Abstract
This study discusses the summary, investigation and root causes of the top four sentinel events (SEs) in Saudi Arabia (SA) that occurred between January 2016 and December 2019, as reported by the Ministry of Health (MOH) and private hospitals through the MOH SE reporting system (SERS). It is intended for use by legislators, health-care facilities and the public to shed light on areas that still need improvement to preserve patient safety.
The purpose of this study is to review the most common SEs reported by the MOH and private hospitals between the years 2016 and 2019 to assess the patterns and identify risk areas and the common root causes of these events in order to promote country-wide learning and support services that can improve patient safety.
In this retrospective descriptive study, the data were retrieved from the SERS, which routinely collects records from both MOH and private hospitals in SA. SEs were analyzed by type of event, location, time, patient demographics, outcome and root causes.
There were 727 SEs during this period, 38.4% of which were under the category of unexpected patient death, 19.4% under maternal death, 11.7% under unexpected loss of limb or function and 9.9% under retained instruments or sponge. Common root causes were related to policies and procedures, guidelines, miscommunication between health-care facilities, shortage of staff and lack of competencies.
Given these results, efforts should focus on improving the care of deteriorating patients in general wards, ICU (Intensive Care Units) admission/discharge criteria and maternal, child and surgical safety. The results also highlighted the problem of underreporting of SEs, which needs to be addressed and improved. Linking data sources such as claims and patient complaints databases and electronic medical records to the national reporting system must also be considered to ensure an optimal estimation of the number of events.
Introduction
A sentinel event (SE) in a health-care organization is defined as ‘any event leading to serious patient harm or death and is caused by healthcare (human error/behavior and/or system failure), rather than the patient’s underlying illness’ [1]. When analyzed and shared within the organization where it occurred, such a triggering event can result in system improvement [2] and influence behavioral change among health-care workers, as described under the term ‘the sentinel event effect’ [3]. Although each SE is unique, its root cause can be commonly encountered in all organizations that have similar systems, policies and procedures, conditions of operation and organizational culture [4].
Sharing knowledge about SEs can minimize the risk of recurrence and promote system improvement not only in the organization where the event occurred, but also in other organizations that might go through the same experience [5]. Indeed, many organizations worldwide conduct research that aims to develop best practices to improve patient safety by reviewing and analyzing the most common reported SEs and disseminating lessons learned to reduce the likelihood of their occurrence in the future [6]. In Saudi Arabia (SA), the Ministry of Health (MOH) oversees the SE reporting system (SERS) and supports the notification, investigation and review of such occurrences across the health system. The MOH SERS was established in late 2010 and requires mandatory notification of all SEs that occur in MOH and private hospitals within 48 h of discovery [7]. The system consists of a secure, web-based portal that allows hospitals to submit information about the type of SE, time, location, contributing factors, subsequent patient outcome and impact on patient care. It also includes a root cause analysis (RCA) that follows a well-structured framework provided by the MOH, which focuses on the causative systems, processes, human factors and planned corrective actions.
The majority of SEs reported are classified into 21 categories as directed in the MOH Risk Management Policies Manual. The RCA information on causative systems, processes and human factors [2] must be completed and submitted to the MOH SERS within 30 working days upon becoming aware of the event. The recommendations that stem from every RCA are meant to be implemented and disseminated within the involved facility to protect patients, staff members and visitors from such unfortunate events. To fulfill this mission, it is also important to distill the information provided by the SERS and track the most frequently encountered SE categories at a national level over time to identify significant trends under each category. It is also important to identify meaningful information about the factors that most frequently contribute to SEs to guide system improvement initiatives nationwide.
Objectives
The purpose of this study is to review the most common SEs reported by the MOH and private hospitals between the years 2016 and 2019 to assess the patterns and identify risk areas and the common root causes of these events in order to promote country-wide learning and support services that can improve patient safety.
Methods
Study design
This is a retrospective, noninterventional study that uses a quantitative approach to review all SE records reported by health-care organizations in SA through the MOH SERS, to assess potential preventability of the most common events. The analysis included data on RCAs reported by the hospitals for each SE.
