A full root cause analysis (RCA) such as that required following a sentinel event is time-consuming, labor-intensive, and expensive. This quality improvement project used a similar but abbreviated process (mini-RCA and action; mini-RCA2) in response to medication errors that caused less serious harm.
In 2018, all medication errors that caused harm due to system failures but were not sentinel events were investigated by mini-RCA2 . The incidence of similar medication errors reported in the year before and in the year after the introduction of mini-RCA2 was compared to determine the impact of this intervention. Similar events were identified by searching the safety reporting system database for reported medication errors by drug name (e.g., Humate® P) and/or event type (e.g., prescribing error—omission of a patient's home medications on admission to hospital). The time and labor costs of this intervention were estimated.
Seven medication errors were investigated by mini-RCA2 . More than 48 members of staff from 11 clinical and nonclinical departments contributed to the identification of 39 system failures and made 42 recommendations, of which 22 (52%) were implemented. This reduced the recurrence of reports of similar events from 35 (0.57%) to 21 (0.36%). Although this 0.21% absolute decrease did not achieve statistical significance, recurrence of similar harm events was reduced from 7 (0.11%) to 0 (p = 0.016). Benefits were greatest when the mini-RCA2 recommendations were fully implemented. This reduced the recurrence of similar events from 9 (0.21%) to 0 (p = 0.007). A total of 251 hours (mean ± SD, 35.9 ± 16.6 hours) were required for this intervention. The associated labor cost was Saudi Arabia Riyal (SAR) 34,181 (US $8256; mean SAR ± SD, 4883 ± 1302 [mean US $ ± SD, $2102 ± $561]).
The use of mini-RCA2 to review medication errors provided a structured process to manage reported events, monitor the implementation of recommendations, and assess the effectiveness of implemented actions. The use of this rapid process to investigate errors that cause harm but are not sentinel events reduced recurrence of similar medication errors. Although the time and cost required for this intervention is not insignificant, the cumulative benefit to patients, healthcare professionals, and the organization are greater.
A drug-induced injury caused by medical intervention is an adverse drug event (ADE). Adverse drug events include allergic reactions, adverse drug reactions, overdoses, and medication errors. The National Coordinating Council for Medication Error and Prevention (NCCMERP) defines a medication error as: “Any preventable event that may cause or lead to inappropriate medication use or patient harm, while the medication is in the control of the health-care professional, patient, or consumer.” The terms medication safety incident, medication error, and adverse drug event are synonymous and are often used interchangeably.[2–6]
A sentinel event is a patient harm event that causes death, permanent harm, or severe temporary harm that required intervention to sustain life. Healthcare regulatory agencies, such as The Joint Commission on Accreditation of Healthcare Organizations, have mandated a comprehensive analysis of sentinel events since 1997.
ADEs are relatively common causes of sentinel events.[2–6] As such medication errors can cause considerable morbidity or mortality, inflate the cost of patient care, and burden healthcare systems with substantial legal and operational expenses.[2–6]
Alarmingly, the profligate trajectory of medication use is likely to dramatically increase the incidence of ADEs and the risk to patients. Yet, neither the definitions of ADEs and harm outcomes, nor the methods used to detect them, are standardized. So, the true incidence and cost of medication errors are indeterminable. Regardless, the drive to prevent medication errors is a critical patient safety initiative.
Medication errors are preventable. Determining the etiology of ADEs is critical to the formulation and implementation of successful and sustainable systems-based strategies to reduce their incidence. The most commonly used framework for the comprehensive systematic evaluation of the factors that resulted in a medical error is root cause analysis (RCA).[7,9,10] When conducting an RCA the first step involves a situation analysis to define the problem and its impact. A core team with the appropriate skill mix to address the issue must then be formed. This team must systematically collect and analyze data about the problem with the intention to identify the causes and the contributing factors. This process requires multi-disciplinary engagement to be effective.
