Learning from patient safety incident investigations

Dan Cohen, chief medical officer of Datix, throws down the gauntlet for healthcare risk managers: improve the quality of investigations so that real opportunities for learning can be identified and actionable improvement strategies can be developed, put in place and sustained.

Risk management professionals, and in my view everyone involved in providing healthcare, should be considered a risk management professional; all have enormous challenges to overcome. Injuries and deaths related to receiving healthcare services continue to plague quality improvement efforts, and what seems lacking to me is a sense of collective professional urgency about this. There is too much complacency among our ranks, with risk management viewed by many healthcare professionals as the responsibility of a small cadre of administrative staff who somehow “manage” risk, whatever that means.

The latest studies regarding deaths related to healthcare reveal that over 250,000 people die in America every year due to engagement with the healthcare industry, not because of underlying illnesses but rather as the result of system and process insufficiencies and human errors. Many orders of magnitude more are harmed, often seriously harmed, and some are damaged permanently. The numbers are daunting and define a public health crisis of enormous proportions. Even if one wished to debate the validity of this projection and used instead the figure of just under 50,000 deaths noted in the Institute of Medicine report of 1999, this is still a public health crisis; it touches the lives of thousands of mothers, fathers, grandmothers, grandfathers, children, sons, daughters, friends and lovers, even enemies.

"If the healthcare industry is to emulate the characteristics of higher reliability industries then it, and we, must diligently strive to learn from investigations of safety incidents."

If that were not enough, there is a risk that the situation may get even worse if projections of the impact of illness burden and complexity related to the pernicious and insidious overweight and obesity endemic are to be considered. The clock is ticking and inertia prevails in far too many settings.

There have been some notable improvements in patient safety and risk reduction through standardization of processes, the implementation and utilization of evidence-based checklists for the performance of procedures and caring for patients, and enhancing communication skills. Yet enormous challenges persist, especially in the realm of diagnostic errors, all of which represent delays in accurate diagnoses and provision of correct and appropriate therapies. We have much to learn and much to do, and in my view we remain too slow off the mark in addressing this crisis in care.

Higher reliability and why rigorous investigation matters

Quintessential characteristics of higher reliability industries are the principles of risk anticipation, cultural risk awareness and commitments to understanding and correcting flaws in systems, processes and human performance factors that can result in harm and even death.

To achieve highest quality in healthcare we must strive to embed these same characteristics, supplemented by understanding the inherent contributions and interactions of structures and processes that must align to achieving best outcomes. It is only by truly delving deeply into causality that opportunities for improvements can, and will, be identified.

Unfortunately, even that might not be enough because in healthcare, all too often, actionable improvements are not realized from investigations. Too frequently, improvement processes are not put into place and even when they are, they may not be sustained. Fundamentally, if the healthcare industry is to emulate the characteristics of higher reliability industries then it, and we, must diligently strive to learn from investigations of safety incidents.

We must identify the system and process errors and human mistakes that result in injuries and deaths, and we must do so in a systematic fashion that leads to actionable improvements in the provision of healthcare services and thus in patient safety. In other words, we must put into action the improvement opportunities we identify from our patient safety investigation efforts.

Sadly, more often than we would presume, so-called “root cause” analyses are inadequate. They regularly focus on the performance of individuals in “blame and shame” fashion instead of truly understanding the complexities of contributing factors that result in causality at the tip of the needle. Clinicians do not wake up in the morning intending to harm patients, so when harm occurs, we must dissect the reasons for this so we can learn and make progress in our efforts to improve.

Improving incident investigations in healthcare

Although commendable efforts have been made to investigate patient safety incidents in healthcare, attempts to standardize processes of “root cause” analyses have been inconsistently utilized and applied. In too many instances, investigative teams have been under-resourced, team leaders have lacked authority and collaborative leadership skills, an assumption has been made that all clinicians have suitable skills to conduct investigations, and investigation teams frequently have not included important subject matter experts who can bring new perspectives into discussions. Fortunately, new strategies to address these deficiencies have been identified. Enhancing the focus on structures and processes leading to identification of opportunities for learning and putting improvements into place is a key to these strategies.

Even the term “root cause” is a misnomer, as it suggests that there should be easily identifiable and discrete causes for incidents, but that is often not the case. If one pulls a dandelion out of the ground, one immediately appreciates that the long root has numerous rootlets extending outward. In healthcare, especially in the realm of human error, more often than not there are numerous rootlets, or contributing factors that collectively result in causality.

Unless investigations delve deeply to locate the contributing factors affecting human performance, then important learning and improvement opportunities will not be identified and little if anything truly actionable will result from the investigation.

