Every organization needs a structure and a toolkit to support improving safety and quality. Since what cannot be seen cannot be fixed, robust identification of adverse events and sources of risk (risk to patients, to staff, and to the reputation of the organization) should be a priority of every hospital. This chapter begins with discussion of structures and tools to identify adverse events and risk-prone conditions. Once identified, the hospital and staff must then determine the priority items and which techniques will be applied to reducing adverse events and risks. Let us define a few terms for this chapter. Adverse events are instances of harm to patients resulting from medical care. Errors may be characterized as resulting from a flawed plan or from failure of a plan to be completed as intended. Not all adverse events result from error, and not all errors result in harm. A near miss is an error or system failure that is either intercepted before reaching the patient or causes no harm if it does reach the patient. Risk reduction efforts may focus on error prevention or on harm prevention. This chapter will not promote one approach over the other, as these principles and tools apply to both.

Where the culture of safety is healthy, it is easy to see that safety is a priority. People working in the area have a focus on safe practices and supporting one another in being safe and in delivering safe care. They may exhibit a “preoccupation with failure” as described by Weick and Sutcliffe, such that there is a general awareness of and attention to risks. Instead of ignoring small, nagging concerns, workers share those concerns with others, and team members rally to help resolve the concerns. When safety is a priority, physicians respond supportively to concerns about risks to patient safety and do not seek to blame when an error happens. When safety is a priority, staff supervisors and system leaders routinely inquire about safety concerns and take the time to listen, seek a deeper understanding of causes, and demonstrate their commitment to safety through action and by communication back to staff on the response to adverse events and concerns. Questions in Table 12-1 can be useful to assess the effect of culture on identifying problems. See Chapter 7 for a more thorough treatment of culture of safety.

Hospitals must be intentional about the identification of patient safety risks. The National Quality Forum identifies and promotes safe practices in health care (see http://qualityforum.org). Among these is the presence of leadership structures and systems to ensure awareness of safety failures in the organization and the performance gaps that need attention. Government and commercial health care purchasers, insurance providers, and hospital accreditation organizations provide strong incentives for hospitals to invest in infrastructure supporting the identification, analysis, and correction of risk-prone areas. Because information about errors and risks comes from many potential sources, hospitals face challenges to bringing this disparate information together in a meaningful way.

One of the chief challenges is that only a small fraction of true errors are detected by most methods, especially those based on human reporting. Computerized data mining methods tend to greatly increase the number of potential errors identified, but the specificity of these reports can be low. For example, a rule to identify nephrotoxicity from medications may look for a rise in the serum creatinine during hospitalization. Most of the cases identified will not result from an adverse drug event, and additional resources will be required to further investigate.

Knowing that the detected errors are but a sample of true errors and risk points, hospital safety leaders must decide whether there is value to applying additional resources to enhance detection. Bringing together the multiple sources of information allows for a broader view, and this effort benefits from having a central safety committee or a safety officer to whom the information funnels.

Some methods are reactive and retrospective, being generated in response to specific events that come to attention. In contrast, systematic methods tend to identify latent or hidden errors or risk points and are more often proactive. Both types of approaches are necessary in the overall strategy to identify risks.

Reactive Methods

Event-Reporting Systems

Event-reporting systems rely on workers bringing events to attention through their reports. In a healthy culture of safety, workers report freely and openly, with few barriers. Workers do not fail to report events that seem to represent small harms or risks of harm, because they are aware that these are important opportunities to learn. Indeed, a robust reporting system will collect a significant number of system failures or errors that were successfully intercepted or that did not result in harm, known as near misses. Near misses are golden opportunities to identify risk-prone conditions or processes and to intervene before harm results. An effective reporting system enhances the engagement of frontline staff in patient safety by providing an identified channel for their observations. To be effective, workers must be aware of the system and the value that leaders place on their reports. Timely acknowledgment and expressions of appreciation reinforce the desired reporting behaviors.

Ease of reporting is key to maintaining a low reporting threshold. Paper reports and verbal reports via telephone recording have advantages of speed, though the information has to be transcribed and aggregated separately. Electronic reporting systems may prompt for more precise and complete information from each report and may produce structured reports from which data are more easily analyzed. Electronic systems may also enable immediate notification of appropriate personnel. For instance, an event reported as causing significant patient harm may generate an automated communication to a risk manager, safety officer, or hospital leader, facilitating the response to the event.

How leaders and managers respond to aggregate data from event-reporting systems will send strong signals to the staff and to physicians. Because event reports are dependent on willingness to report and are unlikely to reflect true incident rates, leaders and managers should exercise caution in inferring that a high number of events reported represents worse safety in one area versus another. In fact, the number of reports may be more indicative of the culture of safety than of safety itself. However, it is human nature to conclude that higher numbers of reports indicates worse safety. Managers may worry that reports reflect poorly on their performance and discourage use of the reporting system. Leadership attention to the use of data for learning rather than for judging is critical to the reporting system effectiveness.

The options for reporter identity protection deserve intentional thought when designing an event reporting system. An open system makes no attempt to protect the reporter’s identity, so colleagues and supervisors can know who reported. This kind of system can work where the culture of safety is strong enough that there is no retribution for reporting and, in fact, reporting is rewarded, whether by peer appreciation or formal recognition. Where the reporter is identified, an event report can be followed by further investigation, and an open and frank discussion of the event promotes learning. If reporters face criticism or retaliation in even a few instances, however, willingness to report can be severely affected.

A confidential reporting system allows identification of the reporter only to responsible system administrators who can follow up on the event with the reporter. Confidential reporting may overcome reluctance of some people to report and enhance detection of some kinds of sensitive issues. For example, a person reporting inappropriate sexual comments or behavior may be reluctant to have his or her identity known to the person whose behavior is being reported. Through attempts to protect the identity of the reporter, however, the investigation may have more limited scope that fails to obtain the whole picture.