A three-step protocol for the analysis of SE records was used. First, a review was undertaken of all events reported in order to validate their inclusion and classification by category. Then, an analysis was undertaken of the overall SE characteristics by region, time, setting and outcomes. The final step was a more in-depth review of the top four SEs. This step involved analysis of the narratives for the key root causes for each category so that the cause of harm could be established and the preventability discussed.
Setting
A multidisciplinary research team established checklists and reviewed the records of all reported SEs. All validated SE records were categorized and analyzed for variables, such as event type; location, time, date and day of the week; patient demographics; primary diagnosis; comorbidities; type of intervention and event outcome. A cross-check of the possible root causes was conducted by two researchers based on a predetermined checklist for common root causes: ineffective communication, shortage of staff, shortage of equipment and supplies, lack of staff competency, policies and procedures, protocol noncompliance and deficiency factors. When there was a discrepancy in identifying possible root causes, the two reviewers would retrieve the records to cross-check and then send them for a third-tier review by an expert in RCA and contributory factor classification.
To validate the research strength, limitations and accuracy of the conclusions, this study follows the guidelines for reporting studies on routinely collected health data (RECORD) [8].
The sample
Study size
The study population included patients of all ages who had received inpatient medical or surgical care at any MOH or private health-care organization in SA, including specialty hospitals, psychiatric hospitals and specialized health centers, between January 2016 and December 2019. Health care in SA is provided by both the public and private sectors. The public sector provides health services through the MOH and other government agencies, including referral hospitals such as King Faisal Specialist Hospital and Research Centre, and military, security and university hospitals [9]. The total number of hospitals in SA in 2019 was 498, with the MOH managing 57% of them. They are of varying size and located in 13 geographic regions. Of the remaining facilities, 164 hospitals are private health-care facilities and the remaining services are provided by other government sectors [10]. According to MOH statistics, more than 11 million hospital admissions nationwide were recorded during the study period, as illustrated in Table 1 [10].
Hospital admissions from 2016 to 2019
| Total number of inpatients | MOH | Private | |
|---|---|---|---|
| 2019 | 3 059 162 | 1 423 805 | 1 635 357 |
| 2018 | 3 070 871 | 1 410 051 | 1 660 820 |
| 2017 | 2 712 648 | 1 377 215 | 1 335 433 |
| 2016 | 2 889 205 | 1 648 675 | 1 240 530 |
| Total | 11 731 886 |
| Total number of inpatients | MOH | Private | |
|---|---|---|---|
| 2019 | 3 059 162 | 1 423 805 | 1 635 357 |
| 2018 | 3 070 871 | 1 410 051 | 1 660 820 |
| 2017 | 2 712 648 | 1 377 215 | 1 335 433 |
| 2016 | 2 889 205 | 1 648 675 | 1 240 530 |
| Total | 11 731 886 |
Hospital admissions from 2016 to 2019
| Total number of inpatients | MOH | Private | |
|---|---|---|---|
| 2019 | 3 059 162 | 1 423 805 | 1 635 357 |
| 2018 | 3 070 871 | 1 410 051 | 1 660 820 |
| 2017 | 2 712 648 | 1 377 215 | 1 335 433 |
| 2016 | 2 889 205 | 1 648 675 | 1 240 530 |
| Total | 11 731 886 |
| Total number of inpatients | MOH | Private | |
|---|---|---|---|
| 2019 | 3 059 162 | 1 423 805 | 1 635 357 |
| 2018 | 3 070 871 | 1 410 051 | 1 660 820 |
| 2017 | 2 712 648 | 1 377 215 | 1 335 433 |
| 2016 | 2 889 205 | 1 648 675 | 1 240 530 |
| Total | 11 731 886 |
Due to insufficient information, it was not possible to use codes or algorithms to validate or classify the whole study population by patient characteristics or reason for admission. Only the SE database population was studied for validation and analysis.
Variables
All reported SEs were analyzed by means of a checklist that was adapted from the RCA report template and the contributory factors grid. Contributing factors were identified by answering questions related to patient, staff, team factors, task and technology, work environment, organizational and management factors and institutional context factors. The ‘London Protocol’ for system analysis of clinical incidents has validated these variables as being contributory factors that influence clinical practice [11].
