In the 1970’s, the rate of commercial airline crashes had plateaued at an unacceptably high level of about one every few months. Investigations determined the safety attitude of pilots was often an important culprit. In response, several airlines introduced the tenets of a high reliability organization (HRO), including a new form of teamwork training that exploits team peer pressure to eliminate all preventable errors. Almost immediately, the rate of crashes approached a rate of zero and has remained so for the last 20 years. This outcome reflects a culture that has spread beyond the cockpit to everyone in the airline industry to achieve “zero errors”. Such a vision has proven to be highly motivating. Today, flight crews, pilots and employees of other HRO (military, nuclear power plants) don’t just agree with this training, they have become evangelists.
In the 1980’s, Japanese automakers redesigned their own system of manufacturing to drive down costs and US companies soon followed suit. After having such a clear impact in the auto, airline and other high risk industries, the mandate for system redesign in US hospitals came to national attention in the late 1990’s. It was generally agreed that bad system design in hospitals has caused problems with cost, quality and unnecessary patient harm. Hospitals have yet to respond. Mark Chassin, MD President and CEO of Joint Commission, describes it best in his own words: “We know of no hospitals that have achieved high reliability across all their activities”. Clearly, hospitals have not bought into the vision of “zero errors”. Many leaders feel healthcare is too complex for this to be a realistic goal and that setting “zero errors” as a goal would just set the organization up for failure.
The attitude that all hospitals are not accountable to the goals of an HRO is the same that pilots had in the 1970’s. Allowing this view to persist is a leadership crisis, plain and simple. Leading by creating an ambitious vision is a core competency of CEOs of the highest performing companies. Hospitals are no exception. A visionary leader sees the real failure as setting goals that aren’t ambitious enough because that amounts to failure before you even start. When it comes to patient safety and avoiding preventable deaths, no one should be content with only modest improvements. The first step towards zero patient harm is to set it as the only acceptable goal. It is the only way to prevent complacency.
Hospitals have a moral obligation to redesign their dysfunctional systems. The basic idea of redesigning them according to the principles of HRO is straightforward – it means designing all processes used by staff and administration that directly or indirectly touch patients based on the current safest and best practices known at the time. Those rules are then regularly updated through a frequent cycle of review. Hospitals that I’ve worked at rarely engage in this type of formal, proactive rulemaking and just default to a patchwork of rules based on tacit knowledge of “the way we do things here”. This is interrupted occasionally and haphazardly by ad hoc responses to the most recent safety crisis. Safety judgements in hospitals have too often been influenced by politics, economics, bureaucracy, legal issues and incompetence rather than a passion about saving patient lives.
To become an HRO, a hospital must first reconcile widely different perspectives held by its administrators and clinicians. An administrator’s job is to allocate resources to projects that provide a good return on investment. Investments to prevent patient harm often appear unattractive. It is difficult to get an accurate accounting of events that were prevented or didn’t happen. Uncovering the value of these investments requires a deeper understanding of the clinical impact. One hospital in North Carolina recognized that their board members were often making these types of decisions without the required clinical knowledge and created an “Immersion Day”. In this program, the leaders spent a day in scrubs, behind the scenes, immersed in the nuances of care delivery. It was highly popular and led to major shifts in the funding decisions of those that participated. One participant said he learned more about hospitals and health care from 10 immersion hours than after many years sitting on the board. In other words, senior hospital leaders given a 10 hr snapshot of medicine started making decisions that their physicians probably begged them to make for many years. Perhaps these boards should have spent their 10 hrs learning to become better listeners.
An even more remarkable thought…most hospitals make high level decisions that influence patient safety without the benefit of even a single day of medical exposure. Decisions that impact safety are made by those highest in the chain of command, usually without consulting physicians or other professionals with the greatest relevant expertise. If a physician made decisions on behalf of their patients this far outside their scope of practice, they could expect to be censured by their state medical board and/or peer review committee and at risk for malpractice. This standard for how hospital executives make decisions puts a double curse on any hope of becoming an HRO. Those in charge not only lack the ability to produce the necessary decisions to drive an HRO culture forward but also the expertise required even to know they aren’t producing them.
