Presentation
Using Systems Engineering Initiative for Patient Safety (SEIPS) to improve clinician situational awareness for Osteogenesis Imperfecta
DescriptionSummary:
Osteogenesis Imperfecta (OI) is a congenital defect of collagen synthesis that leads to brittle bones. Lack of awareness around a child’s OI condition among caregivers can cause unintended injury during non-OI clinic visits. While Clinical Decision Support (CDS) can help improve situational awareness for clinicians, locally developed CDS in the electronic health record (EHR) often fails to achieve its goals and may even cause further harm and alert fatigue due to mismatches between the design and actual workflows. In this abstract we describe (1) development of CDS through application of the Systems Engineering Initiative for Patient Safety (SEIPS) and (2) potential ways to integrate such research methods in operational contexts.
Background:
Osteogenesis Imperfecta (OI), also known as Brittle Bone Disease, is a congenital defect that affects collagen production which affects bone formation, bone strength, and the structure of other tissues1. Depending on the severity of disease patients can experience breaks in their bones during healthcare activities that involve routine positioning. The OI team in our pediatric healthcare system heard concern expressed from caregivers about lack of awareness around their child’s condition and concern for unintended injury during non-OI clinic visits. To address the unintended consequence and improve situational awareness around OI our team developed Clinical Decision Support (CDS).
Our multidisciplinary CDS team includes physicians, nurses, a pharmacist, and a human factors engineer. It includes representation from outpatient and inpatient contexts and ranges from intensive care units to general care and ambulatory specialties. For each CDS request, we have a standard framework for deployment and implementation that begins with using the IHI Model for Improvement, the Five Rights of CDS, and formative testing of CDS before deployment. However, in this case our team had no expertise in workflows of key user roles including how staff in radiology and phlebotomy in an outpatient context interacted with the EHR system and the patient. We recognized that a thorough understanding of the work system and the complex interactions is critical to the development of a usable CDS system. To address this gap in knowledge we used the Systems Engineering Initiative for Patient Safety (SEIPS) model People, Environment, Tools, Tasks (PETT) scan to understand the socio-technical workflows and what type of change could potentially improve situational awareness of staff.
Setting:
Large, urban, tertiary-care pediatric health system in the Southeastern United States including 3 free-standing Children’s hospitals with 673 beds, >40,000 inpatient discharges per year, >200,000 ED visits per year, and >1,000,000 total visits annually.
Methods:
The SEIPS model PETT scan is a checklist and documentation tool used to understand a work system, specifically its people, environments, tools and tasks2. It can be used to understand the presence of barriers or facilitators for each component. For this CDS implementation, a clinical informaticist used the checklist observing clinical workflows and interviewing key stakeholders to ensure understanding of interactions between the healthcare workers (HCW) and patients as well as the HCWs and the EHR system. Notes were member checked to ensure validity and themes were initially mapped to SEIPS concepts by the informaticist and verified/finalized by CDS team members.
Results:
There were three major HCW roles that consistently interacted with OI patients: registration, phlebotomy, and radiology technician.
Registration:
These HCWs had the least on-going interaction with patients and families. Their role is to check-in the patient for the appointment, confirm appointment details, give medication identification paperwork/armband, and collect any co-pays. They do not interact with the EHR beyond the specific duties assigned and have limited access to information within the EHR systems. They do not give any additional identification (such as an allergy armband) beyond a standard patient armband.
Phlebotomy:
These HCWs do a chart review of labs in the EHR prior to calling back a patient. Their chart review is limited to looking for ordered labs specifically and comparing to any paper lab orders if additional brought in by the patient. They then use a paper audit sheet for every patient to capture information that is not able to currently be documented in the EHR system or the Lab Information System. When interacting with the patient, they look for the standard patient armband, confirm if need pain management (such as cold spray), and review the labs they plan on obtaining.
Radiology Technicians:
The workflows for radiology technicians have more variance based on the modality used. For example, appointments for patients in Nuclear Medicine tend to have a deeper chart review then for appointments for an x-ray. However, in each modality, deep chart review was based more on individual staff members and modality rather than a consistent trained method. One tool that was used by all the radiology technicians regardless of modality, was the Snapshot Board. This shows patient appointment information on the top and a procedure report on the bottom half of the screen. Every single tech looks at this report prior to bringing the patient back for their procedure.
Discussion:
The PETT scan allowed our team to understand how patients and HCWs in outpatient areas interacted with each other and the EHR system. We were able to come up with some recommendations for our clinical partners to help with OI patient identification. For registration, it isn’t within the scope of their workflow to have them identify OI patients. For phlebotomy, leveraging the piece of paper that they use for every patient is likely to be the most effective intervention. For radiology technicians, putting identifying information on their report that they look at for every patient should be the best leverage point. While the PETT scan took more time than the traditional approach, it prevented us from spending time designing and building something in the EHR that our users would not have seen or would not have been able to act on. Further, to meet the speed requirements of operational requests, we did not do rigorous analysis of interviews and observations but used member checking to verify key concepts and mapped themes to SEPIS concepts enabling quick CDS and process design.
