Two-Year Profile of Preventable Errors in Hospital-Based Neurology

Abstract

Background and Objectives

Medical errors are estimated to cause 7,000 deaths and cost 17–29 billion USD per year, but there is a lack of published real-world data on preventable errors, in particular in hospital-based neurology. We sought to characterize the profile of errors that occur on the inpatient neurology services at our institution to inform strategies on future error prevention.

Methods

We reviewed all cases of preventable errors occurring on the inpatient neurology services from July 1, 2018, to June 30, 2020, logged in institutional error reporting systems and reviewed at departmental morbidity and mortality conferences (M&MC). Each case was characterized by primary category of error, level of harm as determined by the Agency for Healthcare Research & Quality Common Format Harm Scale version 1.2, primary intervention, and recurrence within 1 year, with a final censoring date of June 30, 2021.

Results

Of 72 cases, 43 (60%) were attributed to errors in clinical decision making and 20 (28%) to systems or electronic health record–related errors. The majority of cases resulted in in-conference education on systems-based errors (29%) at departmental M&MCs followed by in-conference education on clinical neurology (25%). Among errors classified primarily as clinical, 28% were addressed via systems-based interventions including in-conference education on systems issues and changes in written protocol. In 23 cases (32%), a similar error recurred within 1 year of the presentation. In total, 7 cases (10%) resulted in a change in written protocol, none with recurrences.

Discussion

Systems-based interventions may reduce both clinical and systems-based errors, and protocol changes are effective when feasible. Given the important goal of optimizing care for every patient, quality leaders should conduct continuous audits of preventable errors and quality improvement systems in their clinical areas.

According to an Institute of Medicine (IOM) report, medical error is estimated to cause 7,000 deaths and cost 17–29 billion USD in the United States per year.¹ Approximately 400,000 hospitalized patients experience at least 1 preventable harmful event annually.² Despite concerning global estimates, real-world data on in-hospital errors on nonsurgical services are rarely disseminated and dissected in the literature, with a few notable exceptions.^3-7 Characterizing and understanding the profile of errors that occur within a medical specialty may inform strategies on future error prevention.

In this study, we report on all preventable errors that occurred on the inpatient noncritical care neurology services at our institution from July 1, 2018, to June 30, 2020. Based on our findings, we describe the appropriateness of the interventions in response to these errors and consider the advantages and disadvantages of our morbidity and mortality conference (M&MC) format within the context of recommended practices in the literature.

Methods

We reviewed all cases of preventable errors in our institutional error reporting systems (Sustainability Tracking, Assessment & Rating System™ [STARS®]⁸ before August 4, 2020, and RL solutions⁹ thereafter) and those logged in our departmental M&MCs from July 1, 2018, to June 30, 2020. For context, our error reporting system allows any hospital personnel to submit errors using an online system, including physicians, advanced practice providers, nurses, and ancillary staff. Our departmental M&MC model was reconfigured in July 2018 with the aim of capturing all errors occurring on the general neurology and stroke services. Before that, M&MCs were held on an ad hoc basis to discuss individual cases arbitrarily thought to warrant discussion. The new model required that senior residents log all morbidities and mortalities during their time on service in preparation for their assigned date of presentation. All senior residents presented on their service regardless of the number of errors logged. Protocol changes and other important advice emerging from M&MCs were communicated to clinical faculty and residents via email. Unpreventable errors may be discussed at M&MCs, but were not logged or included in this analysis. Preventable errors occurring in the neurocritical care unit were discussed in a separate M&MC with a different logging system and were therefore excluded from this analysis.