The intimate collaboration between physicians and administrators required for an HRO culture does not exist. For the last 100 years, hospital departments have operated in distinct silos with little collaboration. Silos are a destructive problem for any businesses. Leadership breaks this down by creating a common vision for everyone in the organization. Hospitals have not articulated such a vision. Instead, they often prioritize business goals such as employing physicians to shore up their referral base. Many have used strong-arm tactics that pits doctors as competitors if they don’t acquiesce (e.g. stop patient referrals, cancel medical directorships, take away preferred times for surgical procedures, open up urgent care facilities right next door to their practices). When frustrated by physicians that don’t respond to their authority, more than a few hospital leaders have chosen to do things blatantly unfair just to demonstrate their power and control. Such unfairness in the workplace is common and hinders employees’ commitment to their organization and has festered into a growing contempt and fundamental lack of trust. This has been shortsighted because talented staff become engaged when they are inspired by a vision, not coerced to perform using money or threats. True collaboration means sharing the authority to make decisions and that requires surrendering some power. No one shares a scintilla of power in an environment of such low trust.
On the other end of the spectrum are the clinicians on the front lines. The mere act of taking care of just one victim of a preventable error inculcates a radically different perspective on this topic. Many staff involved in a bad patient outcome can quite accurately articulate the underlying system problems that contributed and know in their heart of hearts that these problems are unlikely to ever be addressed. They carry their guilt about their role in causing harm at the same time they worry about their other patients being effected by uncorrected problems in the system. Clinicians deal on a daily basis with the harsh reality of personal accountability for bad events – the persistent risk of lawsuits, a “blame and shame” culture of peer review and infamous Morbidity and Mortality conferences. Being deemed “fit” for practice entails accepting this reality. However, the emotional intensity of harm can also create tunnel vision in physicians that crowds out the ability to think big picture. It’s the same form of cognitive bias (called “hyperbolic discounting”) seen in the alcoholic who makes the decision to binge drink after discounting the severity of regret he will have on the following morning. Physicians get drunk on the immediate patient feedback about whether they did a good job while administrators must remain sober as they wait over longer timeframes to see the results of their actions. Appreciating the different needs of a specific patient vs. a whole organization is a skill critical to running a hospital as it is any business. Through a distorted sense of professional obligation to patients, some physicians shun such a skill and take on an irrational “us vs. them” favoritism against those who consider the big picture.
Neither side is correct, but both realize that a chronic stalemate is unsafe. Anyone trapped in an unsafe state either quits or acquiesces to a state of complacency that accepts bad events as the “price of doing business”. Apathy and denial creates a state of cognitive dissonance in those that care passionately about the safety of their patients. This is a mental state much like that of a nun in the Catholic Church as the scandal emerged of Priests guilty of child abuse. Only one idea could have kept a nun going to work at a place of ongoing, unspeakable horror: “If no one else is speaking up about this, maybe I’m the one interpreting things wrong.” The power of groupthink (i.e. go along to get along) is hard to overstate. It too easily resolves cognitive dissonance by striking down prior strong convictions about safety and replacing it with a new lower level of expectations known as normalized deviance.
A variety of political and legal disincentives prevent administrators from listening to the staff and physicians involved in an adverse event. First of all, knowing the details can force administrators to understand how they might have helped prevent the error. This takes away plausible deniability. Administrators might accept the theory that things they control might contribute to patient harm but rarely admit that any of their decisions/actions contributed to harm in a specific patient. Doing so might win respect among the front lines but would also create legal risk. A variety of state laws, medical boards, and ethical guidelines are specifically directed at concerns over the corporate practice of medicine and aim to protect the independence of physicians. Acknowledging that there are decisions by administrators that influence the care of patients might run afoul of a strict interpretation of those rulings. Moreover, if the patient that was harmed decides to sue, the legal theory of vicarious liability establishes that hospital administrators are responsible for actions normally considered to be physician malpractice depending on their degree of control they had over the event. Hospitals tend to be viewed less sympathetically by juries than a physician. This tendency yields greater size awards when a medical error is identified. Defense lawyers hired by the hospital prefer to keep the blame for malpractice isolated on physicians.
I’ve noticed that right after a really big error on a patient occurs, many of the politically savvy hospital leaders make a conscious effort to steer clear of the weeds that bog down the front lines. Their more pressing goal at this point is to develop and present a corrective action plan to the board of directors, who also have disincentives to hearing the full story. In this setting, the damage control plan that is the most psychologically satisfying is to pin the event on individual clinician(s) and “remove the bad apple”.