Although complex, time-consuming, and labor-intensive, a well-conducted RCA often identifies the system failures and human factors that contributed to an error.[7,9,10] However, despite significant efforts to increase patient safety and widespread uptake of RCA, adverse event rates have remained essentially unchanged.[11,12] Thus, RCA alone is not sufficient to improve patient safety. Further initiatives are required. So, in response to problems with the RCA process, the National Patient Safety Foundation (NPSF) released a report entitled “RCA2 : Improving Root Cause Analyses and Actions to Prevent Harm.” Notably, the NPSF renamed the process “Root Cause Analysis and Action” (RCA2). This emphasized that the outcomes of the RCA (i.e., corrective actions and risk mitigation) are as important as the process itself.
Furthermore, sentinel events are just the tip of the iceberg. Many medication errors cause harm that does not achieve sentinel status. Some healthcare settings use RCA methodology to investigate errors that are not sentinel events.[9,13] For example, a hospital in Western India reported that medication errors occurred in the treatment of 117 of 300 patients between June 2012 and April 2013. Many important observations were made after RCA was performed for each of these errors. For example, the RCA of the prescribing errors, 62 (53%) in this cohort, found that these were due to not prescribing the dosage form (33, 53%), illegible handwriting (21, 34%), and prescription of the wrong brand name (8, 13%).
Institutions that perform more RCA will obviously develop more recommendations to improve patient safety. However, the data suggesting that this approach reduces harm are limited to observational studies. So, healthcare regulators do not mandate detailed investigation and analysis of these less harmful errors. As a result, nonsentinel events are rarely investigated to the same extent. This wastes precious opportunities to expose the flaws in the system uncovered by ADEs that do not appreciably affect patient outcomes.
A full, formal RCA is demanding, time-consuming, labor-intensive, and expensive. Performing a full “formal” RCA for all harm events is not feasible. However, the use of a similar but abbreviated process (eg, mini-RCA2) for errors that cause less serious harm may improve patient safety.
The aim of the study was (1) to assess the feasibility of performing mini-RCA2 for all medication errors due to system failures that caused harm but were not sentinel events and (2) to assess whether this intervention may contribute or not to reduce patient harm.
This retrospective, observational study was approved by the institutional review board of the King Abdullah International Medical Research Center, Riyadh, Saudi Arabia. No patient identifiable data from individual patients are included in this paper.
The setting for this quality improvement project is a 1600-bed academic tertiary care center that promotes a “just culture.” This emphasizes that mistakes are usually due to suboptimal systems and links discipline to behavior, not harm. Thus, in the event of a serious untoward incident, the directly involved staff are not blamed. Instead, the stakeholders focus on “what went wrong?” rather than “who did this?” This model allows individual accountability whilst promoting a learning culture. Thus, a just culture creates an environment where individuals feel comfortable reporting errors. This helps the organization learn from its mistakes.
All reported medication errors are reviewed by Medication Safety Program (MSP) staff (one medication safety officer and two clinical auditors) within 24 hours of reporting. The medication safety officer was previously a pharmacist and the two clinical auditors were previously pharmacy technicians. However, they are now full-time staff in the department of quality and patient safety and have no other roles or responsibilities.
The NCCMERP classification is used to classify the severity of the reported ADEs. The MSP officer validates and initiates intervention for all ADEs that caused harm. Sentinel events are escalated to the director of the Department of Quality and Patient Safety who initiates and leads the immediate response. Other harm events were investigated at the discretion of the MSP officer.
In January 2018, the MSP at King Abdulaziz Medical City (KAMC) started a quality improvement project designed to fully investigate all medication errors due to system failures that caused harm but were not sentinel events. All ADEs were assessed to determine whether any latent system failures contributed to the incident and whether hospital policies on medication use were followed. If policies were not followed, no latent failures were identified, and the incident was considered to be due to human error alone by the medication safety officer and the staff involved; further independent comprehensive systematic assessment was considered unnecessary. In accordance with established hospital policies, these events were addressed at the departmental level. Therefore, ADEs caused by human error alone were excluded from this initiative.
