Identification of Organizational Factors Affecting the Safety of Operations: Foundation for Extending Human Reliability Analysis Methods

Abstract

Organizational factors (OFs) have been identified as a major root cause of accumulated errors and questionable decisions made by personnel and management in different safety-sensitive industries. They have long been viewed by the probabilistic risk assessment community as important contributors to operational safety. Yet, the ambition to quantify OFs remains one of their “grand challenges.” Such a gap also exists in Human Reliability Analysis (HRA). Current HRA methodologies incorporate OFs as performance influencing factors (PIFs), such as training adequacy. However, current OFs definitions present inconsistency and cause overlaps among PIFs. Moreover, the mechanism on how OFs may affect human errors needs further exploration. Therefore, here, our objective is twofold: developing a comprehensive list of OFs affecting operational safety, through an exhaustive literature review, and categorizing those OFs and determining the categories that show the highest influence in the reviewed literature. This will be a foundation to incorporate OFs into HRA.

Background and Objectives

Organizational factors (OFs) encompass the organizational structures, processes, and behaviors that influence the actions of individuals at work. These factors affect the likelihood of accidents and impact their severity. OFs have been identified as the root cause of accumulated errors and questionable decisions made by personnel and management in different safety-sensitive industries, including the oil and gas, nuclear power, and transportation sectors (e.g. Pate-Cornell & Murphy, 1996; Tabibzadeh & Meshkati, 2014).

Discussions on organizational aspects as contributors to operational safety were intensified after major catastrophes such as the Three Mile Island in 1979 in the nuclear power industry, the Piper Alpha accident in 1988 in the oil and gas industry, and the Space Shuttle Challenger accident in 1986 in the aviation industry. Despite these discussions, the importance of a systemic approach to safety that considers the human, organizational, and technological factors, and the complexity of the interrelationships among them needs to be highlighted more than before, as major man-made catastrophes have continued to occur, for example, the Ethiopian Airlines 302 crash in 2019, the Fukushima accident in 2011, and the BP Deepwater Horizon blowout in 2010.

Organizational problems have long been viewed by the probabilistic risk assessment (PRA) community as important contributors to operational safety in high-risk industries. Yet, despite advances in OFs-related disciplines, the ambition to quantify OFs remains one of PRA’s “grand challenges” (e.g., Wellock, 2021). In addition, the field of Human Reliability Analysis (HRA) shares milestones with OFs research. HRA methodologies aim at identifying, modeling, and qualitatively and quantitatively assessing human errors during system operations. It is demonstrated that OFs significantly influence human errors: a considerable number of human errors could have been prevented if the organization had implemented appropriate precautions before incidents.

Some of the current HRA methodologies, such as the SPAR-H (e.g., Gertman et al., 2005), incorporate OFs as Performance Influencing Factors (PIFs), for example, procedures quality and training adequacy (Alvarenga et al., 2014). However, the consideration of OFs within HRA lacks a consistent, theoretically based taxonomy, a causal model, and a robust quantification framework. In one aspect, the definitions of current OFs present inconsistency and caused overlaps among PIFs. Moreover, HRA communities lack an exploration of the mechanism on how OFs may affect human errors. Therefore, in this study, our objective is twofold: (1) to form a comprehensive list by identifying the OFs affecting operational safety in high-risk industries through comprehensive literature review and (2) to classify those OFs into main categories and determine the categories that show the highest influence in the reviewed literature. The results benefit to clarify the research gap of OFs in current HRAs, and further to incorporate the OFs into HRA. It also helps address the challenge of PIFs’ inter-dependencies caused by OFs in HRA communities in the future.

Method

We conducted an exhaustive literature review to identify a comprehensive list of OFs that have affected the safety of operations in different high-risk industries, as a foundation to incorporate those factors into HRA more effectively and cohesively. We did not restrict our search to a single industry; instead, we explored various safety-sensitive industries, encompassing oil and gas, aviation, transportation, nuclear power, and healthcare sectors. The OneSearch option provided by California State University, Northridge’s Library, which includes subscriptions to several pertinent databases such as Scopus, Engineering Village, ProQuest, and PubMed, was utilized.

