If you want to go far,go together: A guiding principle for learning from medical errors

When a patient is harmed by health care, there is a strong expectation that the organization will learn from the mistake to prevent it from happening again.^1–3 Health care organizations are obliged to examine what went wrong, understand why, and change their systems so that future patients are protected.

Patients and their families share this expectation. In disclosure discussions with the physician, they want the facts about what happened and someone to take responsibility. They also want an apology and assurance that action will be taken to prevent the same event from happening to others. ⁴ They would be shocked to know that we do not do this well.

The reality is disappointing. It is bracing to realize that health care may not be any safer today than it was 25 years ago, even after a generation of patient safety efforts.^5,6 Some things are better. Data from the American Hospital Association and Vizient show that hospitalized patient in Q2 2025 were nearly 30% more likely to survive compared to Q4 2019, and there were 24% fewer central line-associated blood-stream infections and 25% fewer catheter-related urinary tract infections. ⁷ However, adverse events were identified in nearly one in four hospital admissions, ⁸ and the incidence of sentinel events like wrong site surgery is stubbornly persistent. ⁹

Many, if not most, incidents are not reported.^10,11 One of the most discouraging reports is from a study by Classen et al. ¹² who found that traditional voluntary reporting captured only 1% of adverse events found using the Global Trigger Tool. A 2025 report from the US Office of Inspector General (OIG) found that hospitals did not capture half of patient harm events in their reporting of incidents that occurred among hospitalized Medicare patients. ¹³ If these figures are correct, only a small proportion of harm events are available to learn from.

There are barriers at the individual and organizational levels that explain the failure of institutions to learn from their mistakes. ¹⁴ Individual barriers include the constant time pressures that make it difficult for a health care worker to report an event. They may be uncertain about what should be reported, and their outcome expectancy may be low—having not seen anything come from past reports, they believe that reporting won’t result in change. Fear of blame discourages workers from reporting to their superiors. At the organizational level, systems tend to place a higher value on getting things done than on making improvements, and managers signal to workers that productivity is more important than identifying problems.

The full African proverb quoted in the title of this paper is, “If you want to go fast, go alone. If you want to go far, go together.” This can apply to when resourceful individuals devise patches for broken systems, which allows these flaws to persist, preventing organizational learning. For example, when a nurse faced with a stocked-out medication borrows a dose intended for another patient, this ensures that the supply will still be missing the next day.

The usual starting point for transforming an adverse event into information that improves patient safety is the Event Reporting System (also referred to as Reporting and Learning System). Several steps are needed for this to happen. These include (1) recognition of a preventable harm event, (2) the decision to report, (3) successful reporting of the event, (4) receipt and review of the report, (5) identification of lessons and actions to be taken, (6) dissemination of the lessons, (7) changes in practice, (8) monitoring of outcomes. The completion of these steps is influenced by the prevailing culture, of both the microsystem and the organization, and the robustness of organizational infrastructure and management.

The successful execution of each of these steps requires certain conditions. The recognition step requires the awareness that an error has occurred that harmed the patient. The decision to report is influenced by the individual's fear of blame or punishment, time pressures, and the psychological safety of reporting. Reporting systems and software vary in accessibility and usability. Would-be reporters can be thwarted at this step. Reports must be received by a functioning system and be effectively reviewed and processed. Identifying contributing factors and how to address them requires experience with information gathering and systems thinking. Lessons need to be disseminated to multiple levels within the organization so individuals and teams can understand and act on them. Infrastructure is needed to implement changes. Implementation needs to be evaluated to confirm improvement and monitored to ensure sustainment.

John Eisenberg, ¹⁵ one of the fathers of patient safety, depicted the gap between potential quality and actual care as a series of voltage drops. Each step in the diagram represents a point where the intended benefits can trickle away. When applied to the goal of learning from mistakes, even well-designed systems are prone to losing power before safety can be improved.

