino JE, Zervos MJ, Peyrani P, Cely CM, Ford KD, Scerpella EG, Ramirez JA; Improving Medicine through Pathway Assessment of Critical Therapy of Hospital-Acquired Pneumonia (IMPACT-HAP) Investigators. Implementation of guidelines for management of possible multidrug-resistant pneumonia in intensive care: an observational, multicentre cohort study. Lancet Infect Dis 2011;11:181189.
9. Sinuff T, Cook DJ, Randall J, Allen CJ. Evaluation of a practice guideline for noninvasive positive-pressure ventilation for acute respiratory failure. Chest 2003;123:20622073.
10. Hill NS. Practice guidelines for noninvasive positive-pressure ventilation: help or hindrance? Chest 2003;123:17841786.
11. Classen DC, Mermel LA. Specialty society clinical practice guidelines: time for evolution or revolution? JAMA 2015;314:871872.
12. Sheares BJ, Mellins RB, Dimango E, Serebrisky D, Zhang Y, Bye MR, Dovey ME, Nachman S, Hutchinson V, Evans D. Do patients of subspecialist physicians benefit from written asthma action plans? Am J Respir Crit Care Med 2015;191:13741383.
13. Maitland K, Kiguli S, Opoka RO, Engoru C, Olupot-Olupot P, Akech SO, Nyeko R, Mtove G, Reyburn H, Lang T, et al.; FEAST Trial Group. Mortality after fluid bolus in African children with severe infection. N Engl J Med 2011;364:24832495.
14. Andrews B, Muchemwa L, Kelly P, Lakhi S, Heimburger DC, Bernard GR. Simplified severe sepsis protocol: a randomized controlled trial of modified early goal-directed therapy in Zambia. Crit Care Med 2014;42:23152324.
15. Goepfert MS, Richter HP, Zu Eulenburg C, Gruetzmacher J, Rafflenbeul E, Roeher K, von Sandersleben A, Diedrichs S, Reichenspurner H, Goetz AE, et al. Individually optimized hemodynamic therapy reduces complications and length of stay in the intensive care unit: a prospective, randomized controlled trial. Anesthesiology 2013;119:824836.
16. Hannah ME, Hannah WJ, Hewson SA, Hodnett ED, Saigal S, Willan AR; Term Breech Trial Collaborative Group. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet 2000;356:13751383.
17. Hogle KL, Kilburn L, Hewson S, Gafni A, Wall R, Hannah ME. Impact of the International Term Breech Trial on clinical practice and concerns: a survey of centre collaborators. J Obstet Gynaecol Can 2003;25:1416.
18. Glezerman M. Five years to the Term Breech Trial: the rise and fall of a randomized controlled trial. Am J Obstet Gynecol 2006;194:2025.
19. Goffinet F, Carayol M, Foidart JM, Alexander S, Uzan S, Subtil D, Bréart G; PREMODA Study Group. Is planned vaginal delivery for breech presentation at term still an option? Results of an observational prospective survey in France and Belgium. Am J Obstet Gynecol 2006;194:10021011.
20. ACOG Committee on Obstetric Practice. ACOG Committee Opinion No. 340. Mode of term singleton breech delivery. Obstet Gynecol 2006;108:235237.
21. van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, Vlasselaers D, Ferdinande P, Lauwers P, Bouillon R. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:13591367.
22. Kavanagh BP. Glucose in the ICU—evidence, guidelines, and outcomes. N Engl J Med 2012;367:12591260.
23. Finfer S, Chittock DR, Su SY, Blair D, Foster D, Dhingra V, Bellomo R, Cook D, Dodek P, Henderson WR, et al.; NICE-SUGAR Study Investigators. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360:12831297.
24. Muizelaar JP, Marmarou A, Ward JD, Kontos HA, Choi SC, Becker DP, Gruemer H, Young HF. Adverse effects of prolonged hyperventilation in patients with severe head injury: a randomized clinical trial. J Neurosurg 1991;75:731739.
25. Laffey JG, Kavanagh BP. Hypocapnia [review]. N Engl J Med 2002;347:4353.
26. Kavanagh BP, McCowen KC. Clinical practice: glycemic control in the ICU. N Engl J Med 2010;363:25402546.
27. Pronovost P, Needham D, Berenholtz S, Sinopoli D, Chu H, Cosgrove S, Sexton B, Hyzy R, Welsh R, Roth G, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med 2006;355:27252732.
28. Haynes AB, Weiser TG, Berry WR, Lipsitz SR, Breizat AH, Dellinger EP, Herbosa T, Joseph S, Kibatala PL, Lapitan MC, et al.; Safe Surgery Saves Lives Study Group. A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl J Med 2009;360:491499.
