Evolution of a patient safety program in an academic medical center

Background

This paper documents the three development phases taken at a large Academic Medical Center to improve the hospital's safety culture through the creation of a robust patient safety program. Since the publication of To Err is Human: Building a Safer Health System in 1999, ¹ there has been a steady evolution in the understanding that most harm in healthcare results from system problems rather than human behavior. An improving safety culture depends on leaders requiring strong and safe systems of care to be put in place in support of employees. Nationally, change has not come easily as shown in several follow-up studies.^2,3 Establishing a strong safety program not only benefits patients, but also the hospital's publicly reported data, reputation scores, and pay-for-performance programs.

National standardized culture of safety surveys highlighted the hierarchical and punitive culture in healthcare. Intimidating behavior by leaders, and caregiver's fear of reporting adverse events, has historically led to an underreporting of harm and failure to correct the underlying causes. Caregivers do not come to work with the intention of causing harm to patients, but faulty systems may lead them down dangerous paths. The principle of “blameless vs blameworthy” harm has helped give a construct for leaders to manage communication and harm with frontline workers. ⁴ A further tool highlighted has been “Just Culture” which balances safety with accountability. ⁵ The challenge for any healthcare system is how to meld all these concepts into an effective and functional safety program.

The origin and impetus for this program was the Medical Center's decision to make Clinical Quality Improvement the #1 strategic priority. This decision was made because the Academic Medical Center had earned a 1-Star CMS ⁶ designation and received a Leapfrog Safety Grade of an F ⁷ in 2015. Improvement was approached using high reliability as the platform incorporating the three pillars of Leadership, Safety Culture, and Performance Improvement. ⁸ Safety culture requires leadership to make it a priority and to be engaged in implementation; and for all caregivers to embrace necessary change and embed patient safety in their daily work. This program grew by incremental steps, creating sustainability, and benefiting patients and the organization.

This paper will outline the steps taken, lessons learned, and effectiveness of such a program. The patient safety program evolved in three phases over five years:

Error Prevention Training

Methods

The common theme to each phase of program development was a consistent infrastructure to set priorities, develop and implement programs, track and review progress, and celebrate successes. Specific to this patient safety program, phases were organized and overseen by the Chief Quality and Patient Safety Officer and Chief Medical Officer working in collaboration with hospital leaders. The program's success depended on these leaders building engagement and participation of the entire workforce adopting patient safety as a core mission. From phase 2 onward, the hospital worked in collaboration with The Joint Commission Resource (formally the Center for Transforming Healthcare) using high reliability as the platform to provide standardization of systems to help achieve the goal. ⁹

The strategic priority, Clinical Quality Improvement, was established as a data-driven program. The guiding principle was to display simple, timely, actionable data drillable at the institution/department/unit levels. The Quality & Patient Safety Scorecard (Scorecard) was created as the one source of truth for data used for improvement and internal reporting. The Scorecard had three domains: patient safety, patient outcomes, and patient experience. Each domain has metrics which include baseline data and annual targets. Data were updated on the 15^th of each month with the prior month's performance. The evolution of safety data, such as the volume, review, and location of harm events and specific scorecards for hospital-acquired infections, patient falls, and pressure injuries was a key component of the program. This new “source of truth” provided the North Star for all caregivers and hospital leadership.

Results

Phase 1: Error prevention training

The total number of employees trained during the initial training from 2014 to 2015 was 7797. The annual average number trained from 2016 to 2022 was 220. The program was well received with consistent positive feedback. Over time, there were minor updates, and its ongoing use was achieved by leaders pushing for participation. This program impacted the clinical quality improvement strategy by preparing caregivers to engage in improvement priorities.

During this time, safety culture improvement was documented from the AHRQ Patient Safety Survey as shown in Figure 1:

OPEN IN VIEWER

Figure 1.

Culture of safety survey results 2016–2018 which shows a 20% average improvement in the dominant domains.

Overall perception of safety (2016–2018) improved and there was an improvement in three focus areas: event reporting, nonpunitive response, and hand-off communication. Of note, all other dimensions of the patient safety survey showed improvement year over year, but by 2018 only 3 of 12 dimensions scored above the AHRQ mean.

