Publicatie
Publicatie datum
Patient safety and medical devices: interacting contributing factors leading to unintended patient harm.
Baartmans, M.C. Patient safety and medical devices: interacting contributing factors leading to unintended patient harm. Amsterdam: Vrije Universiteit, 2024. 203 p. Proefschrift van de Vrije Universiteit
Download de PDF
Dit proefschrift laat zien dat medische technologie op verschillende manieren betrokken is bij gebeurtenissen die leiden tot onbedoelde potentieel vermijdbare schade aan patiënten. Bestaande methoden voor het onderzoeken en analyseren van dergelijke gebeurtenissen brengen de rol van de technologie onvoldoende aan het licht. Daarom werd de GAM ontwikkeld, waarin het systeemdenken en een veiligheidsergonomisch perspectief zijn ingebed. Een ziekenhuisoverstijgende GAM-analyse van calamiteiten toonde aan dat kenmerken van de technologie vaak interacteren met factoren uit andere socio-technische domeinen zoals de organisatie, de patiënt en de professional. Gezamenlijk leiden deze factoren tot schade aan de patiënt. Aanvullende diepgaande analyses op specifieke onderwerpen waarbij gespecialiseerde instrumenten werden ingezet hielpen om onze
inzichten nog verder te vergroten en meer specifieke verbetermaatregelen te formuleren.
Uit deze bevindingen volgen enkele aanbevelingen voor praktijk en beleid, waaronder adviezen voor het inrichten van toekomstige calamiteitenanalyses en strategieën om het aantal schades aan patiënten waarbij medische technologie een rol speelde te verminderen. Toekomstig onderzoek zou I) prospectieve methoden moeten omvatten om de veilige toepassing van medische technologie te onderzoeken, II) zich moeten richten
op de veiligheid van de inzet van technologie buiten de muren van het ziekenhuis, en III) potentiële patiëntveiligheidsrisico’s moeten bestuderen die verband houden met medische software en informatietechnologie in de gezondheidszorg, waaronder de opkomende toepassing van kunstmatige intelligentie.
inzichten nog verder te vergroten en meer specifieke verbetermaatregelen te formuleren.
Uit deze bevindingen volgen enkele aanbevelingen voor praktijk en beleid, waaronder adviezen voor het inrichten van toekomstige calamiteitenanalyses en strategieën om het aantal schades aan patiënten waarbij medische technologie een rol speelde te verminderen. Toekomstig onderzoek zou I) prospectieve methoden moeten omvatten om de veilige toepassing van medische technologie te onderzoeken, II) zich moeten richten
op de veiligheid van de inzet van technologie buiten de muren van het ziekenhuis, en III) potentiële patiëntveiligheidsrisico’s moeten bestuderen die verband houden met medische software en informatietechnologie in de gezondheidszorg, waaronder de opkomende toepassing van kunstmatige intelligentie.
Aim
Hospitals are high-technical environments where medical devices are essential for providing patient care. Often the application of medical devices is safe, but occasionally devices are involved in events resulting in unintended patient harm. In some cases this patient harm could potentially have been prevented. However, much is unknown about how events leading to potentially preventable patient harm emerge and what role the device plays. This thesis aims to increase knowledge and provide deeper understanding on how medical devices are involved in unintended patient harm in Dutch hospital care, and identify leads to further improve patient safety.
Methods
To fulfill this aim various studies were conducted, namely
I) a patient record review study on the incidence and nature of patient harm where medical devices are involved,
II) a systematic literature review on one method currently applied to study the causes patient harm,
III) the stepwise development of a new sentinel event analysis method, and
IV) a retrospective aggregated cross-hospital analysis of sentinel event reports identifying contributing factors including those factors that relate to devices.
Results
Based on data from reviewing 2,998 patient records from 20 Dutch hospitals, we found a potentially preventable Adverse Eevent involving a Medical Device (AEMD) occurred in 1.9% (95CI 1.5%- 2.5%) of patients who died during their admission in Dutch hospitals in 2019. Perforations, bleedings, and infections after invasive (endo)scopic procedures, heart- and vascular implants, and tubes for enteral feeding or drainage were recurring topics identified in these AEMDs. The contribution of the device often laid in provoking a known complication, which in most cases was considered non-preventable. Preventability often not necessarily related directly to device application, but rather laid in the subsequent management of the perforation, bleeding or infection. Current methods for studying such events not always successfully clarify the role of medical devices. Therefore the novel GAM was developed, in which a human factors perspective and systems thinking approach are embedded. Aggregate cross-hospital GAM-analysis of sentinel events from 28 Dutch general hospitals indicates that characteristics of the device itself often interact with contributing factors related to other socio-technical domains such as the organization, patient, and professional. Jointly these factors trigger patient harm. Additional in-depth analysis on specific topics using specialized instruments can help to further increase our understanding and assist in formulating more specific measures for improvement.