The team assigned the “watch dog” role and the task of uncovering the real issues that cause safety problems is the hospital’s QI/QA department. An important system design problem is that this department is directly supervised by the hospital administration. Not surprisingly, their investigations rarely point their finger at system problems under the control of the administration. A similar organizational reporting chart influenced one of the greatest management failures in US government history, the NASA Challenger crash. A Presidential Commission concluded that QA teams placed under the supervision of those they were supposed to check allowed NASA to bypass safety requirements and this was the root cause of the Challenger disaster. Hospital investigations in this setting often create false diagnoses and corrective action plans that are doomed from the start.
When a death occurs in a hospital that was preventable, it doesn’t have to be stated out loud. You see it in the faces of those involved. Deep down, everyone – nuns, nurses, technicians, doctors and administrators – knows that when a problem exists, it exists whether it is acknowledged or not. The lack of transparency alienates staff and cause them to lose trust in their managers. Even more problematic is when errors leading to harm get the attention of lawyers. Once this happens, the staff that were involved are often required not even to acknowledge the underlying issues and just march on as if nothing happened. More skilled workers in this situation eventually leave to find work elsewhere. Those that remain are discouraged from getting involved in future problem solving and are at risk for becoming a “second victim” that further limits their effectiveness.
The last best hope for bringing hospitals back into focus is to become an HRO. The first step is to acknowledge those with the most to lose – the existing hospital leadership team. To paraphrase Albert Einstein, asking the existing team to lead us towards an HRO amounts to an attempt to solve our problems with the same thinking we used when we created them. Their track record is undeniable – no sense of urgency on the topic of patient safety and no credibility with the front lines as someone likely to get such a job done. Hospital leaders are currently acting like other leaders faced with growing scandals – think Catholic Church in the 1990’s and financial institutions of the early 2000’s. These organizations usually got in their bad spot by pretending not to know something. Hospitals today are pretending that their poorly designed systems have not been causing safety problems for decades and their slow decisions are not the bottlenecks that prevent redesigning those systems. They behave this way in large part because the existing system has protected them and a new culture is an existential threat.
George Orwell understood, “No one ever seizes power with the intention of relinquishing it”. The leadership team capable of creating an HRO must be prepared to quickly remove the old guard. The most effective tool for doing this is high velocity decision making. The new motto must be: “a wrong decision aimed at patient safety is better than no decision at all”. Bad decisions that don’t really improve things or cost too much can be undone. Outside healthcare, the importance of rapid decisions by CEOs is well understood. Jeff Bezos wrote in his annual letter to the board of Amazon – “Most decisions should probably be made with somewhere around 70% of the information you wish you had. If you wait for 90%, in most cases, you’re probably being slow. If you’re good at course correcting, being wrong may be less costly than you think, whereas being slow is going to be expensive for sure.” Harvard Business Review describes a decisive CEO as one that recognizes not to wait for “perfect information” and makes the call when there is around 65% certainty. I can’t imagine any other leadership style capable of yielding an HRO, which in turn is the best way to make the old guard irrelevant.
Another hallmark of the successful HRO leader is a refusal to accept the same, tired, simplistic answers to correcting our current sad state. One of those simple solutions that hasn’t work is adding physicians into the ranks of hospital executives. Conventional wisdom states that this will improve “physician engagement”, break down departmental silos and close the gap between administration and physicians. When those with medical knowledge communicate important hospital decisions to other physicians and staff, it will be more readily accepted since they have more confidence in a physician as their manager. A prototypical example is Toby Cosgrove, the world renowned cardiac surgeon who is on everyone’s list as best hospital CEO because of how he changed their culture. Dr. Cosgrove sums up his influence over physicians at Cleveland Clinic with one word: credibility. Those at Cleveland Clinic seem to buy into the idea that a highly credible physician is a quick fix to the complex problem of engagement. Every one of the candidates proposed to replacing the retiring Toby Cosgrove are well recognized physicians.
