Social Network Analysis of Diversion of Controlled Substances in Canadian Hospitals: A Study Protocol
TimeThursday, April 152:00pm - 3:00pm EDT
One of the growing challenges for the Canadian healthcare system is dealing with the increasing economic and social burden of managing opioid abuse and addiction. Canada and the United States are two of the largest per-capita consumers of opioids in the world and there is a growing concern that many of the opioids that are being used for recreation are being stolen from healthcare facilities. Diversion, whereby controlled substances are lost or stolen from healthcare facilities, is a recognized problem. Canadian hospitals have been shown to lack the safeguards needed to detect and resolve inconsistencies in their inventories of controlled substances due to a large proportion of these incidents being classified as “unexplained losses”. Many system factors such as technologies, workflow processes, and environments within the medication use processes (MUP) impair healthcare facilities’ ability to detect and mitigate diversion, enabling its continuation. Previous research has identified technical factors contributing to in-hospital diversion risk. For example, hospitals that store controlled substances in narcotic cupboards with a physical lock and key allow nurses to access all inventory stored on the unit and use an honor system so that nurses accurately write down what was withdrawn and when. This is a weaker security measure than hospitals that use automated dispensing cabinets (ADCs), which require users to log in, and only grant them access to specific drawers for the medication they are seeking. As a result, hospitals with ADCs have a stronger technical safeguard against diversion by limiting access and improving the audit trail.
While technical factors contributing to diversion risk and safeguards has been previously described in the literature, there remains a gap in understanding which social factors contribute to drug diversion and how healthcare workers (HCWs) interact within the MUP. For example, close relationships between working colleagues can promote increased trust and therefore a reduced vigilance in adhering to witnessing procedure. Witnessing procedures are essential to ensure processes such as the disposal of unused opioids occurs as documented, rather than simply ‘pocketed’ and documented as wasted; falsely reporting drug waste is a common method of drug diversion.
Therefore, a greater understanding of how social relationships relate to technical risk factors is essential. Social Network Analysis (SNA) is an analytic approach which can reveal the types of social interactions, relationships, and tasks between HCWs that contribute to diversion risk within the MUP. Through the identification of which social factors contribute the most to diversion risk SNA can further be used to inform the planning, execution, and evaluation of any network alerting intervention needed to mitigate diversion risk in the MUP.
The proposed project aims to reveal the types of social relations and tasks between HCWs that contribute to drug diversion within the MUP of healthcare facilities, using SNA. The objectives will be to 1) map the communication and task networks of the MUP to describe their social network structure; 2) identify the influential actors within the MUP by assessing social interactions (talks with, gives advice to) and relationships (ties, tasks, and brokerages) of the staff by roles; and 3) understand HCWs attitudes towards diversion and safeguards in the MUP.
This study will employ exploratory sequential mixed methods using previously collected clinical observations data and cross-sectional network surveys in three units of two large hospitals in Toronto, Canada to identify drug diversion risks and contributing factors. This will allow for the combination of technical and social risk factors to produce a comprehensive sociotechnical analysis of diversion risk in hospital settings.
The study will consent purposively recruited HCWs involved in the MUP of each unit for cross-sectional surveys. The survey will collect data on demographics, social relations, and attitudes on drug diversion. The transcribed free-form field notes of the clinical observations will be used to describe the social network structure of each unit though contextually describing the environment and how HCWs complete tasks. Additionally, this data will be used to develop a detailed understanding of participants’ social interactions and relationships, as well as interactions with safeguards related to the MUP.
Sequential mixing of the two datasets will be done to ensure that there is a robust amount of network information available to map and analyze the communication and task networks of the units. The identified contributing factors will be modelled using SNA approaches. Specifically, SNA allows for identification of the type, importance, and quantity of relationships in a social network, which can subsequently be depicted using sociograms and network descriptive statistics to measure properties such as centrality (importance and influence of HCWs in the network) and density (level of connectedness within a network by HCWs). Determining network centrality of HCWs within the MUP is important in surveilling high risk HCWs and identifying vulnerabilities in the network (i.e. HCWs routinely used as a witness, auditor, or administrator of controlled substances in the MUP).
This study will reveal relationships between social interactions and tasks contributing to drug diversion risk in the MUP, which has a direct impact on patient and HCW safety due to opioid abuse. The outcomes of this study will include the development of an SNA method for drug diversion modelling that reflects the complex sociotechnical interactions between contributory factors in the MUP. This type of modelling can be used as a proactive risk management tool by providing the knowledge of which contributing factors may need to be controlled or mitigated for diversion prevention and provide unique contributions to the research literature in this area.
SNA may provide an entirely new avenue for research, development and testing of interventions that may reduce sociotechnical risk factors for diversion. For example, current pharmacy practices already firewall the personnel ordering controlled substances from those receiving, to minimize the risk of someone ordering and stealing the received shipment without anyone else knowing. Similar instances of social firewalling, or staff rotation, may help minimize diversion risks, and will be best guided by the results of SNA.