Statistical methods
Data were first summarized using frequency distributions and cross-tabulations. Statistical analysis was performed using SPSS, version 20. Numbers and percentages were used in the descriptive statistics.
Data sources
The data were collected entirely from the MOH SERS. No reporting from patients themselves and their families or any other form of automated surveillance of clinical patient records or complaint data for detecting SEs could be included in this study, as not only are these methods not yet routinely used by all health-care organizations, but there is also lack of linkage between the existing databases. All the SEs included fulfilled the definition of the Saudi Taxonomy for Patient Safety for SE.
Bias
The study design limited the researchers’ understanding of how the reported causes induced SEs. As this study relied exclusively on the SERS as the source of data, the researchers did not observe any of the events first-hand or conduct any of the RCAs. Also, some reports were randomly missing data for certain variables, which may not only have resulted in biased interpretations, but also have overlooked the reasons why such data were not reported. Cases with missing information were not included in the sample, which may have weakened the power of the results.
Data access and cleaning methods
The data were extrapolated from a restricted access database and de-identified. All records were verified for accuracy and completeness by the research team to ensure the data collected were consistent with the definitions and quality criteria established at the outset of the study.
Linkage
No linking with other data sources, such as claims databases or patient files, was performed.
Results
Participants
The study population consisted of 727 patients who had been reported as being exposed to an SE while admitted for care to hospital facilities. The reported SEs affected 158 patients in 2016, 153 in 2017, 188 in 2018 and 228 in 2019.
Descriptive data of the reported SEs
The highest number of SEs notified occurred in MOH hospitals, with 87% of the SEs being reported from MOH hospitals and 13% of SEs being reported from private hospitals (Figure 1). More than half of the SEs were reported from general hospitals, followed by maternity and children hospitals then medical cities, as shown in Figure 2. Consistently for all 4 years, the day on which SEs were most recorded was Thursday, with Wednesday recording the second highest, followed by Tuesday with the third highest.
Percentage of reported SEs by the health-care sector.
Number of SEs by type of health-care facility.
Key results for the main SE categories reported
The four main SE categories reported were unexpected death, maternal death, unexpected loss of a limb or a function and retained instrument or sponge (Figure 3).
Number of SEs by type.
Unexpected death
The number of unexpected deaths reported per year increased from 2016 to 2019, with the majority occurring in the MOH hospitals. Most of the cases were in the 40- to 64-year age group, followed by the under 1-year-old age group. Most of the unexpected deaths occurred in the general wards, followed by the adult critical care wards and emergency departments. The leading causes of death were bleeding/Gastrointestinal (GI) bleeding, hypoxemia, cardiac arrest and respiratory disorder. There were five main root causes behind unexpected death. First was ineffective communication, both between hospitals (interfaculty) and between staff, followed by shortage of staff, especially ICU consultants, then shortage of equipment and supplies, such as medications, Computed Tomography (CT), Doppler machines and monitoring machines. The fourth main root cause was lack of staff competency in monitoring critical patients and use of equipment, and the last was noncompliance with policies and procedures/protocols, such as in implementing high alert medication safety measures, airway management, compliance with telephone and verbal order policies and proper documentation.
Maternal death
Maternal deaths decreased from 2016 to 2019, with 89% of the cases being reported by MOH hospitals. Most of the maternal deaths occurred in the ICUs, followed by Operating Rooms (ORs), with 72% of the maternal deaths involving delivery by cesarean section. Most of these patients were in the 31- to 40-year age group. The top causes of maternal death were postpartum hemorrhage and pulmonary embolism (PE). Reported comorbidities were gestational diabetes (21%), preeclampsia/eclampsia (21%), sickle cell disease (17%), Hemolysis, Elevated Liver Enzymes and Low Platelet count (HELLP) syndrome (8%) and obesity (8%). The top root causes were the nonavailability of and nonadherence to clinical guidelines for venous thromboembolism and anticoagulant use, management of high-risk pregnancy and placenta previa, staff shortages, lack of staff competency and ineffective communication between staff.