Unique leaders like Cosgrove certainly have a better shot than anyone at bridging the hospital-physician divide. However, picking a doctor at your hospital and hoping to reproduce the results of Toby Cosgrove is like predicting that the all-star quarterback at your high school will end up as good as Tom Brady. Most physician-executives have been more of an Uncle Tom than a Tom Brady. They serve as powerful symbols of the problem but haven’t provided a solution any more helpful than Uncle Tom was to slavery. The ability to speak medical jargon is not what it takes to transform a hospital into an HRO. It is true most physicians are less compelled by business logic than on specific unmet needs of their patients. However, physicians respond to scientifically sound arguments, whether medical or economic. Any one in active practice must have a working knowledge of the unique jargon of nurses, therapists, technicians, nutritionists, social workers. Likewise, no doctor is surprised by administrators that communicate using a business perspective.
Adding further to the problem…physicians with the talent of Cosgrove are not the ones clamoring for a shot at leadership. Remember, this is a thankless job. Many of those that you would be asked to lead – e.g. arrogant and increasingly angry physicians – are fiercely independent and don’t want to be managed by anyone. Nonmedical CEOs are granted automatic allegiance based on their position of authority. In stark contrast, physicians are morally justified to disobey you if doing so is not beneficial to their patients. You will be asked to control costs that no one is able to quantify beyond a vague recognition that they are spiraling out of control at the same time reimbursements are plummeting. And your salary will be 25% less than the average CEO of nonmedical firms and your job security worse. While most organizations benefit by the cream rising to the top, those floating to the top of a troubled industry like hospitals are not the cream but often the dreck, a phenomenon described in the management literature as “cesspool syndrome”. Indeed, the poorest performers are often the least aware of their own incompetence, a form of cognitive bias (i.e. Dunning-Kruger effect) that is described on public radio as the Lake Wobegone Effect, named after the town where “all the children are above average.”
Another area where simplistic solutions are applied is the high rate of physician burn-out. Dr. Noseworthy, the CEO of Mayo Clinic, feels that burnout is from frustrations over the cumbersome electronic health records. This is true only to the extent that investing so heavily in such a horrible system is symptomatic of a decision making process that repeatedly fails by not incorporating the clinicians’ perspective. A much more fundamental concern to physicians is that this process harms patients on a daily basis and yet no one is willing to say the Emperor has no clothes. The famous psychoanalyst Carl Jung said problems that we do not make conscious emerge later as fate. The fate of a system prohibited from making its most serious problem conscious is unnecessary patient harm. The fate of individuals trapped in this dysfunctional system is burn out. Burn-out comes from the EHR and multiple other dysfunctional interactions endured every day that all promote duplication of work effort, quick fixes, and miscommunication.
Roadblocks like these reinforce a culture where the HRO shift cannot occur because the leaders don’t know or care enough to push such a major change forward. The only solution is a change in the culture. This will not happen by merely bringing physicians into the world of executives. The Immersion Day experiment illustrated that the conversion to an HRO occurs at the bedside, not the boardroom. What is needed is to bring executives into the world of physicians, a shift in approach described as creating a physician-administrator dyad. A dyad relationship is unfamiliar to those from a traditional command and control hierarchy. It requires servant leadership so that both points of view are respected and contribute to all major decisions that directly or indirectly affect patients. At the bedside, executives can learn from physicians who naturally play the role of servant leaders of their patients. The legitimacy of the authority of the servant hospital administrator rests solely on their use of administrative skills to help solve problems for patients.
Scientifically grounded methods can make this cooperative dyad flourish: 1) develop a shared purpose, 2) create an open, safe environment to speak up about harm, and 3) breakdown silos by insisting on fairness and equity in the application of rules across departments. The roadmap has been established in HRO in other fields: the process starts with a common vision of zero error. If this is to be the shared purpose for an HRO hospital, the first and most important step is an honest investigation of all bad patient outcomes. These need to be soul searching investigations. All stakeholders should demand to know what really happened and whether their poorly designed systems had any impact. Administrators will be surprised how these debriefings will engage hospital staff, particularly for those that have felt unable to provide their vulnerable patients with the safest possible care. The servant administrative leader will naturally ask: “what can I do to advance this engagement?”.
