Dealing with an institutional crisis

When calamity strikes, despite excellence in diligence and attention to detail, what do you do? Dan Cohen, chief medical officer for Datix, offers his advice.

The healthcare environment is inherently unsafe. Our structures and processes are not always aligned for success and are plagued by inefficiencies and insufficiencies. Thrown into this mix are the liabilities that well-meaning people bring to the workplace, particularly the complex human factors that affect cognition, behavior and performance.

Add to this the communication challenges that permeate the soil within which best clinical outcomes must be cultivated, and one can easily get a sense of just how complicated and dangerous providing and receiving healthcare services can be.

As a result of this complexity and the multitude of variations in structures, processes and human factors issues, patients will continue to be harmed and some will die. It has been recently reported that as many as 250,000 patients may die and, by extrapolation, several orders of magnitude more will be harmed each year as a result of receiving healthcare services.

These figures are the subject of much debate. Yet even if “only” 10,000 people were to die each year, healthcare-associated deaths would still constitute a public health crisis of enormous proportions. The fact is that baseball stadiums full of mothers, fathers, grandparents, children, friends and lovers are harmed or die every year.

Despite our best intentions and our best planning, hospitals and healthcare systems are from time to time affected by sudden crises that are unexpected and potentially catastrophic for patients, staff and institutions. These predictably unpredictable crises often represent dramatic failures in structures and/or processes. Alternatively, they may simply represent lapses in human performance, which on rare instances may result in catastrophic incidents, resulting in deaths and disabilities.

The challenges of healthcare-associated infections

The frequency and severity of healthcare associated infections remains a major challenge for healthcare professionals, the institutions in which they work and, most importantly, for patients. Although there have been some dramatic reductions in many infections, including those caused by methicillin-resistant Staphylococcus aureus (MRSA) and Clostridium difficile, much more work remains to be done and we must double our efforts.

We are challenged by the need for continuous surveillance, and on occasion we are sobered by dramatic outbreaks despite our best efforts and despite all of our training and diligence. Infectious disease outbreaks are among the most feared catastrophic incidents because they often affect the most vulnerable patients, resulting in terrible consequences and enormous institutional liability. Although our planning for infection control is well-established and all hospitals have infection control programs, policies and procedures, one can predict with reasonable certainty that at some time an outbreak will occur. Forewarned is forearmed, and all institutions should have in place not only concise plans for emergency infection outbreak responses, but as important, emergency response plans for all major safety incidents that can be rapidly implemented: see Respectful Management of Serious Clinical Adverse Events (Second Edition). IHI Innovation Series white paper. Cambridge, Massachusetts: Institute for Healthcare Improvement; 2011 (available on www.ihi.org).

It has been well established that hand-washing can be an effective preventive measure as part of any hospital infection control program. Yet what is also known is that any particular single clinician/patient interaction is extremely unlikely to be the source of an infection, let alone an outbreak affecting many patients. The fact that this is the case is actually part of the problem as considerable complacency about risk sets in and affects human behavior.

When time and again nothing really happens, when patients do not become infected even if clinicians forget to wash their hands, the importance of hand-washing becomes unconsciously undervalued. This is particularly the case in high-acuity environments where task saturation and distractions are common. Under these circumstances, deviation from standards of care for hand-washing becomes normalized as part of our behavior, and this “normalization of deviation³” actually feeds complacency. See: Banja J. The normalization of deviance in healthcare delivery. Bus Horiz 2010;53(2): 139-148 (doi:10.1016/j.bushor.2009.10.006)

Case study

A major academic medical center, one with a reputation for an outstanding infection control program, experienced a neonatal outbreak of highly virulent Group A beta-hemolytic streptococcal infection, Streptococcus pyogenes, affecting seven infants over a two-week period, two of whom died. This institution had recently been cited by an accreditation agency as an example of excellence for its high diligence and attention to detail in its infection control program.

The outbreak occurred in February during a peak of respiratory illnesses in the community. The index case was a six-day-old, 32-week premature infant with respiratory distress syndrome (RDS), born after premature rupture of membranes and a spontaneous vaginal delivery. The mother was entirely well, although she had had a recent minor urinary tract infection and post-partum she remained healthy.

Because of moderate RDS, the infant was intubated and received continuous positive airway pressure and mechanical ventilation. An indwelling umbilical artery catheter was inserted in accordance with a standard protocol. On the fifth hospital day, though the child had been making substantial progress, she became ashen and mottled in association with hypoglycemia and a basal temperature of 96.6^oF, all signs of classic neonatal sepsis. She underwent a sepsis workup and both blood cultures and a culture of her umbilical stump revealed Streptococcus pyogenes, a common respiratory pathogen that can potentially cause serious infections in vulnerable patients. Unfortunately, this little girl died of fulminant septic shock within 24 hours. Within four days, three subsequent cases of streptococcal illness were identified, as were an additional three colonized patients.