29. Ferrer R, Artigas A, Levy MM, Blanco J, González-Díaz G, Garnacho-Montero J, Ibáñez J, Palencia E, Quintana M, de la Torre-Prados MV; Edusepsis Study Group. Improvement in process of care and outcome after a multicenter severe sepsis educational program in Spain. JAMA 2008;299:22942303.
30. Ioannidis JP. Why most published research findings are false. PLoS Med 2005;2:e124.
31. Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R, Reinhart K, Angus DC, Brun-Buisson C, Beale R, et al.; International Surviving Sepsis Campaign Guidelines Committee; American Association of Critical-Care Nurses; American College of Chest Physicians; American College of Emergency Physicians; Canadian Critical Care Society; European Society of Clinical Microbiology and Infectious Diseases; European Society of Intensive Care Medicine; European Respiratory Society; International Sepsis Forum; Japanese Association for Acute Medicine; Japanese Society of Intensive Care Medicine; Society of Critical Care Medicine; Society of Hospital Medicine; Surgical Infection Society; World Federation of Societies of Intensive and Critical Care Medicine. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Crit Care Med 2008;36:296327.
32. Kahn JM, Bates DW. Improving sepsis care: the road ahead. JAMA 2008;299:23222323.
33. Dixon-Woods M, Bosk CL, Aveling EL, Goeschel CA, Pronovost PJ. Explaining Michigan: developing an ex post theory of a quality improvement program. Milbank Q 2011;89:167205.
34. Bion J, Richardson A, Hibbert P, Beer J, Abrusci T, McCutcheon M, Cassidy J, Eddleston J, Gunning K, Bellingan G, et al.; Matching Michigan Collaboration & Writing Committee. “Matching Michigan”: a 2-year stepped interventional programme to minimise central venous catheter–blood stream infections in intensive care units in England. BMJ Qual Saf 2013;22:110123.
35. Urbach DR, Govindarajan A, Saskin R, Wilton AS, Baxter NN. Introduction of surgical safety checklists in Ontario, Canada. N Engl J Med 2014;370:10291038.
36. Welker JA, Huston M, McCue JD. Antibiotic timing and errors in diagnosing pneumonia. Arch Intern Med 2008;168:351356.
37. Larson DM, Menssen KM, Sharkey SW, Duval S, Schwartz RS, Harris J, Meland JT, Unger BT, Henry TD. “False-positive” cardiac catheterization laboratory activation among patients with suspected ST-segment elevation myocardial infarction. JAMA 2007;298:27542760.
38. New York State Department of Health. Hospital sepsis protocols. 2013 May 1 [accessed 2015 Oct 3]. Available from:
39. Jacobson PD. Transforming clinical practice guidelines into legislative mandates: proceed with abundant caution. JAMA 2008;299:208210.
40. Tobin MJ. Of principles and protocols and weaning. Am J Respir Crit Care Med 2004;169:661662.
41. Mohr NM, Collier J, Hassebroek E, Groth H. Characterizing critical care physician staffing in rural America: a description of Iowa intensive care unit staffing. J Crit Care 2014;29:194198.
Correspondence and requests for reprints should be addressed to Brian P. Kavanagh, M.B., Department of Critical Care Medicine, Hospital for Sick Children, 555 University Avenue, Toronto, ON, M5G 1X8 Canada. E-mail:

Supported by research funding from the Canadian Institutes of Health Research (CIHR) (B.P.K.); B.P.K. holds the Dr. Geoffrey Barker Chair in Critical Care Research.

Author Contributions: B.P.K. and M.N. both contributed to the conception, design, and writing of the manuscript.

Originally Published in Press as DOI: 10.1164/rccm.201502-0314CP on September 22, 2015

Author disclosures are available with the text of this article at


No related items
Comments Post a Comment

New User Registration

Not Yet Registered?
Benefits of Registration Include:
 •  A Unique User Profile that will allow you to manage your current subscriptions (including online access)
 •  The ability to create favorites lists down to the article level
 •  The ability to customize email alerts to receive specific notifications about the topics you care most about and special offers
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record