Phase 2: Event reporting and review

Prior to 2015, UMMC only tracked data on serious harm events which averaged 5–6/month. Sentinel events were investigated and managed by CRM. The restructured event reporting program brought a different philosophy by educating all caregivers to report any event where patient care did not go as planned. The goal was to markedly increase what was reported, make events learning opportunities and advance the UMMC safety culture.

In 2016–2017, following the restructuring of I-CARE, an education program was implemented for all caregivers emphasizing the priority to report and how to use the easy, new system. The new system for review of events required initial sorting of reports by CRM, events with harm or significant concern were investigated by CRM, and less serious events were referred and managed at the unit level. The goal was to create local ownership by training managers or their designee to review and close events in a standardized and timely manner. The large number of less serious events were managed this way and escalated to CRM if indicated.

In parallel with enhanced event reporting, new systems were developed to track, report, and analyze the data generated. The Scorecard was used to share event reporting data. The main elements of data were: total events reported, percent reviewed in <7 days, temporary or permanent harm, and need for additional treatment. These data were drillable by hospital units. Figure 2 summarizes the changes for hospital inpatient units in reporting pattern with review completed in a timely manner (seven days) and event severity in the initial five years of the program:

OPEN IN VIEWER

Figure 2.

Event reporting 2018–2022 shows the increase and stabilization of reported events, the improved reviews and the small, but continued, percentage with harm.

The overall volume of events reported after the initial I-Care update in 2017 was 600/month and increased to 800/month. Event reviews completed in <7 days steadily increased to >90%. The percentage of reported events that needed any additional treatment varied between 15% and 18% and reported harm events between 2.7% and 5.4% with >95% being temporary harm. The small number of events reported with permanent harm (9–17 per year) fluctuated and was one reason for the implementation of Chasing Zero: Eliminating Harm.

Phase 3: Chasing zero

The timing for implementation of Chasing Zero: Eliminating Harm as a patient safety program was chosen after the initial two phases above were established. Chasing Zero was implemented three years into the overall patient safety program. Concepts from the first two phases, an event reporting system and reliable Scorecard data, made this program possible at both the institutional and unit levels.

As shown in Table 1, in the first year of the program there were a total of 11 Chasing Zero events leading to a reset of the clock. The longest period between events in this timeframe was 95 days. The event summary on the Clinical Intranet was limited to <100 words and structured to:

State the event,

Give a brief, deidentified description, and

Educate with lessons learned.

OPEN IN VIEWER

Table 1.

Chasing zero harm first year reporting on the clinical intranet shows how the “time since” was tracked and the number of events for which the clock stopped in that year.

Clinical intranet: Chasing zero (2017 data)
Longest period between events: 95 days
Never event	2017 count
Irretrievable loss of irreplaceable specimen	1
Medication error resulting in serious injury or death	1
Administration of wrong blood product	1
Fall resulting in serious injury or death	2
Unintended retention of foreign object	3
Wrong patient/site/procedure	3
Grand Total	11

An Example of a Chasing Zero Summary:

Event : Wrong Site Procedure—A drain was placed for intra-abdominal fluid collection under radiologic guidance in the wrong location necessitating the subsequent placement of a second drain in the correct location.

Findings:

Documentation of the problem requiring intervention was lacking.

The order was not clear as to be requested location of the drain.

Communication between the ordering provider and proceduralist did not occur for clarification of the order.

Action: Accurate and complete documentation and communication are essential to optimize patient-centered care.

The selected events were limited to six to maintain focus and cover a broad spectrum of services. Events were reviewed and adjudicated by CRM and the Patient Safety Officer. Minor adjustments were made to the list in year 2.

Conclusions

Primum non Nocere is a basic tenet in healthcare, but as emphasized in the 1999 To Err is Human and 2001 Crossing the Quality Chasm papers from the Institute of Medicine, UMMC was falling short of this goal. These papers from IOM ignited patient safety activities by creating awareness and introducing improvement systems applicable to healthcare. While some have decried the lack of progress in the subsequent two decades, this is probably not fair as awareness has increased reporting of adverse events resulting in apparently static numbers. More importantly, awareness has led many centers to pay attention to patient safety and develop programs such as outlined in this paper.