Mini-RCA2 methodology was used by the MSP to investigate nonsentinel events due to system failures as soon as possible after the event. This initiative complemented the preexisting protocolized pathway for investigation of sentinel events. The methodology of full RCA2 has been described previously.[8,10] A mini-RCA2 is essentially an informal, abbreviated RCA2 . The differences between a full RCA2 and a mini-RCA2 conducted in our organization are described in Table 1. Similar to RCA2 , the mini-RCA2 process involves identifying the root causes of the ADE. It integrates system thinking and involves all relevant stakeholders to identify potential improvements that could decrease the risk of recurrence of the ADE.
After the initial analysis, a preliminary report is prepared by using a standardized template (Fig. 1). This initial report includes all of the root causes that were identified during the mini-RCA2 and is shared with all stakeholders. It also recommends interventions to prevent the recurrence of the incident (ie, an action plan). Within 72 hours of the initial analysis the medication safety officer prepares the final report of the mini-RCA2 in collaboration with the MSP Chairman. The final report may include other recommendations deemed appropriate. Each report specifies a time frame for implementation of each recommendation. Specific departments are made accountable for either implementing or rebutting each recommendation.
The medication safety officer sends the final mini-RCA2 report to all departments involved in the incident and uploads the report to a database to facilitate follow-up (Access, Microsoft). The medication safety officer contacts each department to ensure that the recommendations have been actioned as specified in the final report.
If recommendations have not been implemented the medication safety officer informs the chairman of the MSP. At our institution the chairman of the MSP is currently also the director of the Department of Quality and Patient Safety. In this latter capacity he expedites implementation of the recommendations by contacting the relevant stakeholders, resolving any conflicting opinions, and informing the higher administration. The effectiveness of the action plan is assessed by monitoring the recurrence of similar ADEs through the ongoing assessment of reported medication errors.
Study of the Intervention
The database of mini-RCA2 reports was reviewed to determine the number of medication errors that were investigated with this methodology between January 1, 2018, and December 31, 2018. The number, specialty, and grade of healthcare professionals involved in each mini-RCA2 were determined. The total number of recommendations made and the number that had been implemented were determined.
The numbers of similar ADEs reported 12 months before and after each mini-RCA2 were compared. All reported medication errors are classified by specific event type in the Safety Reporting System database. Similar ADEs events were identified by searching the Safety Reporting System database for reported medication errors by drug name (eg, Humate P) and/or event type (eg, prescribing error—failure to order a patient's home medications on admission to hospital). These data were then aggregated to provide summary statistics.
To determine the time required for this intervention the authors created a framework to identify the tasks performed during a mini-RCA2 . These tasks included reviewing the incident report, conducting the investigation, the mini-RCA2 meeting, the preparation of the report, reviewing the report, communicating actions, and follow-up. Most of the work was performed by the medication safety officer and the MSP Chairman who were involved with most of these tasks. They were asked to recall the time they spent on each mini-RCA2 . The other participants only reviewed the incident report, attended the mini-RCA2 meeting, and reviewed the final report. The duration of the single mini-RCA2 meeting for each incident was strictly limited to 1 hour. The time that the other participants spent reviewing the incident report and the final mini-RCA2 report was extrapolated from the time the MSP Chairman spent on these tasks. All times were rounded up to 15-minute intervals.
An economic analysis was then performed. To estimate the labor cost, the time spent by each member of staff involved was multiplied by their average hourly salary (based on the midpoint of the salary scales).
Data are presented as frequency and percentage and were compared by using Fisher exact test. All statistical analyses were performed with Excel 2016 (Microsoft). The time required for each mini-RCA2 is presented in hours and the associated financial cost is presented in Saudi Arabia Riyals (SAR) and US dollars ($). The exchange rate is pegged at 3.75 SAR to $1.
