Triggering Management for Quality Improvement

The policy community has been concerned about medical errors since before the famous Institute of Medicine report, To Err Is Human. Concerns about medical errors have led to many efforts to improve the precision of estimates of these errors, with the concept of the Global Trigger Tool being the latest such effort. In the article by Kennerly et al. (2014), published in this issue of Health Services Research, an intentional search for medical errors in a large private hospital system with the Global Trigger Tool identified concerns at much higher rates than voluntary reporting (Kennerly found 61.4 adverse events per 1,000 patient days, with less than 5 percent of these events previously identified through voluntary adverse event reporting). Identification of adverse events through administrative claims was even lower.

Other applications of the Global Trigger Tool methodology have noted similar results. In a smaller sample of patients from three hospitals with existing, intensive patient safety programs, Classen et al. (2011) noted a similar discrepancy between voluntary reporting, the Agency for Healthcare Research and Quality (AHRQ) Patient Safety Indicators, and the Global Trigger Tool. Adverse events detected with the Global Trigger Tool occurred in 33.2 percent of hospitalizations, with again less than 5 percent detected through voluntary reporting. The AHRQ Patient Safety Indicators detected only 9 percent of adverse events. Similar discrepancies in ascertainment are observed among Medicare beneficiaries (U.S. Department of Health and Human Services 2010).

These findings beg the question: should all health care systems adopt the Global Trigger Tool method in an intentional search for adverse events? Or should health systems continue to rely on voluntary reporting or claims-based indicators, which appear to substantively underperform in comparison to intentional search?

One concern regarding the trigger tool approach is the cost. Application of a trigger tool through nursing review of records is labor intensive and may be costly in comparison with voluntary reporting. The costs of a trigger tool – based system are estimated at approximately $2 per admission (Kennerly et al. 2013). This cost contrasts with the substantial incremental costs of care due to adverse events (U.S. Department of Health and Human Services 2010). However, as we discuss later, the economic benefit of a method like the Global Trigger Tool is directly related to the ability to use these new data to effect meaningful change within a health care organization.

A second concern is that the high rates of adverse events detected by the trigger tool methodology, and identification of adverse events that may not result in lasting patient harm, can empower skeptics of safety and quality efforts. Kennerly et al. (2014) noted that while health system executives responded with alacrity to the high rates of adverse events, clinicians were more likely to meet the findings with skepticism. In particular, Kennerly et al. (2014) note that clinicians often viewed adverse events as not preventable, and thus not representing a departure from the standard of care. In this context, it is important to note that historically, blood stream infections were simply considered an accompanying risk of central line catheters, until intentional efforts at preventing these “not preventable” adverse events resulted in dramatic decreases in central line infection rates (Pronovost et al. 2006). Health system leaders could leverage the findings of a trigger-based review system to shift perceptions around adverse events and foster a more patient-centered safety culture.

A related concern is that application of the trigger tool methodology may focus too much attention on ascertainment and too little on improvement. To the extent that individuals within the system focus on adjudicating preventability, or whether deviation from standard of care occurred, distraction from the primary goal of improving care may result. As noted by Kennerly et al., strong leadership is required to shift prevailing cultural norms to view adverse events as serious departures from expected care.

Thus, application of the Global Trigger Tool within a health care system has the potential to represent an important first step along the road to better and safer care delivery. If we accept that intentional search for adverse events using a trigger-based methodology can provide the most robust measure of clinical performance, what is required for improving care at scale?

Despite the increasing attention from the medical and health policy communities on reducing adverse events, little evidence of large-scale change exists. To the contrary, Landrigan et al. (2010) found no significant change in the overall rate of harms over a 6-year period, despite the increasing awareness of patient safety issues. The persistence of adverse events in care is not limited to the U.S. health care system. A recent analysis of hospitals in the Netherlands found that the rate of preventable adverse events remained unchanged over a 5-year period (Baines et al. 2013).

