INTERSECTION OF HEALTH IT AND PATIENT SAFETY HARDEEP SINGH, MD, MPH - PowerPoint PPT Presentation

MEASURING THE INTERSECTION OF HEALTH IT AND PATIENT SAFETY HARDEEP SINGH, MD, MPH H OUSTON VA C ENTER FOR I NNOVATIONS IN Q UALITY , E FFECTIVENESS & S AFETY M ICHAEL E. D E B AKEY VA M EDICAL C ENTER B AYLOR C OLLEGE OF M EDICINE Twitter: @HardeepSinghMD 1

Technology Hype Cycle 2

So Why the Disillusionment? 3  Health IT radically changes practice  Implementation & use inherently prone to failure  Safety benefits require a journey!  Overhaul of underlying system & processes  We never prepared for unintended consequences

5

Errors Reported by Front Line Docs 6  “Dependence on accessibility of hospital -based EHR led to delayed diagnosis of life-threatening problem when test result performed at hospital (that I was unaware had been performed) was not transmitted to my office”  “Discontinue fentanyl patch was ordered in EHR and IV started. Patch was deleted from MAR (medication administration record) so nurse didn’t know it was on when started the drip. Patient died of overdose.”

Safety Begins with Measurement 7 We need to measure Health IT safety for improvement But we cannot measure what we cannot define!

Defining Health IT Safety – 3 domains 8  Domain 1: Safe health IT:  Events unique/specific to health IT Sittig & Singh N Engl J Med. 2012 Nov 8;367(19):1854-60

9

Defining Health IT Safety – 3 domains 10  Domain 1: Safe health IT:  Events unique/specific to health IT  Domain 2: Using health IT safely:  Unsafe or inappropriate use of technology  Unsafe changes in the workflows that emerge from technology use Sittig & Singh N Engl J Med. 2012 Nov 8;367(19):1854-60

No Fail-Safe Communication 11  Evaluation of 1,163 outpatient abnormal lab & 1,196 abnormal imaging test result alerts  7% abnormal labs lacked timely follow-up  8% abnormal imaging lacked timely follow-up  Follow-up in acknowledged vs. unacknowledged alerts? Singh et al Am J Med 2010 & Singh et al Archives of Int Med 2009

Not Really A Technology Problem! 12

13

And More Digital Data Is on the Way 14  Smartphone  Wearables  “Patients can now continuously monitor their data real- time and send it to their docs”

Defining Health IT Safety – 3 domains 15 Domain 1: Safe health IT :  Events unique/specific to EHRs  Domain 2: Using health IT safely:  Unsafe or inappropriate use of technology  Unsafe changes in the workflows that emerge from technology use   Domain 3: Using health IT to improve safety Leveraging health IT to identify unsafe care processes  and potential patient safety concerns before har m Sittig & Singh N Engl J Med. 2012 Nov 8;367(19):1854-60

Lessons from Health IT Research 17  EHR-based systems better than paper  Not achieving full potential  Need sociotechnical approaches to measure and improve safety Sittig and Singh JGIM 2012; Arch IM 2012

8-dimensional Socio-Technical Model of Safe & Effective Health IT Use 18 Organizational Policies, Workflow & Communication Procedures, & Culture External Rules & Regulations Content Personnel Hardware & Software Measurement & Monitoring Sittig Singh QSHC 2010

Singh Sittig BMJ Qual Saf doi:10.1136/bmjqs-2015-004486

Measurable Aspects of Health IT 20 Type of HIT-related safety concern Examples 1. Instances in which HIT fails during use or is otherwise Broken hardware or not working as designed. software “bugs” 2. Instances in which HIT is working as designed, but the Usability issues design does not meet the user’s needs or expectations. 3. Instances in which HIT is well-designed and working Duplicate order alerts correctly, but was not configured, implemented, or that fire on alternative used in a way anticipated or planned for by system PRN pain medications designers and developers Sittig Classen Singh J Am Med Inform Assoc. 2014 Oct 20

Measurable Aspects of Health IT 21 Type of HIT-related safety concern Examples 4. Instances in which HIT is working as designed, Medication order for extended and was configured and used correctly, but release morphine inadvertently interacts with external systems (e.g., via changed to immediate release hardware or software interfaces) so that morphine by error in interface data is lost or incorrectly transmitted or translation table displayed . 5. Instances in which specific safety features or Hospitalized patient functions were not implemented or not inadvertently receives 5 grams available (i.e., HIT could have prevented a of acetaminophen in 24 hours safety concern). because maximum daily dose alerting was not available Sittig Classen Singh J Am Med Inform Assoc. 2014 Oct 20

Proactive Measurement 22  ONC- sponsored “Safety Assurance Factors for EHR Resilience (SAFER) project”  Proactive risk assessment and guidance  “1 st draft” of best practices and knowledge  Self-assessment; not meant to be regulatory  Focused on high-risk areas  Nine guides — all freely available http://www.healthit.gov/safer Singh et al BMC Med Inf 2013

Policy and Practice Implications 23  Robust measurement and monitoring of health IT safety essential  Certain risk areas now well defined and amenable to measurement for QI/safety purposes  Health IT safety measurement should be an essential component of overall patient safety strategy

Thank you… 25 Funding Agencies Department of Veterans Affairs  Agency for Health Care Research & Quality  National Institute of Health  Office of National Coordinator (SAFER Guides)  Multidisciplinary team at Houston-based  VA Health Services Research Center of Innovation Contact Information… Hardeep Singh, MD, MPH hardeeps@bcm.edu @HardeepSinghMD