Harm Events Investigated Using the Mini-RCA2 Methodology
Of the 2927 medication errors reported in 2018, 13 errors (0.44%) caused harm but were not sentinel events. Six ADEs were due to human error, while seven (54%) resulted from system failures and were investigated by using mini-RCA2 methodology (Table 2). For contextualization, in 2018, a total of 2912 events did not cause harm, one sentinel event was investigated by higher administrative personnel using the full RCA methodology, and another sentinel event was investigated by the medication safety officer, under the supervision of higher administrative personnel, using the full RCA methodology.
The patient affected by the error involving Humate P was critically ill and was receiving treatment in the pediatric intensive care unit at the time of the incident. The clinician reporting the incident suggested that the ADE had caused harm. So, this incident was included in the initiative, and mini-RCA2 was performed. Although the patient subsequently died, during the course of the mini-RCA2 it became clear that the error did not harm the patient. Whilst ultimately the medication error was not thought to have harmed the patient, this was not clear at outset.
Engagement of Healthcare Professionals in Mini-RCA and Root Causes Identified
The mini-RCA2 for seven ADEs involved 48 healthcare professionals (median, 6; range, 3–13) from 11 clinical and nonclinical departments (Table 3). This initiative identified a total of 39 system failures that were the root causes of the harm events investigated (median, 6; range, 2–10).
Recommendations Made Following the Mini-RCA2
The recommendations made after each mini-RCA2 and their stage of implementation on January 7, 2020, are shown in Tables 3–5. Several recommendations were made after each mini-RCA2 (median, 6; range, 2–10). Of the 42 recommendations made following the mini-RCA2 , 22 (52%) had been implemented by January 7, 2020. The time scale within which each recommendation should have been implemented was defined by the team conducting each mini-RCA2 . Of 22 recommendations (52%) that should have been implemented within 1–3 months (short term), 16 (73%) had been. Of 12 recommendations (29%) that should have been implemented within 3 to 6 months (intermediate term), four (33%) had been. Of eight recommendations (19%) that should have been implemented within 6 to 12 months (long term), two (25%) had been. After extensive discussion, two of the recommendations (5%) were ultimately rejected by the stake holders and the higher administration. Thus, on January 7, 2020, a median of three recommendations had been implemented after each mini-RCA2 (range, 0–6) and 18 recommendations were still pending.
Major changes to systems, policies, guidelines, work flow, awareness, and training were made. However, despite regular follow-up with the departments responsible for each recommendation, many important medication safety initiatives are outstanding. Recommendations were fully implemented for only two events (prescribing error—omission of a patient's home medications on admission to hospital and Humate P dose incorrect; see Tables 3 and 4). The implementation of some recommendations was delayed because they were incorporated into larger ongoing quality improvement projects. For example, the recommendations made after the omission of the midazolam infusion have been incorporated into a large multi-disciplinary initiative to enhance medication safety within the operating theaters.
Analysis Used to Determine the Impact of Mini-RCA2 on the Recurrence of Medication Errors
The data from the SRS database on the number of medication errors that occurred 1 year before and 1 year after each index event is summarized in Table 4. The first event occurred on January 25, 2018, and the last event occurred on December 31, 2018. To determine the total number of medication errors reported during the periods under consideration the SRS database was interrogated for events from January 25, 2017, until December 30, 2018 (“before”) and January 26, 2018, until December 31, 2019 (“after”). This was used as the denominator for the aggregate data below.
Impact of Mini-RCA2 on the Recurrence of Medication Errors
In the years before each index event a total of 6124 incidents were reported. Including the seven index events, 35 similar events (0.57%) occurred. In the years after each index event, a total of 5885 incidents were reported. Twenty-one similar events (0.36%) occurred. The reduction in the recurrence of similar reported events (14, 0.24%; p = 0.11) did not achieve statistical significance. However, consideration of the impact of this initiative on incidents that caused harm is of great interest.