S. pyogenes is an uncommon but potentially devastating cause of neonatal sepsis. As part of this hospital’s comprehensive response plan for infectious disease outbreaks, an urgent investigation was undertaken, with initial primary focus being on 1) a maternal source of infection, or 2) the insertion of the umbilical artery catheter.

The mother was now well, at home, her pelvic exam was as expected for a woman six days post-partum, and cultures of vaginal discharge/fluids did not reveal S. pyogenes. No breaks were identified in the protocol for umbilical artery catheter insertion and the individuals involved in the catheter insertion were healthy, with no skin lesions or respiratory symptoms. Cultures of equipment, nares, throat, and nail beds of the staff involved in the catheter insertion were all negative, although these individuals were removed from patient care as a precaution, pending culture results.

Two additional infants became ill and both had been patients in the neonatal intensive care unit (NICU) before moving to the regular newborn nursery, a step-down unit for the NICU. One of these infants died. Subsequently, a third possibility was considered: transmission from nursery staff not involved in the catheter insertion procedure.

Cultures were performed of skin, nares and throats of all nurses, technicians and residents and these staff were removed from the NICU for 48 hours and placed on antibiotics, pending culture results. Cultures were performed on all infants in the NICU and regular newborn nursery and umbilical stump cultures were positive in three additional instances. These infants were asymptomatic and thus likely only carriers, although they were treated as a precaution.

A very strict hand-washing enforcement policy was put in place, involving direct observation by staff colleagues. The local public health department was notified, as was the Centers for Disease Control, from whom expert advice was sought. The newborn nursery was thoroughly disinfected and the NICU was closed to new admissions for 96 hours. All infants were treated with antibiotics and close follow-up was instituted for discharged infants, especially those who were found to be infected or colonized. The hospital held a news conference and via various social media channels, distributed information to the press and community about newborn S. pyogenes infection and measures being taken to contain the outbreak.

The investigation revealed that all of the infected or colonized infants had, for some varying periods of time, been on treatment beds that were side by side in the NICU, with two nurses providing most of the care to both patients. Cultures of the nares from one nurse, and one nurse only, revealed the pathogen in question, and it turned out that this nurse’s six-year-old child had recently had a throat infection that was treated with antibiotics, though a throat culture had not been taken. This nurse had not reported her child’s infection to her nursing supervisor, though her child’s doctor had diagnosed her son with a ‘strep throat’.

This young nurse was known for her meticulous attention to detail, though she could not be 100 percent certain she had always washed her hands between touching the side-by-side patients, nor that she had not inadvertently touched her nose with her hands. As we all know, no one can be certain of everything all the time and we all make mistakes. This nurse was removed from direct patient care until she had completed a course of antibiotics and was culture negative on two occasions. Sadly, she became very despondent, a “second victim⁴,” and she received professional psychological counseling as part of her hospital’s response plan.

An investigation into structures and processes surrounding this incident identified three areas of concern: 1) the nurse implicated as the source of infection should have advised her nursing supervisor that she had a child at home with presumptive streptococcal pharyngitis, given the high family carrier rate in such circumstances; 2) the hospital did not have a written policy or training program that emphasized and required staff to report family illnesses to supervisors; and 3) the necessity to put in place additional warning prompts and reminders about hand washing, especially in circumstances where nurses are caring for side-by-side patients, and to provide access to cleansing foam at both the foot or head of each neonatal intensive care unit bed in side-by-side care circumstances.

The hospital was sued and accepted considerable liability for two deaths and two other infections that resulted in prolonged hospital stays, morbidity and severe parental anguish.

Conclusions

This case study reveals how even in institutions with outstanding infection control programs, and with highly trained and committed staff, the possibility of an infectious disease outbreak always exists. More often than not, human factors and human behavior will be identified as contributing factors resulting collectively in causality. The excellent and prompt response of this hospital’s infection control team possibly halted an even more extensive epidemic that could have had substantially more dire consequences. Furthermore, the hospital’s emergency incident response plan proved invaluable, meeting the needs of patients, staff, the hospital, and the larger community.

Risk modulation and risk prevention require an in-depth understanding of human performance factors, especially the impact of high-acuity workload factors, distractors and task saturation on human performance. Professional complacency, particularly in high-intensity environments is, sadly, a pernicious influence that is difficult to combat and that cannot be overemphasized in training.

Normalization of deviation is common. Healthcare professionals need to appreciate that although our intentions may be benevolent, because of our intrinsic human liabilities, we are all inherently dangerous. Bad stuff happens, we can prevent most of it, but we cannot eliminate all risks. Providing health care is just too complicated and our most vulnerable patients are those most at risk.

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