Debriefings work when the environment is safe for being honest about the actual underlying problems. Hospitals are haunted by stories of staff that have been blamed and shamed as the result of simple human error. We have all witnessed a clinician who lost a job, raise, promotion, or a seat at the table because they naively spoke their hearts about a problem in a politically sensitive situation. One of the most effective ways to overcome this fear of politics is for leaders to admit their own errors that contributed to a bad event. Humility creates psychological safety in others to admit their own faults. When hospital administrators join this debrief as a dyad partner, they start to see how their own bad decisions (usually errors of omission, e.g. not allocating resources for a critical piece of equipment, appropriate staffing or timely training) directly influence patient safety. As their learning progresses, they will emerge from the first stage of competency (unconscious incompetence), begin to say “I don’t know the answer” and discover clinicians are happy to help.
The final stage in an HRO is to insist on fairness and equity in the application of rules. It is fair to accept that not all physicians will be sold on HRO concepts. The tree of safety yields invisible fruit – a few less deaths or complications over a distant future. Physicians are under pressure to increase their work productivity and may not want to engage with a program that seems ineffective according to an RVU standard. Many already protect their patients via intricate workarounds developed over years of practice and won’t be quick to abandon the sunk costs they’ve invested in them. Resolving the tensions between productivity and safety is a common challenge of any safety program. Moreover, the open book management style of an HRO isn’t for everybody. There are old school surgeons who were trained to be the captain of the ship and feel most comfortable with this. With any major change, one should expect the so-called “late laggards” that will be convinced only after they see clear evidence of benefit.
If fairness and equity are necessary for an HRO to succeed, then hospital leaders and administrators must be its champion and lead by example. This starts by the administration inviting the same harsh accountability that has been applied to physicians for years. Administrators are prone to errors (of omission or commission) that can have major financial or clinical consequence. Investigations into these errors should be performed by a QI/QA team that includes physicians and reports directly to the board of directors. Their task would be to determine if there was appropriate expertise to make the decision found to be errant and if not whether input was obtained from those with relevant expertise. If the answer is no, a corrective action plan should be in place to make sure similar flawed decisions don’t happen again. Those that refuse to participate in this process should be made immediately available to a non-HRO hospital.
It is also fair to expect that an HRO would only succeed in the long run if it helps achieve goals that are more traditional of administration. An administrative dyad would employ the Hawthorn principle (observed behavior gets better) and use scientifically valid information to help their physician partner overcome reluctance to adopt new ways of doing things. For instance, β-Blocker, statin and aspirin after MI became part of the established guidelines for ACC and the AHA in the mid 90’s. These meds were not widely adopted until administrators showed physicians their poor compliance data using the Health Plan Employer Data and Information Set (HEDIS). Similarly, administration could also reduce the reluctance of surgeons to adopt basic safety measures such as the OR checklist and make them less dependent on the sales reps for advice. They could decipher when physicians are “crying wolf” by playing the safety card. The possibilities for improvement are endless after once the “us vs. them” silo separating these two groups of professionals is removed.
The most important goal of the administration is the bottom line and the business case for becoming an HRO is actually very solid. Victims of preventable errors/deaths require substantially more resources during their hospital stay than patients without complications, which reduces the margins for these cases substantially. With only 20 fewer adverse events per year, the hospital will save over $1 million in direct costs. Nurses that no longer have to develop complicated workarounds to keep their patients safe will undoubtedly have improved job satisfaction and less job turnover. Fewer adverse events and greater attention to safety will improve patient satisfaction. Patients that experience a complication are less likely to consider lawsuits or to become disgruntled and spread unfavorable views about this hospital. That can lead to increased reimbursement through the Medicare value based purchasing program and generate more revenue through word of mouth referrals to the hospital. Physicians are likely to be more satisfied in an environment with improved clinical outcomes and higher staff morale. An engaged workforce, like any effective team, will provide the extra discretionary work on behalf of their patients and the hospital in order to produce a team win.
In order for hospitals to achieve the breathtaking potential of an HRO, we must be honest about how depressingly far we are from even the very first stage of this transition. I routinely have meetings with CEO and other administrator where they fail to fulfill a request for resources that I know and they should know is clearly best for patients. During these discussions, I can’t bring myself to blame them personally. Usually, I feel more like I am talking to a company rather than a person, or perhaps someone who could no longer distinguish between the two. In this case, the corporation has effectively absorbed its administrators into a moral universe that interprets its fiduciary responsibilities in solely financial rather than clinical terms. Much like tobacco company executives, they have long ago made the decision to turn a blind eye to the potential damage of this attitude. With that in mind, our task is clear. We must be the beacon of light that allows us to leave the foggy shore and begin the journey.