Impact of Mini-RCA2 on the Recurrence of Harm Events
In the years before each index, seven of the 35 similar incidents (0.11%) caused harm but were not sentinel events. These seven incidents were the harm events that were included in the present quality improvement initiative (ie, mini-RCA2 was performed). However, none of the 21 similar events reported in the years after each index event caused harm. The statistically significant absolute decrease in recurrence of similar harm events (7, 0.11%, p = 0.016) represented a relative risk reduction of 1 (Table 4).
Effect of Implementation of the Mini-RCA2 Recommendations on Event Recurrence
When the recommendations from the mini-RCA2 had been fully implemented, the benefits were greater. In the years before the two index events for which all mini-RCA2 recommendations were implemented, a total of 4344 ADE reports were submitted. Including the two index events, 11 similar events (0.25%) occurred. In the years after each index event, a total of 3323 incidents were reported. No similar events were reported. This represents a statistically significant absolute decrease (11, 0.25%) in recurrence of similar events (p = 0.007), with relative risk reduction of 1 (Tables 4 and 5).
Resource Utilization and Economic Analysis
Table 6 displays the time and the labor costs of each mini-RCA2 . A total of 251 hours (mean ± SD, 35.9 ± 16.6 hours) was required for this intervention. Most of the time was spent on the investigations by the medication safety officer (99 hours, 39.4%). Fifty hours of other healthcare professionals' time was required for the mini-RCA2 meetings. The total cost of the labor required for this quality improvement project was SAR 34,181 (US $8256; mean SAR ± SD, 4883 ± 1302 [US $ ± SD, $2102 ± $561]). However, the cost varied considerably. The mini-RCA2 for the ADE involving Humate P was complex, took the most time (58.75 hours; 23.4%), and was the most expensive (SAR 7681, US $2048; 22.5%). The mini-RCA2 for the ADE involving levetiracetam required the least work (13 hours; 5.2%). However, the labor costs were lowest for the mini-RCA2 for the ADE involving the midazolam infusion (SAR 1895, US $505; 5.5%).
Feasibility and Sustainability
Mini-RCA2s were conducted by the medication safety officer for all ADEs that caused harm but were not sentinel events in 2018. All preliminary reports were submitted within 72 hours of the mini-RCA2 . However, the final reports were submitted between 6 and 84 days after the incident (mean, 47.5 days). Delays were usually related to requests for minor modifications to the report from the MSP Chairman and individual stakeholders.
It would not be feasible for this number of medical errors to be investigated by higher administrative personnel using the full RCA methodology. However, unless the incidence of ADEs that cause harm increases significantly, investigation with mini-RCA2 led by the medication safety officer is feasible and sustainable.
A structured mini-RCA2 response to all ADEs that caused harm (ie, not just those that resulted in sentinel events) significantly reduced the recurrence of similar reported events. The benefit was greatest if all of the recommendations of the mini-RCA2 were implemented. Although the costs associated with conducting a mini-RCA2 were considerable, the impact on patient safety far outweighed this.
Although medication errors are common, harm is rare. In 2018, 2927 ADEs were reported at KAMC, but only 15 (0.51%) caused harm. Two of these ADEs were sentinel events. Six errors that caused harm but were not sentinel events were due to human error alone. So, seven ADEs due to system failures were investigated with the mini-RCA2 methodology. The identification of 39 system failures by this initiative generated a huge amount of knowledge about suboptimal systems that would otherwise have been lost.
Identification of these system failures prompted several interventions to improve the safety of our organization. While some similar events recurred despite mini-RCA2 , this intervention significantly reduced the incidence of recurrence of similar events and eliminated harm from similar events. Simply going through the motions of RCA and developing an action plan without implementing the recommendations will have no benefit. In the present study, the benefit to patient safety was greatest if all of the mini-RCA2's recommendations were implemented. In some cases, the recurrence of similar errors was eliminated completely after the implementation of the recommendations of the mini-RCA2 . Furthermore, it is possible that this initiative may have reduced the recurrence of other events. This observation reinforces the need for follow-up to ensure that the recommendations made following RCA2 are fully implemented.
