The incongruity between the Hippocratic principle of doing no harm and the lack of physician engagement in patient safety in American hospitals has not been adequately acknowledged or addressed. In large measure, the problem is rooted in the prevailing model for clinical peer review that provokes fear and fails to substantially improve quality and safety. Even though a model is available that corrects this dysfunction, the problem of efficient case identification remains unsolved. Physicians are almost always aware of adverse events affecting their patients, but are reluctant to report. The aviation industry successfully advanced passenger safety by engaging pilots in self-reporting through an offer of immunity from sanctions. The Patient Safety Act offers comparable protections for self-reporting of adverse events and hazardous conditions that have not been appreciated. By promoting self-reporting of adverse events and adopting better clinical peer review process, with the support of a Patient Safety Organization relationship, hospitals could readily engage physicians and accelerate the organizational learning required to improve patient safety.
Adverse events and hazardous conditions remain all too common in American hospitals. Adverse events cause patient harm manifest as prolonged hospitalization, permanent injury, life-sustaining intervention or death. About 7% of hospitalized Medicare patients suffer from preventable adverse events. Another 7% experience preventable temporary harm that requires treatment, but does not prolong the length of stay (e.g., delirium, hypoglycemia, hypervolemia). (1)
While this issue has gained national attention, some obvious contributory problems have not been addressed. In particular, Wachter observed, “Although physicians’ involvement is crucial for patient safety, many—perhaps most—practicing physicians remain unengaged.” (2) There is little data on the actual extent of physician engagement in safety. In a survey of members of the American College of Physician Executives, only 34% reported that physicians at their respective organizations were very supportive of quality and patient safety improvement projects. (3) Improvements in patient safety will be difficult to achieve where physicians fail to assist with the identification of problems, where they resist the adoption of proven practices such as surgical and ICU check lists, where they make only meager efforts at clinical peer review, where they don’t contribute to process improvement work, and where they fail to provide leadership. In practice, physicians uncommonly report adverse events (4). Participation in peer review has been described as fair or poor among 24% of hospitals (5). Regardless of the overall level of physician engagement in quality and safety, widespread interest to improve engagement has been cited by others (6)
How is it that the medical profession has not fulfilled its fundamental accountability, embodied in the ancient Hippocratic principle: Above all, do no harm? Physicians and health care leaders should be looking closely at this incongruity. This brief essay will explore the potential for engaging physicians in patient safety by improving clinical peer review process and by invoking the protections of the Patient Safety Act to make it safe for physicians to self-report adverse events and hazardous conditions.
Contemporary quality improvement methods emphasize the study of “defects” to generate the organizational learning that leads to process redesign. Every adverse event and near miss could potentially contribute to learning for the involved individuals, the organization and, occasionally, the profession. While medical error involving judgment, skill or patient management is common, most preventable adverse events are multi-factorial and may involve hazardous conditions latent in the system of care. Even where diagnostic errors play a role, preventive strategies need to be buttressed with system solutions. (7, 8) Blaming individuals is easy, but counter-productive. Thus, when analyzing adverse events, the most important work is to identify and fix process issues in the system of care.
If physicians are not generally engaged in patient safety, by inference the situation must contribute to a sizable gap in organizational learning. The primary forum in which physicians have traditionally analyzed adverse events is the clinical peer review process. Virtually all hospitals have a peer review program in order to meet Medicare conditions of participation and minimize vicarious malpractice liability. Retrospective medical record review is universally practiced. In about 60% of facilities, the peer review program also includes morbidity and mortality conferences. Peer review involves single specialty or multi-specialty committees of physicians in decision-making. (5) It typically operates in parallel with and complements inter-disciplinary organizational processes for adverse event investigation and analysis, but at lower resource intensity and much higher case volume.
Two national studies have revealed wide variation in clinical peer review practices, along with a high level of dysfunction and substantial opportunity for improvement. (5, 9, 10) In many hospitals, case review looks narrowly at whether the involved physicians met the standard of care, confuses clinical performance evaluation with the determination of competence, and fails to address clinical process issues. Adverse organizational culture, including a punitive style of peer review, is also associated with lower hospital quality and safety. A more effective quality improvement-oriented “QI model” has emerged, which asks broadly, “What can we learn from this case to improve clinical performance.” (11) Most programs could readily adopt elements of the QI model, such as recognition of excellent clinical performance, identification of clinical process improvement opportunities, timeliness of review, standardization of peer review process, attentive program governance, and integration of peer review with other hospital improvement activity.
Even so, the problem with case identification must be solved. Hospital reporting systems may identify fewer than 10% of adverse events. (12) Virtually all hospitals primarily rely on so-called “generic screens” for substandard care that were pioneered in the California Medical Insurance Feasibility Study during the 1970s. These screens are fundamentally flawed and inefficient for use in peer review. (13, 14) Much effort (15-18) has failed to find a more efficient strategy for identifying adverse events, although a few teaching hospitals have had encouraging results with resident reporting (19-21), and rapid response teams may be useful for both identifying and analyzing an important subset of life-threatening events. (22-24)
The natural and obvious solution, self-reporting, has been neglected. Physicians generally know when a patient suffers an adverse event, whether or not they fully understand their own contribution. The unspoken assumption that physicians will not voluntarily report adverse events signifies an environment in which lack of trust prevails. In most hospitals, it is not safe for a physician to call attention to an adverse event, much less to admit an error to peers.