Unexpected loss of limb or a function
The majority of unexpected losses of limb or function were reported by MOH hospitals, and most of the cases occurred following procedural/treatment interventions such as cesarean sections, intravenous medication administration, labor management and central line insertion. Around 40% of the cases were in the under 1-year-old age group. The impact of unexpected loss of limb or function was either the loss of function in 60% of cases or loss of organ in 40% of cases. The root causes were deficiencies in patient assessment, ineffective communication between staff, lack of staff competency in managing critical pediatric patients and cardiotocography readings, noncompliance with policy and procedure for central line insertion, physicians’ privilege standards, lack of proper documentation and lack of policy regarding the management of high-risk pregnancies.
Retained instrument or sponge
Finally, around half of the incidents of retained surgical instruments or sponges were retention of gauze/swab/pad/towel, and the majority of events occurred during laparotomies, followed by normal vaginal delivery and orthopedic procedures. The majority of cases resulted in loss of function. The root causes were safety culture factors such as staff not speaking up when they suspect missing instruments and lack of cooperation between surgeons and nurses, in addition to equipment quality issues, shortage of staff, lack of adherence to the policy for surgical counting, and timely updates on patient condition and the results of x-ray reports by radiologists.
Discussion
Principal findings
According to the results above, there was an increase in the number of reported SEs in 2019 compared to previous years. This increase may not indicate an increase in the number of SEs that occurred as much as it may reflect a relatively better adherence to the policy of mandatory reporting of such events through the SERS. The incidence by type of SE in SA changed from unexpected death (37.4%), maternal death (20.8%), unexpected loss of limb or function (14.8%), major medication errors (7.4%) and retained surgical instruments or sponge (7.4%) for the period 2012–15 [12] to unexpected death (38.4%), maternal death (19.4%), unexpected loss of limb or function (11.7%) and retained surgical instruments or sponge (9.9%) for 2016–19. This change reflects encouraging results for the medication safety programs and strategies implemented by the MOH during recent years, such as the Saudi Food and Drug Authority launching the adverse event reporting system and the introduction of tools that can reduce medication errors, such as electronic prescribing and computerized physician order entry with clinical decision support systems, barcodes to track medications and other interventions.
The data also suggest that safer patient care is more likely to be delivered during the weekend (Friday and Saturday), which contradicts the findings of studies that reported a less effective response to adverse events on weekends [13, 14]. The study showed that PE was the leading cause of unexpected and maternal deaths, which reinforces the findings of other authors [15]. Most of the unexpected deaths appear to occur in general wards, which implies that the clinical deterioration and unexpected life-threatening events that occur over the course of a patient’s stay in general wards could be predicted and prevented by closer monitoring and optimal care. This situation was conclusively shown in a recent study conducted in SA that investigated the factors that influence the outcome in terms of mortality rate of patients who received critical care response team (CCRT) service in the general ward [16]. According to the authors, there is a need for increased vigilance of older patients, those who spend more time in critical care units and those with low oxygen saturation. Also, the topics of CCRT implementation, effectiveness and best practice standards in the Saudi health-care facilities are some of the areas that need in-depth analysis, as these issues are still highly institution-dependent, and to the best of the authors’ knowledge, research activities in these areas are still lagging behind the international level.
In addition, the complications of pregnancy and patient condition postdelivery, such as postpartum hemorrhage, disseminated intravascular coagulation and HELLP syndrome, would indicate that these patients had not received adequate prenatal care, most probably due to lack of awareness. Indeed, a previous study that focused on preconception and prenatal health behavior among Saudi pregnant women confirmed the need for preventive health-care interventions that target young women and those with a low educational level [17].