Despite the barriers described above, it is possible for a local incident to be recognized and converted into improved care across an institution and beyond. For example, high-flow nasal cannulas came into wide use a few years ago to deliver high-flow humidified oxygen. These devices are bulkier than standard nasal cannulas. At one hospital, soon after they were introduced, an intensive care unit nurse noticed that it had caused a pressure injury to a patient's nose. The nurse repositioned the device and added padding, and alerted other personnel to the problem, including aides, respiratory therapists, and physicians. The nurse also posted an incident report on the institution's online event reporting system, noting temporary harm. The report was read by the patient safety nurse assigned to examine all incident reports and routed to appropriate hospital committees. The Skin and Wound committee proposed adding padding between the device and patient and circulated this recommendation to Respiratory Therapy and all nursing units. The recommendation was also sent to the manufacturer of the device. Patient Safety subsequently conducted a compliance audit to assess the implementation of this remedy.

The papers in this issue of the Journal illustrate many aspects of why it is so challenging to learn from adverse events. Kippenbrock and Emory ¹⁶ conducted a secondary analysis of data from a large sample of US clinicians. They found that advanced practice nurses rated patient safety lower than physician assistants and physicians, and reported errors more frequently. The authors recommended team-building interventions to increase cohesiveness among clinicians around patient safety.

Ferorelli and colleagues ¹⁷ emphasized the importance of understanding health care professionals’ attitudes to adverse event reporting. They presented evidence for the reliability and validity of the Italian version of the Reporting of Clinical Adverse Events Scale and recommended that it be used to enhance workers’ understanding and reporting of adverse events.

Kane and colleagues ¹⁸ discussed measurement problems that make it difficult for health systems to understand the impact of their safety improvement efforts. They recommended that safety events be captured and triangulated using multiple sources of data to provide reliable estimates, that institutions be vigilant about biases in the measurement of events, and that definitions of events be harmonized to allow comparison.

Kapur and colleagues ¹⁹ found that while nurses in a New Delhi hospital had adequate knowledge and positive attitudes toward reporting medication administration errors, reporting was hindered by fear of being punished and of losing the trust of patients and families. The authors suggested that fostering a culture of non-punitive reporting, as well as training and streamlining the reporting process would increase reporting.

Harrington and colleagues observed that in the United States, a key factor in the inconsistent implementation of communication and resolution programs (CRPs) has been physician hesitation to embrace the approach. ²⁰ These programs incorporate health care organizations' response to patient harm, including disclosure, event review, quality improvement and, when required, an apology and offer of compensation. The authors identified the central role of emotional barriers to physician engagement, including fear, uncertainty, and shame. They recommended that institutional leaders incorporate data to persuade physicians that CRP is integral to the clinical mission and is in the interest of both patients and physician.

They might be encouraged by the study by Sokol-Hessner and colleagues, ²¹ who reported on a systematic review of the effectiveness of CRPs. They found either positive or neutral effects of CRPs on organizational liability and cost outcomes. If this became widely known, it should help institutions appreciate the benefits of adopting CRPs.

Berry observed that investigation of patient safety incidents in the British NHS relies on systemic learning, with personal accountability based on “no blame” and “just culture” frameworks. ²² However, he suggested that after harm these frameworks may fail to meet expectations of patients and families. He presented an alternative approach termed “responsibility culture” which strikes more of a balance between systemic and personal causation. He acknowledged that substantial change would be need within organizations to minimize fear and scapegoating among medical staff.

Macrae explored problems that limit the effectiveness and impact of safety recommendations. ²³ He argued that greater rigor is needed in defining the safety risks that need to be addressed, and proposed principles to support a more systematic and integrated approach to developing and applying safety recommendations.

Figueroa and O’Connor ²⁴ noted that local safety events often fail to be communicated more widely within a hospital to prevent repeat events. They developed a process that units can use to assess the risk of a similar event occurring in their work area. Their preliminary findings suggest the process helped other units estimate their own risk, and when appropriate, make plans to prevent recurrence.

In conclusion, the expectation that health care organizations will learn from patient harm is well established — and frequently unmet. Harm remains common, many events go unreported, and few incidents are successfully converted into system improvement. Barriers at both the individual and organizational level disrupt the chain of steps needed to move from a harmful event to safer care. Individual barriers such as time pressure, fear of blame, and poor reporting systems combine with weak organizational follow through ²⁵ and tendencies to prioritize throughput over learning. The articles in this issue address this challenge from multiple angles—measuring harm, improving reporting, supporting disclosure, and spreading lessons—and collectively point toward the same conclusion: meaningful learning requires culture change, better infrastructure, and sustained leadership commitment.