The lack of openness about adverse events not only hampers efforts to improve safety, it causes untold collateral damage. Patients who have been harmed are denied the satisfaction of knowing that the organization will be able to prevent a recurrence. Clinicians who discover they have made an error frequently bear their guilt and anguish alone, compromised in their ability to fully explore causative factors and protect colleagues from similar experiences. (25)
In contrast, the aviation industry successfully engaged pilots in safety by promoting self-reporting of errors and dangerous conditions with impressive results. In conjunction with Crew Resource Management training, self-reporting was also instrumental in shifting from a culture of “Captain as King” to one of teamwork, making crisis management more effective. The critical enabling factor was the end-user benefit of relative immunity from sanctions. (26)
One anesthesiology peer review program has sustained high rates of self-reporting (90% of cases reviewed, 70% of events identifiable by all means) and excellent clinical outcomes. (27) The authors conclude that "Anesthesiologists will comply with a system of self-reporting if they understand the process, if there is institutional and departmental encouragement and support for the process, and if the process is non-punitive and can result in real improvements in patient care."
Appreciating the importance of these caveats which link directly to the QI model for peer review, can this experience be generalized? If so, how?
It is not possible to move from an atmosphere of fear, avoidance and silence to one of trust absent a major new influence. Leadership and good communications are essential, but may be insufficient. The broad Federal protections of the Patient Safety Act can provide the needed catalyst. When invoked, they give strong support for self-reporting by simultaneously protecting the reporter from reprisals and the information from discovery.
The Patient Safety Act created a general framework to protect voluntary initiatives for improving quality and safety in all health care settings through reporting to Patient Safety Organizations (PSO) with oversight by the Agency for Healthcare Research and Quality (AHRQ). (28) The Act’s strong confidentiality and privilege protections were intended to facilitate the standardized, large scale data aggregation and sharing that would be capable of advancing a culture of safety in health care, along with more rapid learning and improvement. The role of the PSO is to work with multiple health care organizations and their associated clinicians to identify, analyze, and reduce the risks and hazards associated with patient care. Thus, the PSO brings additional expertise and rich data to complement and extend the work that hospitals are already doing to minimize adverse events.
The Patient Safety Final Rule was implemented January 19, 2009. Its effects are just beginning to be seen as infrastructure work continues. AHRQ released the first version of Common Formats for acute care event reporting in 2010 and is not expected to begin receiving electronic reporting from PSOs to its Network of Patient Safety Databases until late 2011.
About 80 PSOs are currently listed by AHRQ. (29) A provider may enter into more than one PSO relationship. Although most PSO relationships have initially focused on sharing adverse event details, the Act allows any healthcare provider organization to include clinical peer review information in its Patient Safety Evaluation System (PSES). PSES is a regulatory term used to describe the provider’s process and protected space for collecting, managing and analyzing information about patient safety events for the purpose of reporting to a PSO. Since hospitals already have such processes in place, it is not difficult to modify them in accord with the intended PSO relationship. The regulations do not in any way restrict the use of peer review, root cause analysis and other methods of event analysis outside the PSES. They allow flexibility in the structure of the hospital-PSO relationship. For example, the PSO can provide additional expertise to facilitate event analysis or the hospital can maintain its own process and simply report the results. There is no significant downside to PSO reporting. The leadership just needs to become familiar with the subject. At least one readily accessible physician’s guide to PSO reporting offers additional information. (30)
Thus, it is relatively simple to establish a PSO relationship to promote and protect self-reporting of adverse events and hazardous conditions. The essential operational detail to be worked out with the PSO is the mechanism for reporting. Although electronic methods best support data aggregation, they are not required under the regulations. The hospital also needs to modify the peer review policy with language such as: "No medical staff member shall be subject to disciplinary action in relation to cases that are self-referred for peer review, in the absence of reckless disregard for patient safety." While this policy alone might be useful to promote self-reporting, the umbrella of Federal protections is indispensible in states with poor statutory protections for peer review. Furthermore, the commitment to PSO reporting reinforces the intent of the program, testifies to the integrity of leadership, and provides additional collateral to build trust among skeptics. Moreover, under the Affordable Care Act, beginning in 2015, hospitals with more than 50 beds must have a PSES to contract with a qualified health plan. The potential benefits of self-reporting offer a good reason not to procrastinate.
Self-reporting connects the individual to the group. In effect, each report acknowledges individual limitations, reveals vulnerability to system defects, invites exploration of root causes and preventive strategies, provides an opportunity to give solace to the “second victim”, and contributes to organizational learning. Initially, self-reporting would enhance rather than supplant other methods of adverse event identification. For self-referral to payoff, it needs to be complemented by peer reviewer training in basic quality improvement principles and methods. Graber has clearly demonstrated that this is easy to do and, of itself, greatly improves reviewer participation and program effectiveness. (31) There must also be a simple mechanism to hand off issues that might exceed a peer review committee’s scope of authority or resources. The growth of a learning environment promotes a culture of safety, in which hazardous conditions may be more readily recognized and corrected before an event occurs.
The ability to learn from experience is essential for organizational progress. The tools for removing a major block to learning in health care are now at hand. They could quickly bring physicians back to the forefront of the patient safety movement. They may be of equal value for nursing, particularly where there is commitment to the American Nursing Association’s Magnet Program, which mandates a nursing peer review process. Hospital and physician leaders should consider promoting self-reporting of adverse events and adopt QI model clinical peer review process. This is not the only way to engage physicians in safety, but it builds on familiar processes, corrects significant problems, requires minimal investment, and is simple to implement. Longer term, it may pave the way for the development of new methods of inter-disciplinary teamwork to analyze and reduce adverse events. Peer review must be done. Why not seek to harness its full potential?