Cases had been previously characterized by the departmental quality improvement (QI) director based on primary category of error, level and duration of harm,¹⁰ and primary intervention. The primary category of error was categorized into diagnostic (e.g., error or delay in diagnosis), treatment (e.g., error in medication administration), preventive (e.g., failure to provide prophylactic treatment or inadequate monitoring during treatment), and other (e.g., system failure), as detailed in the IOM Committee on Quality of Health Care in America's landmark 2000 publication.¹ Level and duration of harm were categorized based on the Agency for Healthcare Research & Quality (AHRQ) Common Format Harm Scale version 1.2¹⁰ (the latter defined as permanent if greater than 1 year and temporary if less). Primary intervention was categorized as (1) change in written protocol, (2) in-conference education on clinical neurology, (3) in-conference education on systems issues, (4) discussion with leadership of a different department, and (5) other per our departmental policy. For the purpose of this study, errors were considered recurrent if repeated within 1 year of case presentation, with a final censoring date of June 30, 2021. Determinations were made based on consensus between the previous and current departmental QI directors and chief of the division of hospital neurology.

Standard Protocol Approvals, Registrations, and Patient Consents

This project was undertaken as a QI initiative. As it satisfied 19 criteria set by the Institutional Review Board classifying this work as QI, it did not require formal supervision by the board, as per institutional policies.

Data Availability

Anonymized data not published within this article will be made available by request from any qualified investigator.

Results

Seventy-two preventable errors were recorded during the study period. Forty-three (60%) were classified as primarily related to clinical decision making, 20 (28%) were primarily related to systems or electronic health record (EHR)-related errors, and 5 (7%) were related only to a procedure (Figure 1). Errors were most frequently classified as resulting in no harm followed by mild harm, together constituting 80% of cases. The majority of cases resulted in in-conference education on systems-based errors (29%) followed by in-conference education on clinical neurology (25%). Among errors classified primarily as clinical, 28% were addressed via systems-based interventions including in-conference education on systems issues and changes in written protocol. In 23 cases (32%), a similar error recurred within 1 year of the presentation. In total, 7 cases (10%) resulted in a change in written protocol, none with recurrences. Three cases were related to the coronavirus 2019 (COVID-19) pandemic (e.g., ambiguity in EEG workflow for patients infected with COVID-19). Others included changing how tube feeds are ordered in the EHR to reduce miscommunication with nursing, changing documentation practices for pulmonary function testing results to make critical data more easily accessible, changing communication methods with other hospitals to ensure that essential information is transmitted during patient transfers, and changing primary team assignment in the postanesthesia care unit to close gaps in care transition. Error types, level of harm, and interventions are summarized in Table 1.

Figure 1. Preventable Errors by Primary Category Occurring on Inpatient Neurology Services From July 1, 2018, to June 30, 2020.

Clinical errors (dark blue) are subdivided into diagnostic, treatment or monitoring, and preventive errors.

Table 1.

Cases of Preventable Errors Occurring on General Neurology and Stroke Services Between July 1, 2018, and June 30, 2020

Discussion

This study characterized the profile of errors that occurred over 2 years on the general and stroke neurology services of a university-affiliated teaching hospital. Contrary to the notion that in-hospital errors are most frequently attributed to systems issues,¹¹ errors were most frequently attributed to clinical decision making in our series, a phenomenon rarely described in the literature.¹² Within the category of clinical errors, preventive errors were most common, for example, failure to order relevant monitoring laboratory studies after initiation of a new medication, usually resulting in no harm. However, a notable subset of clinical errors led to systems-based responses. For example, a clinical error made by a junior resident ordering the wrong dose of a medication may result in formal changes to the structure of resident supervision or creation of an EHR checkpoint. This suggests that errors attributed to clinical decision making may still be addressed via systems-based solutions.

One purpose of M&MCs is arguably to improve the collective knowledge of the audience with the goal of reducing errors related to clinical judgment. Although valuable, this may not necessarily result in long-term change due to trainee and staff turnover, which may explain the higher risk of recurrence for clinical-based errors compared with systems-based errors; although the latter may result in permanent safeguards against future similar events (e.g., via protocol change), the effects of clinical education to a large group are unpredictable and only impactful to the attendees at a particular conference.