The approach taken has been through incremental steps, moving forward to the next phase in development, as the principles became established and accepted by caregivers. It is often said “Timing is Everything,” and this was definitely true for this program. If this program had started with Chasing Zero it would have failed as this concept is hard to accept de novo, but is better accepted once basic building blocks are established. The initial Error Prevention Training program established a common language and began the necessary culture change as the first step in what was to become the overall program.

In conjunction with the incremental approach, ensuring that the patient safety program was a data-driven program led to quality and safety improvement. Having accurate, reliable data were necessary for engagement and led to the development of a data-driven infrastructure. Starting with basic quality/safety metrics, this showed that improvement could be achieved using a data-driven approach. UMMC physicians, nurses, and all support service caregivers had visibility to this data with structured Scorecard reviews by department and location.

Event reporting in hospitals evokes mixed responses, but was a critical step in developing a sustainable culture of safety. The concept of blameless vs blameworthy adverse events helps address this and emphasizes the role of flawed systems in healthcare and the need to eliminate these. Equally, a Just Culture received attention as a means of addressing errors and establishing a level playing field for all employees. When I-CARE was restructured and increased reporting of events was used as a metric, there was pushback. It was important to emphasize that reporting was nonpunitive and was the only way to identify events to use as learning opportunities. As shown in the Culture of Safety survey data in Figure 1, the organization moved in the right direction. The increased reporting led to an improved understanding of why following up on events was important, positive interactions during RCAs, and overall participation in action planning to reduce adverse events.

Has the goal of eliminating harm, as articulated in the Chasing Zero campaign been achieved? Not yet, as the total number of reported harm events, Figure 2, has not changed significantly. Individual locations or programs have made significant strides in this direction: for example, Children's Hospital went 223 days without a CLABSI in 2023, and the Pressure Injury prevention team saw a 48% reduction in severe PIs in 2022–2023. In the literature, there has been debate on the usefulness of Chasing Zero with some declaring this an unattainable goal, but the experience at this hospital would still advocate for the stated goal of a Chasing Zero program.

Most importantly, care for patients has improved. Caregivers routinely see that with simple, timely, and actionable data on the intranet Quality Scorecard. In addition, it has been validated by external sources as the Medical Center has seen significant improvement in the Leapfrog Safety Grade (Figure 3).

OPEN IN VIEWER

Figure 3.

UMMC's leapfrog patient safety grades 2015–2023 shows progressive improvement from F to a high B.

How does this experience fit with other published work since To Err is Human? The concept that a “A One Size Fits All,” approach for patient safety programs does not work was put forth in 2019 by Mark. ¹¹ Different institutions have different cultures, leadership styles and can implement change at varying paces. The importance of timing was emphasized in this paper and supports that concept. Other patient safety programs such as the World Health Organization Global Patient Safety Action Plan have successfully focused on specific targets for improvement like medication safety. ¹² Collaborative patient safety initiatives, such as Solutions for Patient Safety, A Children's Hospitals Collaborative have been very successful in pediatrics and was the basis for Error Prevention Training. ¹³ Another growing area of emphasis over the past two decades has been the role of Change Management as an important tool in hospital improvement activities. Being able to incorporate components from all these programs fostered the presented system.

How generalizable this is to other hospitals will be based on their ability to take key elements of the phases presented and modify to fit their environment. The ability to develop this program in a safety net hospital, which was clinically busy and resource-limited, was achieved by keeping the phases simple and would impose little limitations for others to adopt. The program was developed and executed with minimal staff and financial resources, and the Error Prevention Training was led by volunteer trainers.

Using High Reliability as the platform, the Academic Medical Center built a robust and credible safety program producing improvements in patient safety and publicly reported quality and patient safety data. Leadership commitment to fostering a safety culture was foundational. Requiring Error Prevention Training for all Medical Center employees reinforced that patient safety was truly an institutional priority. With a changing culture and trust in the motivation behind Chasing Zero, performance improvement at the front lines became possible.