The keywords “organizational factors” (contains an exact phrase in title) AND “safety” (contains in any field) were used, resulting in 341 search items. After refining the results to only include articles, conference proceedings, book chapters, and reports that were written in English, we found 321 search results. Further assessment was carried out based on two inclusion criteria: (1) the study addressed safety and (2) the abstract captured OFs affecting safety or there was an indication that some OFs were captured in the full text of the publication. Additionally, we viewed the references of those publications to find more possible relevant sources.

A total of 278 references were found suitable. Each of the 278 references were reviewed and their identified OFs were captured. Moreover, some data analysis was conducted to identify the most influential OFs. However, before conducting this analysis, we categorized our captured OFs based on their definitions into some clusters.

Later, we conducted a study to identify the elements for modeling organizational factors in HRA. The research team conducted a thorough review of 269 papers. The review process was divided into three distinct phases. Initially, we identified 269 relevant papers. Among these, 47 were deemed directly relevant. In the final review stage, 24 papers were classified as highly relevant to the research. Papers that addressed both HRA and OFs were categorized as highly relevant, whereas those that focused on one aspect while mentioning the other were considered medium relevant. The team’s systematic approach aimed to identify the most pertinent publications, with a primary focus on understanding the HRA methodologies used and the specific industries addressed in the existing literature. Ultimately, our goal was to uncover any gaps in the current research landscape related to HRA elements in modeling OFs.

Findings

Utilizing the described literature review in the Method section, studies across several safety-sensitive industries were captured, with most studies being in the healthcare, nuclear power, oil and gas, and aviation sectors. Additionally, studies across other high-risk industries, such as maritime transportation and railroad were reviewed. After reviewing the 278 references, approximately 1,100 unique OFs were identified following some data cleaning. Subsequently, the frequency of each organizational factor was determined through data analysis, considering the number of references that cited that particular OF.

In the next phase, those OFs were categorized based on the definitions provided for them and their similarity. Five main categories were created encompassing safety culture/climate, communication and coordination, leadership and management support, education and training, and risk and safety management. The remaining OFs were grouped into the “other” category, as they lacked specific connection with each other.

We used the frequency of being cited by reviewed references as an indication of influence. The five above-stated categories of OFs had frequencies of 94, 93, 84, 77, and 10, respectively. Among the “other” category, “resource management” was the OF with the highest frequency of 13.

Based on this analysis, we identified that the four categories of safety culture/climate, communication and coordination, leadership and management support, and education and training are the most influential OFs clusters to be used toward HRA in our future study. The Bayesian Belief Network (BBN) will be used as the methodology to incorporate these OFs into HRA and capture their impact on human performance/human error in a quantitative manner. Furthermore, a BBN will enable the development of a causal-based model in which captured organizational factors are incorporated into the model using a theoretically based taxonomy. The use of the BBN has been demonstrated in HRA methods, such as the Phoenix (Ekanem et al., 2024; Ramos et al., 2021) and HSIA-BN (Cheng et al., 2024), which could provide a causal and methodological foundation for our research.

Takeaways

Organizational factors have been a major root cause of incidents in different safety-sensitive industries, including the oil and gas, nuclear power, transportation, and healthcare sectors. Through a comprehensive literature review of 278 relevant references, the four categories of safety culture/climate, communication and coordination, leadership and management support, and education and training were identified as the most influential categories of organizational factors. This is a unique study, as it conducted an exhaustive literature review to capture a comprehensive list of organizational factors across different safety-sensitive industries to create a solid foundation for our future study when we incorporate those OFs into our human reliability analysis with the main goal of quantitatively capturing the influence of the OFs on human performance. This will fill an important gap as despite advances in OFs-related disciplines, the ambition to quantify OFs remains one of PRA’s and HRA’s grand challenges.