Importantly, there were no recurrent errors within 1 year after case discussions that resulted in a protocol change (e.g., creation of a protocol delineating responsibilities of different teams on the management of patients in the postanesthesia care unit after a neurointerventional procedure). Although only a subset of errors is amenable to protocol changes, our study suggests that great effort should be made to create protocol changes as often as feasible given their effectiveness in preventing future events. Similarly, systems-based interventions resulting in more permanent change are effective, e.g., modifying the EHR system to prevent simultaneous active tube feed and hold tube feed orders.

One key step to understanding the profile of errors is their accurate capture. Our M&MC model has certain advantages and disadvantages compared with others described in the literature. Literature on M&MCs outside of surgical, intensive care, and emergency medicine specialties is relatively sparse given their origin in surgery, for which there are robust guidance on conducting M&MC such as the validated American College of Surgeons National Surgical Quality Improvement Program guidelines.^13-15 In general, QI experts advise (1) careful case selection with or without thematic grouping,^11,15,16 (2) standardized presentation formats,¹⁷ (3) action plans that reflect institutional QI initiatives^16,18 with particular focus on hospital QI metrics,¹⁹ (4) anonymity of involved personnel,²⁰ and (5) inclusion of trainees.²¹

By contrast with the above guidelines, our M&MC model requires the logging and discussion of all errors, including near-misses. This normalizes the practice of addressing errors, enables systems changes to occur even outside the context of adverse patient events, and creates a culture of constant cognizance of patient safety as recommended in the literature.²² As such, our number of individual cases discussed is substantially higher than those reported in analogous studies.^5-7 Our model includes review of every presentation by the QI director, who recommends modifications to case descriptions to reflect a Systems Audit approach²³ with the exception of acute stroke cases for which Ishikawa diagrams are used. The consistent model of senior resident as presenter preserves the anonymity of personnel involved and allows for their exposure to patient safety and QI, in congruence with the 2020 ACGME common program requirement IV.B.1.d.1, which details resident competencies related to practice-based learning and improvement.²⁴

This was a QI study at a single institution, which may limit generalizability. It is also conceivable that human error led to some senior residents neglecting to log all cases of preventable error. The review of errors has a degree of subjectivity despite discussions among multiple departmental leaders to arrive at a consensus (e.g., disagreements regarding level of harm are well described in the literature¹⁰). Unfortunately, even the definition of preventable in patient safety literature is not standardized.²⁵ As such, we advocate that the classification of errors for the purpose of improving institutional quality standards should rely on the input of multiple quality leaders.

Responses to individual errors are dependent on error type and magnitude, which must be taken into account when considering and comparing the effectiveness of different interventions. As well, recurrence within 1 year was selected arbitrarily, and rare errors may recur past this time frame. This study did not attempt to evaluate additional, unquantifiable effects of M&MCs such as promoting a culture of patient safety or serving as a forum for self-reflection and humanization.²⁶

Our study corroborated the notion that at least a subset of preventable errors attributable to clinical decision making are amenable to systems-based interventions and that when feasible, changes to written protocols are most likely to reduce risk of recurrence of a similar future error. This decade has seen an increasing focus on patient safety and care quality in neurology at the national level.²⁷ Given the important goal of optimizing care for every patient, quality leaders should conduct continuous audits of preventable errors and QI systems in their clinical areas. It is well known that medical errors are prevalent, and quality leaders should not be deterred from sharing institutional data and associated insights that can improve patient safety in the larger community.

Appendix. Authors

Study Funding

No targeted funding reported.

Disclosure

The authors report no disclosures relevant to the manuscript. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.

TAKE-HOME POINTS

• → Contrary to the common notion that in-hospital errors are most frequently attributed to systems issues, errors attributed to clinical decision making on inpatient neurology services may also be common and important to recognize.

• → A subset of preventable errors attributable to clinical decision making are amenable to systems-based interventions.

• → Changes to written protocols may be most effective in reducing the risk of recurrence of a similar future error.

• → Quality leaders should conduct continuous audits of preventable errors and quality improvement systems in their clinical areas.