The problem of retained surgical items after surgery is reported mainly for sponges, gauze and towels retained in the abdomen and pelvis following laparotomies and normal vaginal deliveries. This result is also recorded by other studies and suggests an increasing risk for retained surgical bodies associated with these body cavities, especially when an unexpected change in procedure is encountered [18, 19]. Al Wahabi et al. [12] identified that the three main causes of this problem for the 2012–15 period were unavailability of policies, procedures and guidelines, failure to implement them and lack of proper communication. These causes were also found in the current study, indicating that these contributing factors are deeply rooted in the health-care system despite the ongoing efforts by the MOH to improve them. This study also revealed that staff shortages, lack of competencies, issues related to equipment and supplies and lack of a safety culture were among the main root causes. Likewise, nonpunitive response to error, staffing and communication openness are areas for potential improvement according to the latest Saudi Patient Safety Center (SPSC) recommendations that are based on the results of hospital surveys on patient safety culture undertaken by the Patient Safety Culture National Committee [20]. If considered as interventional priorities, the initiatives recommended by the SPSC, such as implementing the Just Culture Algorithm in health-care facilities and encouraging leaders to build a culture of safety, will probably promote a learning environment with tangible positive effects on both reporting rates and the prevention of SEs.
Strengths and limitations
Although hospitals are mandated to report all SEs, it is unknown to what degree the reported cases accurately reflect the number of occurrences and the related root causes. The analysis undertaken in this study was restricted to the MOH SERS, which could have resulted in underestimation of the national rate of SEs and related causes. To identify some of the nonreported SEs, it would be necessary to institute a method for systematically investigating events that are not collected by the reporting system. Identifying other appropriate sources of data to address the research question, such as claim databases, patient complaints databases and even electronic medical records, should be considered for future studies, especially with the increased availability of emergent strategies such as pharmacovigilance, which can better exploit the more accessible data sources [21].
However, a consensus among experts and regulatory agencies on the most appropriate methodological steps and validation processes needs to be established to avoid interpretation errors. In addition, although near miss events that may lead to SEs are not required to be reported through the SERS, including such events among those that require a structured RCA would allow organizations to prevent them from happening in the future with serious consequences or fatality. However, with event reporting still not being anonymous, this area is challenging and requires strong programs and initiatives [22]. Therefore, to increase the generalizability of the results, there is a need to encourage reporting and to integrate the results with other sources of data.
Interpretation within the context of the wider literature
The results from this study provide a more reliable benchmark of the national inpatient SE rate that policymakers, regulators, accrediting organizations, health-care providers and researchers can use when making decisions regarding the allocation of resources and implementation of specific requirements to prevent patient harm. Although the rate of occurrence of SEs in SA is significantly low compared to the rates reported in the literature of other countries, the data on reported events in SA should be considered in light of the abovementioned limitations. Furthermore, according to a literature review of hospital adverse event studies conducted in 27 countries between 1991 and 2017, the estimated rates of reported events seem to be lower than the actual occurrences, ranging from 2.9% to 21.9% of admitted patients, of which 7.3% resulted in permanent disability and 7.3% in patient death, despite the mandatory reporting of events in these countries [23]. This points toward the problem of underreporting of events and near misses in the health-care industry. These authors believe that the eventual discovery of such events is mostly driven by patients’ and families’ complaints rather than voluntary reporting. Unfortunately, failing to report adverse events and especially near misses contributes to the likelihood of the occurrence of SEs. These adverse events and near misses are indeed an opportunity for learning, introducing safety activities and developing preventive actions and tools [24].
Underreporting has always been linked to the safety culture of the organization, due in the most part to fear of blame, guilt, punishment, monetary penalties or acquiring a bad reputation [25]. In addition, authors showed that lack of awareness about the reporting policy and reporting systems in SA may not encourage professionals to report incidents, as they usually involve a higher chain of command during the submission and interview processes, in addition to the complicated process of reporting events [26].
It is important to note that the MOH and the Saudi Central Board for Accreditation of Healthcare Institutes (CBAHI) for accredited hospitals require that all health-care organizations in the Kingdom of Saudi Arabia (KSA) have reporting systems and internal policies that report and disseminate information on all SEs. Well-structured RCAs should be submitted to both the MOH and CBAHI along with action plans to avoid such risks in the future. Despite the clarity of the reporting procedures, the standardization of the framework and process for conducting RCAs and the continuous education and training conducted by the MOH in this area, the low rates of reported SEs revealed in this study indicate that hospitals need more guidance throughout this process. They should be provided with standards for the selection of recommendations and assistance with their implementation by experts, as recommended by Bos et al. [27]. In addition, initiatives or clear legislation that addresses protection of the institution or reporter of the incident are needed to encourage reporting and learning from near misses and adverse occurrences. Ideally, future development of independent reporting systems that are not controlled by any organization with the power to punish reporters could be a good direction shift.