From a management perspective, we clearly have a service with a deeply flawed quality record, with up to one in three discharged patients experiencing an adverse event, based on the results of Kennerly et al. Yet, despite the evidence that this problem exists, it has not been resolved. Policy makers have responded by realigning incentive structures for hospitals to put them at financial risk for quality. Previously hospitals faced minimal financial penalties for complications arising from care; instead, these entities often received increased payments to treat patients who experienced potentially preventable harms. Federal and state policy initiatives now require monitoring of specific adverse events (“never events”), and payers refuse to provide additional payment when these events occur.

Changes in the incentive structure can lead to better alignment of hospital and patient safety interests, but a regulatory approach to this problem is not a panacea. Readmission penalties fall disproportionately on hospitals with a high proportion of patients with low socioeconomic status. Some observers note that variation in coding practices (Welch et al. 2011) may reflect attempts to optimize risk adjustment rather than reduce mortality by improving care delivery (Woolhandler, Ariely, and Himmelstein 2012).

To date, data transparency and regulatory mandates have not solved the quality problems in health care. Given this result, we need to ask the question of how to internalize the ownership of clinical quality as an essential responsibility of management at an organizational level in health care (Glickman and Schulman 2013). In this framework, structure is the capacity of an organization to support high-quality processes (Glickman et al. 2007). Therefore, health system leaders should lead the drive toward improving quality of care, rather than simply focusing on quality assurance and reporting. What are the generalizable lessons from successful quality-improvement campaigns?

Robust measurement, along with a feedback mechanism to clinicians, is the essential foundation to any serious effort to improve patient safety and quality of care. How often do physicians receive data about the quality and safety of care they provide for their patients? Data collection systems also need to support actionable feedback—not only signaling that, for example, catheter-associated urinary tract infection rates are high, but also infrastructure to facilitate data collection in support of efforts to identify drivers of the problem and test solutions. Health system management can send an unequivocal message about institutional priorities by establishing data collection and feedback mechanisms.

Visible leadership at all levels of the organization is also a common feature of quality-improvement frameworks. In the LEAN system (also known as the Toyota Production System), a key concept is that management leads by going to the “gemba,” or the place where the value is created. Rather than developing a single management function assessing quality (e.g., the chief safety officer), this model devolves responsibility for quality across the entire leadership team. Successful examples of care transformation, such as Virginia Mason Hospital (Kenney 2011), Geisenger Health System, and Intermountain Health, highlight the key role of determined leadership from the very top through the roots of the organization.

Management must also adopt a framework for improving quality and disseminate this framework throughout the local system. Several frameworks exist, such as the Model for Improvement (from the Institute for Healthcare Improvement), the LEAN system (Toyota Production System), and Six Sigma. Fundamentally, each of these systems emphasizes a systematic approach to quality improvement based on measurement of phenomena, intentional experimentation to assess the benefits of interventions, and based on those results either continued implementation and/or further iteration. Education and dissemination are important: when management espouses an improvement framework but front-line physicians, nurses, and other care providers do not understand or use the framework, then improvements in patient safety or quality of care are unlikely to occur.

Evidence-based clinical performance tools can provide an important starting point for organizations. For example, Pronovost et al. (2006) have demonstrated the utility of simple checklists for reducing central line-associated blood stream infections and have deployed this checklist at scale. While much of quality improvement is based on local personnel and processes, the adaptation of evidence-based tools to local circumstances provides a clear opportunity for rapid improvement in care delivery and patient safety.

Finally, the importance of institutional culture in assuring patient safety cannot be underestimated. As noted by the Institute for Healthcare Improvement, in a culture of safety everyone in the organization takes action in the face of safety concerns (http://www.ihi.org/resources/Pages/Changes/DevelopaCultureofSafety.aspx). This is a key concept in the Toyota Production Model, where anyone on the production line can stop the line to address quality concerns (Kenney 2011).

This is a critical time in the transformation of medicine as a network of service organizations. More robust measurement tools such as the Global Trigger Tool help us to understand the gaps in our current clinical care model. However, the precision of measurement of adverse events often stands in stark contrast to our ability to engage health care delivery systems to commit to the more difficult challenge of organizational transformation. It is critical that the policy and regulatory communities expand their framing of the problem to help catalyze continued improvement in clinical performance in health care delivery.

Supporting Information

Additional supporting information may be found in the online version of this article:

Appendix SA1: Author Matrix.