Furthermore, many system failures occurred (and recurred) because of suboptimal operating procedures or lack of awareness of established/updated policies. Thus, when an ADE occurs or policies and operating procedures are developed or updated, it is important to educate the relevant stakeholders about the changes to practice required. Our institution uses a multi-modal approach to rapidly disseminate this information. This includes email, intranet banners, computer screen savers, wall-mounted screens, presentation at departmental meetings, and awareness campaigns.
When considering the recurrence of events following RCA2 , it is important to reflect on the quality of the recommendations made. In terms of the hierarchy of effectiveness, force functions and automation or computerization of processes are most likely to prevent errors. Guidelines, policies, education, and reminders are less likely to successfully and sustainably improve patient safety.
Previous studies have questioned the efficacy of RCA.[10–12] Indeed, the NPSF attempted to address this by developing RCA2. Our study reiterates that merely identifying causes of errors does not reduce their recurrence or prevent harm to patients. However, we demonstrate that when appropriate recommendations are made and implemented with the mini-RCA2 methodology, recurrence of similar errors can be reduced and harm may be prevented.
Substantial resources were required for this intervention. It consumed 250 hours of staff time at an estimated cost of more than SAR 34,181 (US $8256). For comparison, in an academic center in the United States the average cost of using RCA to investigate a serious reportable event due to an ADE was US $10,930 in 2013. While the calculation of cost differed from that used in the present study, this illustrates the potential savings in terms of finances and staff time that may be achieved by using the mini-RCA2 methodology rather than a full RCA.
Medication errors cause great harm to patients and burden healthcare systems with considerable legal and operational expenses.[2–6] So, whilst the absolute impact may seem small, the overall benefit of this intervention to patients, healthcare professions, and the hospital is of great clinical and operational importance.
Furthermore, the insights gleaned through RCA can occasionally provide significant benefits above and beyond improvements in patient safety. For example, one mini-RCA2 found that the medication error involving Humate P did not harm the affected patient. However, this was not clear at the outset. This finding was extremely comforting to the staff involved in the incident and the patient's family.
Thus, while the full benefit of this intervention is unquantifiable, the present study demonstrates the feasibility and impact of using mini-RCA2 methodology to investigate medication errors that cause harm. Our observations are similar to those from other organizations that use RCA methodology to investigate all medication errors that cause harm.[10,13] However, no previous studies have described the impact of using an abbreviated RCA2 strategy to investigate ADEs.
The present study describes the investigation of a relatively small number of ADEs that caused harm in a single center. The culture at KAMC encourages and rewards the reporting of errors, and nearly 3000 ADEs were reported in 2018 alone. However, not all medication errors are reported. Furthermore, each mini-RCA2 was led by an experienced medication safety officer trained to perform RCA. This ensured internal validity. However, this in turn limits external validity and generalizability to centers that do not have medication safety officers or personnel besides higher administrative staff who can perform RCA. The economic analysis did not include the nonlabor direct costs (ie, the costs of the infrastructure required to operate the MSP), hospital overheads, or the time taken to implement the recommendations of the mini-RCA2 . So, the total costs associated with this intervention are underestimated. However, most of the cost is associated with staff time, so the true cost is unlikely to be significantly different from the data provided above.
Using mini-RCA2 to investigate medication errors that cause harm but are not considered sentinel events can reduce ADEs. This may improve patient safety and reduce the operational costs of healthcare systems. These benefits far outweigh the cost of implementing this intervention. Beyond the reduction in operational hazards, some mini-RCA2s may ultimately conclude that the patient was not actually harmed by the ADE. In this situation, the relief for all stakeholders is of incalculable value above and beyond that of the systems-level improvements made by the organization.
Source of Support: None. Conflict of Interest: None.
Ghada Hussain Al Mardawi and Rajkumar Rajendram contributed equally to this study.