Of note, even though all health-care organizations are required to report SEs, in the current study, there was a higher reporting rate of SEs in governmental facilities than in the private sector. However, according to the literature, the incidence of SEs is not confined to governmental health-care systems [28], and in terms of quality, neither private nor public hospitals have been proven to be superior [29]. Therefore, before concluding that there is still work to be done to improve the transparency of private organizations in reporting such events, the results of the current study point to the need for further studies that compare the quality of care and safety culture of private and governmental sector delivery in the KSA. There is also a need to apply systematic methods that can reliably capture SEs and measure the transparency of both sectors.
Implications for policy, practice and research
The implementation of the National Sentinel Events Reporting Program is a major step in providing accessibility to a reporting system for all hospitals. However, optimizing of valid records and minimizing of suboptimal reports are significant requirements for better dashboard information. Also, it is likely that not all SEs are reported through the SERS, and alternative methods were not used in this study to identify such events. In the future, inclusion of other sources of data, such as the national program of patient complaints, claim databases, clinical audits and electronic medical records, should be considered to further address the research question and provide more reliable data on SEs. Using such data sources in addition to conducting studies that compare the quality of care and safety culture of private versus governmental hospitals will have the potential to capture the unreported SEs of noncooperative hospitals and could also provide an explanation for the very low reporting rate of the private sector. With such information, hospitals could be pushed to improve their transparency. Furthermore, given the gap between the level of reporting of SEs between SA and other countries, it is necessary to undertake deep research on the issue of underreporting in SA.
There have been several calls by researchers to promote a safety culture by establishing a blame-free learning environment for employees, and this study recommends working on initiatives or legislation that protects the institution or voluntary reporter from legal discovery. This protection will encourage learning from near misses and adverse occurrences. Also, providing hospitals with standards for the selection of recommendations related to the most common root causes for SEs and providing support to facilities to conduct analysis of the SE cases while they are under investigation and not retrospectively after the RCA is submitted will help to improve the RCAs in the concerned organization.
Finally, the main recommendations drawn from the present study are to activate improvement projects in the areas of maternal, child and surgical safety, provide education and training to improve the care of deteriorating patients in general wards and audit the adherence of staff to policies, procedures and evidence-based guidelines, such as ICU admission/discharge criteria, in addition to addressing staff shortages and lack of competency.
Conclusions
This study aimed to investigate the most frequent types of SEs in SA and their root causes to enhance patient safety and improve the quality of care. The top areas identified for improvement are the care of deteriorating patients in general wards, improvement of ICU admission/discharge criteria and a focus on maternal, child and surgical safety.
When compared to the rates of SEs in the wider literature, the findings suggest that these events are underreported in spite of a mandatory requirement for reporting by the MOH and the national accrediting body for health-care organizations.
On the basis of the study outcomes, in addition to the abovementioned recommendations for enhancing the safety of health care provided in the Saudi health-care system, concerns were raised about the barriers to reporting errors in clinical practice in SA, which need to be identified and addressed.
Ethics and other permissions
The Central Institutional Review Board of the Saudi Ministry of Health approved this study (National Registration Number with NCBE-KACST, KSA: (H-01-R-009)-IRB Log No: 20-208M).
Acknowledgements
The research group acknowledges all the health-care facilities for their ongoing contribution to the Sentinel Event Program and Sue Jessop for providing editorial assistance.
Funding
There was no funding from an external source for this study.
Data sharing statement
The data underlying this article will be shared on reasonable request to the corresponding author.
Contributorship
Conception and design: all authors; collection and assembly of data: Haifa Alnaima, Ms. Falaa Alahmari and Ms. Noor Alabdulkareem; analysis and interpretation of data: all authors; drafting of the article: Dr. Mafaten Chaouali; critical revision: Dr. Nasser Altalhi; guidance and final approval: Dr. Tareef Alaama; funding and administrative, technical and logistic support: all authors.
