How Can We Learn from Incidents?

Christine Holzmueller, BLA

Peter J. Pronovost, MD, PhD
Richard Branson, BA, RRT

Patient safety and quality care have become important issues in healthcare, particularly critical care. The Society of Critical Care Medicine (SCCM) is committed to advancing this healthcare initiative and, as a result, created an Advisory Review Panel on Patient Safety. This Panel is made up of ICU physicians, nurses, respiratory therapists, pharmacists, and other safety researchers. The charge of this Panel is to:

• Review cases from the Intensive Care Unit Safety Reporting System (ICUSRS) project (an anonymous Web-based incident reporting system)

• Discuss ways to improve safety using panel member expertise and evidence-based medicine

• Broadly disseminate their findings to the critical care community
  

The ICUSRS project is an incident reporting system that currently collects data on adverse events and near misses in 23 intensive care units in the United States. The reporting process is voluntary, anonymous to reporter and patient, and the data collected is kept confidential. This project is directed by Peter Pronovost, MD, PhD, who is also the chair of SCCM's Advisory Review Panel on Patient Safety.


The Panel's format for disseminating information is to select a case illustrative of a safety issue, provide a case overview, dissect and pull out the multiple system failures that contributed to the incident, and provide safety recommendations designed to decrease the likelihood of recurrence. Using this format, the following is an example of an adverse event analyzed by the committee.

Adverse Event: Inappropriate Ventilation Support

Case Overview
A 63-year-old trauma patient arrived from the emergency department with abdominal distension and deteriorating respiratory status. On initial evaluation, it was found that the patient suffered from duodenal trauma and perforation. The patient's respiratory status progressively declined, prompting the clinical staff to manually ventilate the patient using a bag valve mask. Failure to improve the patient's respiratory status, which was a direct result of his/her abdominal distension, was met with more aggressive manual ventilation. The intensivist arrived, noted the aggressive ventilatory management and worsening abdominal distension and immediately altered the ventilation process. Prior to intubation, the patient aspirated gastric contents into the right lung and subsequently developed ARDS.

In evaluating this incident, five system failures contributed to patient injury. Furthermore, each system failure had an impact on the next, like a domino effect, resulting in the patient developing ARDS. First, let's dissect this case and then use James Reason's 'Swiss cheese model' to capture a visual of this adverse event. 1 In this model, the holes represent latent organizational vulnerabilities; when these vulnerabilities align and defenses break down, harm can occur.

System Analysis

System Failure 1: Poor Communication. Although it is not clear from the report what transpired at the time of transfer, staff should have communicated with senior staff to assess the patient's airway status before transfer. And, airway should have been secured using an endotracheal tube prior to transfer from emergency department to ICU (System Failure 2: Transfer guidelines).

System Failure 3: Lack of supervision during transfer and inadequate training and mix of staffing skills (System Failure 4) led to poor decision making and clinical management. The patient's airway should have been ventilated by face mask until saturations were adequate and then intubated if needed. The patient's deteriorating condition should have been monitored more closely. In the absence of staff trained in intubation, other options should have been available (e.g., laryngeal airways) and manual ventilation altered to limit gastric insufflation.

System Failure 5: Poor ventilation technique. Previous evidence has shown that manual ventilation with a self-inflating bag and face mask was a poor choice because this method can result in gastric distension (an initial diagnosis for this patient), reduced chest wall compliance, alveolar hypoventilation, and an increased risk of aspirating gastric contents (the harm this patient experienced). 2,4

System Failure 6: Inadequate patient monitoring made it difficult for the staff to notice the worsening of this patient's abdominal distension, which caused aspiration and subsequent ARDS. 5 Reason's Swiss cheese model6 is based on the premise that incidents in healthcare are complex, and rarely involve one action or system failure. Each slice of cheese in this model represents a specific element in the delivery of care (e.g., ventilating a patient). The slices are riddled with imperfections or breakdowns in each element of the system (e.g., proper staff training to perform a ventilatory procedure). These breakdowns are represented by the holes in the cheese slices. When the holes line up, an adverse event occurs.

Safety Recommendations

Staff Training and Education:

•  All clinical staff should be trained in basic life support using ECC/AHA guidelines. 7 Trained staff should also be inserviced annually and/or when standards of care change. In the case of ventilation, training should emphasize proper methods for manual ventilation (e.g., smaller tidal volumes, low flows or squeezing bag slowly, low airway pressures and use of cricoid pressure for prevention of abdominal distension). 8,10
  
•  All clinical staff with patient interaction should know all of the options available for airway management (e.g., laryngeal mask airway, cuffed oropharyngeal airway (COPA) and Combitube)11 and consider simulation training to gain experience in the use of the equipment.

Policies and Procedures:

• Change hospital policy to allow allied health professionals to intubate patients in an emergency situation when a physician is not available. In many hospitals, respiratory therapists can perform endotracheal intubation.
  
• Change policy to require patient monitoring of airway pressure during manual ventilation. To avoid gastric distension, note that airway pressure should be less than 20 cm H2O, given you can maintain adequate oxygenation. 12

To improve patient safety, we must learn what is broken, through incident reporting, and use that information to change our systems. In an effort to present the lessons learned from a case, the Advisory Review Panel has adapted this case to a one-page summary sheet (Figure 2) that can be posted on a bulletin board or used in presentations. The Society would like to know which methods of summarizing incidents are most useful to you and would appreciate your feedback on the methods that best facilitate improvements in safety.

The Society will continue to dedicate an article in Critical Connections for disseminating safety suggestions made by the Advisory Review Panel on Patient Safety. Future articles will address both adverse events and near misses and will include case presentations. Our goal is to help you improve patient safety in your ICU.

Acknowledgements
We would like to acknowledge all the other members of the Advisory Review Panel on Patient Safety for their contributions; including, Paul Barach, MD; David Kaufman, MD, FCCM; Lynn Kelso, MSN, APRN, BC; Ruth Klienpell, PhD, RN-CS; Lisa Lubomski, PhD; Diane Lyle, BS, PharmD; Steve Martin, PharmD, FCCM; Patricia McGaffigan, RN, MSN; Nancy Stonis, RN, BSN, MJ; Ann Thompson, MD, FCCM; and David Thompson, DNSc, MS, RN.