Osteogenesis Imperfecta (OI) is a congenital defect of collagen synthesis that leads to brittle bones. Lack of awareness around a child’s OI condition among caregivers can cause unintended injury during non-OI clinic visits. While Clinical Decision Support (CDS) can help improve situational awareness for clinicians, locally developed CDS in the electronic health record (EHR) often fails to achieve its goals and may even cause further harm and alert fatigue due to mismatches between the design and actual workflows. In this abstract we describe (1) development of CDS through application of the Systems Engineering Initiative for Patient Safety (SEIPS) and (2) potential ways to integrate such research methods in operational contexts.
Background:
Osteogenesis Imperfecta (OI), also known as Brittle Bone Disease, is a congenital defect that affects collagen production which affects bone formation, bone strength, and the structure of other tissues1. Depending on the severity of disease patients can experience breaks in their bones during healthcare activities that involve routine positioning. The OI team in our pediatric healthcare system heard concern expressed from caregivers about lack of awareness around their child’s condition and concern for unintended injury during non-OI clinic visits. To address the unintended consequence and improve situational awareness around OI our team developed Clinical Decision Support (CDS).
Our multidisciplinary CDS team includes physicians, nurses, a pharmacist, and a human factors engineer. It includes representation from outpatient and inpatient contexts and ranges from intensive care units to general care and ambulatory specialties. For each CDS request, we have a standard framework for deployment and implementation that begins with using the IHI Model for Improvement, the Five Rights of CDS, and formative testing of CDS before deployment. However, in this case our team had no expertise in workflows of key user roles including how staff in radiology and phlebotomy in an outpatient context interacted with the EHR system and the patient. We recognized that a thorough understanding of the work system and the complex interactions is critical to the development of a usable CDS system. To address this gap in knowledge we used the Systems Engineering Initiative for Patient Safety (SEIPS) model People, Environment, Tools, Tasks (PETT) scan to understand the socio-technical workflows and what type of change could potentially improve situational awareness of staff.
Setting:
Large, urban, tertiary-care pediatric health system in the Southeastern United States including 3 free-standing Children’s hospitals with 673 beds, >40,000 inpatient discharges per year, >200,000 ED visits per year, and >1,000,000 total visits annually.
Methods:
The SEIPS model PETT scan is a checklist and documentation tool used to understand a work system, specifically its people, environments, tools and tasks2. It can be used to understand the presence of barriers or facilitators for each component. For this CDS implementation, a clinical informaticist used the checklist observing clinical workflows and interviewing key stakeholders to ensure understanding of interactions between the healthcare workers (HCW) and patients as well as the HCWs and the EHR system. Notes were member checked to ensure validity and themes were initially mapped to SEIPS concepts by the informaticist and verified/finalized by CDS team members.
Results:
There were three major HCW roles that consistently interacted with OI patients: registration, phlebotomy, and radiology technician.
Registration:
These HCWs had the least on-going interaction with patients and families. Their role is to check-in the patient for the appointment, confirm appointment details, give medication identification paperwork/armband, and collect any co-pays. They do not interact with the EHR beyond the specific duties assigned and have limited access to information within the EHR systems. They do not give any additional identification (such as an allergy armband) beyond a standard patient armband.
Phlebotomy:
These HCWs do a chart review of labs in the EHR prior to calling back a patient. Their chart review is limited to looking for ordered labs specifically and comparing to any paper lab orders if additional brought in by the patient. They then use a paper audit sheet for every patient to capture information that is not able to currently be documented in the EHR system or the Lab Information System. When interacting with the patient, they look for the standard patient armband, confirm if need pain management (such as cold spray), and review the labs they plan on obtaining.
Radiology Technicians:
The workflows for radiology technicians have more variance based on the modality used. For example, appointments for patients in Nuclear Medicine tend to have a deeper chart review then for appointments for an x-ray. However, in each modality, deep chart review was based more on individual staff members and modality rather than a consistent trained method. One tool that was used by all the radiology technicians regardless of modality, was the Snapshot Board. This shows patient appointment information on the top and a procedure report on the bottom half of the screen. Every single tech looks at this report prior to bringing the patient back for their procedure.
Discussion:
The PETT scan allowed our team to understand how patients and HCWs in outpatient areas interacted with each other and the EHR system. We were able to come up with some recommendations for our clinical partners to help with OI patient identification. For registration, it isn’t within the scope of their workflow to have them identify OI patients. For phlebotomy, leveraging the piece of paper that they use for every patient is likely to be the most effective intervention. For radiology technicians, putting identifying information on their report that they look at for every patient should be the best leverage point. While the PETT scan took more time than the traditional approach, it prevented us from spending time designing and building something in the EHR that our users would not have seen or would not have been able to act on. Further, to meet the speed requirements of operational requests, we did not do rigorous analysis of interviews and observations but used member checking to verify key concepts and mapped themes to SEPIS concepts enabling quick CDS and process design.
Event Type
Oral Presentations
TimeWednesday, March 279:10am - 9:30am CDT
LocationSalon A-1
Hospital Environments