Conclusions
Preventability often not necessarily related directly to device application, but rather laid in the subsequent management of the perforation, bleeding or infection. Current methods for studying such events not always successfully clarify the role of medical devices. Therefore the novel GAM was developed, in which a human factors perspective and systems thinking approach are embedded. Aggregate cross-hospital GAM-analysis of sentinel events from 28 Dutch general hospitals indicates that characteristics of the device itself often interact with contributing factors related to other socio-technical domains such as the organization, patient, and professional. Jointly these factors trigger patient harm. Additional in-depth analysis on specific topics using specialized instruments can help to further increase our understanding and assist in formulating more specific measures for improvement.
Recommendations
From the findings of this thesis follow some important recommendations for practice and policy, including guidance on future sentinel event investigations and strategies to reduce adverse events involving medical devices. Future research should I) include prospective methods to study the safe application of medical devices, II) also focus on safety of devices in nonhospital settings, and III) study potential patient safety risks related to medical software and information technology in health care, including the emerging application artificial intelligence.
Hospitals are high-technical environments where medical devices are essential for providing patient care. Often the application of medical devices is safe, but occasionally devices are involved in events resulting in unintended patient harm. In some cases this patient harm could potentially have been prevented. However, much is unknown about how events leading to potentially preventable patient harm emerge and what role the device plays. This thesis aims to increase knowledge and provide deeper understanding on how medical devices are involved in unintended patient harm in Dutch hospital care, and identify leads to further improve patient safety.
Methods
To fulfill this aim various studies were conducted, namely
I) a patient record review study on the incidence and nature of patient harm where medical devices are involved,
II) a systematic literature review on one method currently applied to study the causes patient harm,
III) the stepwise development of a new sentinel event analysis method, and
IV) a retrospective aggregated cross-hospital analysis of sentinel event reports identifying contributing factors including those factors that relate to devices.
Results
Based on data from reviewing 2,998 patient records from 20 Dutch hospitals, we found a potentially preventable Adverse Eevent involving a Medical Device (AEMD) occurred in 1.9% (95CI 1.5%- 2.5%) of patients who died during their admission in Dutch hospitals in 2019. Perforations, bleedings, and infections after invasive (endo)scopic procedures, heart- and vascular implants, and tubes for enteral feeding or drainage were recurring topics identified in these AEMDs. The contribution of the device often laid in provoking a known complication, which in most cases was considered non-preventable. Preventability often not necessarily related directly to device application, but rather laid in the subsequent management of the perforation, bleeding or infection. Current methods for studying such events not always successfully clarify the role of medical devices. Therefore the novel GAM was developed, in which a human factors perspective and systems thinking approach are embedded. Aggregate cross-hospital GAM-analysis of sentinel events from 28 Dutch general hospitals indicates that characteristics of the device itself often interact with contributing factors related to other socio-technical domains such as the organization, patient, and professional. Jointly these factors trigger patient harm. Additional in-depth analysis on specific topics using specialized instruments can help to further increase our understanding and assist in formulating more specific measures for improvement.
Conclusions
Preventability often not necessarily related directly to device application, but rather laid in the subsequent management of the perforation, bleeding or infection. Current methods for studying such events not always successfully clarify the role of medical devices. Therefore the novel GAM was developed, in which a human factors perspective and systems thinking approach are embedded. Aggregate cross-hospital GAM-analysis of sentinel events from 28 Dutch general hospitals indicates that characteristics of the device itself often interact with contributing factors related to other socio-technical domains such as the organization, patient, and professional. Jointly these factors trigger patient harm. Additional in-depth analysis on specific topics using specialized instruments can help to further increase our understanding and assist in formulating more specific measures for improvement.
Recommendations
From the findings of this thesis follow some important recommendations for practice and policy, including guidance on future sentinel event investigations and strategies to reduce adverse events involving medical devices. Future research should I) include prospective methods to study the safe application of medical devices, II) also focus on safety of devices in nonhospital settings, and III) study potential patient safety risks related to medical software and information technology in health care, including the emerging application artificial intelligence.