If the outcome of an investigation, the “root cause” identified, is that Dr. Smith or Nurse Jones made an error, a mistake in judgment or performance, then the result of that investigation has been to assign blame, not to improve healthcare and reduce risk. In my view, if the result of an investigation has assigned “root” causality to an individual, then that investigation has not delved deeply enough to identify the numerous contributing factors that affect human performance and that collectively result in causality.

This is not to say that individuals are not responsible for their actions or inactions or should not be held accountable. Rather, that there are numerous, confounding, and overlapping contributing factors governing human behavior that collectively result in causality. Only by identifying these human factors will improvements be put in place to mitigate their influence and to improve safety.

Leaders and managers need to understand this and to support robust, intellectually succinct and unbiased investigations. We need to do a much better job, and the patients we are privileged to serve are waiting and watching.

Case study

A morbidly obese patient was admitted for a cholecystectomy. Since the surgery was complicated by access problems related to the patient’s obesity, the surgeon converted the planned laparoscopic procedure to an open cholecystectomy. At one point, a tiny bowel puncture was suspected but not confirmed.

Post-operatively, the patient complained of mild abdominal pain, some nausea and anorexia. After three days, she was discharged home. Her exam at discharge, as reported in the surgeon’s discharge note, revealed mild abdominal tenderness, attributed to incisional pain. She was advised to call for any problems and to return for a surgical outpatient appointment in one week.

Thirty-six hours after discharge the patient became suddenly unresponsive, having developed a distended abdomen over several hours. She was readmitted urgently with a diagnosis of bowel perforation and abdominal sepsis. Despite operative intervention and intensive support, she died 48 hours after readmission.

The hospital performed a root cause analysis, and the findings concluded that this “unfortunate incident” was primarily related to the possible bowel perforation, a known complication of surgery. The surgeon had acted in good faith.

The patient’s family filed a malpractice claim, and a second investigation, by an outside surgical consultant, was conducted as part of the hospital’s risk management process. This investigation noted that the patient’s vital signs post-operatively had revealed a slow, yet consistently increasing temperature and pulse and slowly decreasing blood pressure. On the morning of discharge, the patient had complained of anorexia to the nursing staff, had refused breakfast and her temperature was 38oC. The patient’s surgeon was unaware of these symptoms and findings and had not discussed his patient with the nursing staff.

This combination of symptoms and findings, at minimum, should have suggested to the surgeon that the patient had emerging abdominal sepsis related to bowel perforation. This patient’s bowel perforation was a treatable condition and her septic death was possibly preventable. The patient died because the surgeon failed to pay attention to objective signs of evolving infection. The surgeon was responsible and accountable for what had happened.

If this were the end of the story, and the “root cause” was the surgeon’s inattention to details, then few opportunities for learning would have been identified. But this was not the end of the story.

When the surgeon was interviewed as part of a subsequent professional competency investigation, the following factors were identified that may have contributed to his inattention to detail. The surgeon had been double-booked for office duties the day before the patient’s discharge, as his partner had become suddenly ill. The surgeon had taken an additional consecutive night on call that evening as a result of his partner’s illness and had been up until 4am with an urgent surgical case.

He had had two hours of sleep prior to making ward rounds on the day of discharge and was due to travel that day to visit his elderly father, recently diagnosed with acute leukemia, but had not yet had a chance to pack for his early morning flight. Thus he was exhausted, task saturated, emotionally distressed and rushed.

Yes, the surgeon was responsible for the care of this patient. Yes, he was accountable for what happened. Yes, he made mistakes, but human beings make mistakes. He was faced with compelling circumstances that day, and the cumulative effect of multiple “hidden” human factors degraded his performance. His mistake, though deeply regrettable, was understandable given the pressures affecting his life. He was cautioned but not subjected to professional sanctions and was referred for counseling as a “second victim”. Several opportunities were identified for improvements in processes and structures for care and were identified.

Conclusion

Only by improving the quality of investigations can real opportunities for learning be identified and can actionable improvement strategies be developed, put in place and sustained. Investigations must be thorough and must be competently performed, using the right people in the right settings in the right timeframes, and by valuing the contributions of those who devote their efforts to patient safety.

Placing a high value on incident reporting and enhancing the quality of investigations by delving deeply to identify contributing factors are fundamental pillars of the high reliability culture that the healthcare industry must strive to achieve. Establishing this quintessential paradigm is a key responsibility of leadership.

Dr Dan Cohen is chief medical officer of Datix. He can be contacted at: dcohen@datixusa.com