Providing Evidence and Developing a Toolkit to Accelerate the Adoption of Patient Photographs in Electronic Health Records

Project Final Report (PDF, 1.49 MB) Disclaimer

Patient photos and alerts displayed as standard features in the electronic health record (EHR) significantly reduce wrong-patient order errors and have the potential to enhance patient safety. Achieving system-wide implementation with high compliance is feasible by addressing operational barriers.

Project Details -
Completed

Computerized provider order entry (CPOE) systems, while reducing overall medical errors, can increase wrong-patient errors, with one hospital reporting over 5,000 such incidents annually. These errors, which involve medications, lab tests, procedures, and discharge orders, can cause serious patient harm. Identification (ID) verification alerts have been shown to effectively reduce these errors, and the Office of the National Coordinator for Health Information Technology (ONC) recommends verifying patient ID at the time of ordering. Evidence, although limited, indicates that displaying patient photographs in EHRs also reduces wrong-patient orders, as face recognition requires less cognitive effort than reading names; thus, ONC also recommends displaying patient photographs in EHRs. However, only 18 percent of hospitals have adopted this practice, citing major barriers such as a lack of evidence and implementation challenges; notably, patient refusal is not among these barriers. This study evaluated the effectiveness of displaying patient photos in EHR systems to prevent wrong-patient errors when using CPOE systems.

The specific aims of the research were as follows:

  • Provide rigorous evidence for the efficacy of displaying patient photographs in EHRs.
  • Identify factors that impact the effectiveness of patient photographs on the frequency of wrong-patient orders, including characteristics of providers, patients, and orders.
  • Demonstrate generalizability using different EHR systems to evaluate the effectiveness of patient photographs for reducing the frequency of wrong-patient orders across sites and EHR systems.
  • Provide implementation guidance for displaying patient photographs in EHRs.
  • Examine user perceptions and experience of patient photographs in the different EHR systems, including staff, providers, and patients.

Researchers conducted two large multisite studies using randomized controlled trials (RCTs) in two different EHR systems, Allscripts and Epic, each with varying photo functionalities and configurations. Providers were randomly assigned to view screens either with or without a patient photograph. Photos were displayed either passively as a banner on the EHR screen (requiring no action by the user), actively through an interruptive “popup” alert, or in combination. If no photo was available, the alert appeared without a photo. Outcomes were measured using the Wrong-Patient Retract-and-Reorder (WP-RAR) measure, a validated method for identifying wrong-patient orders that is endorsed by major patient safety organizations. Researchers discovered that integrating patient photos and alerts into standard EHR features led to a significant reduction in wrong-patient order errors. Notably, even without patient photos, alerts alone reduced wrong-patient order errors by approximately 12 percent, supporting earlier research and highlighting the efficacy of alerts in error reduction. Overall, 62 percent of order sessions included a photo, showing that a system-wide implementation with high compliance is feasible when operational barriers are addressed. (Note that at the end of the project period, the RCT analysis for Allscripts was still ongoing.) To address concerns about these barriers, the study team developed implementation guidance titled “Implementing Patient Photographs in EHR Systems,” which can help healthcare organizations adopt patient photographs system-wide. This project has the potential to provide healthcare systems with the rationale and resources to accelerate the adoption of patient photographs in their EHRs, and to provide patient safety and regulatory bodies with the evidence needed to require the use of patient photographs in EHRs.

Displaying Patient Photos in Medical Records Reduces Errors, Improves Patient Safety

Theme:

Supporting Health Systems in Advancing Care Delivery

Subtheme:

Optimizing Patient Safety Using Digital Healthcare Solutions

Patient photos displayed in the electronic health record significantly reduce wrong-patient order errors and improve patient safety.

Wrong patient orders can harm patients

While digital solutions like computerized provider order entry (CPOE) systems have reduced medical errors and improved patient safety, the risk of placing orders for the wrong patient still exists. Wrong patient orders can relate to medications, lab and imaging tests, procedures, nursing orders, consults, discharges, and other orders.

“Orders placed on the wrong patient should be a ‘never event,’ as in it should never happen,” said Dr. Jason Adelman, a practicing clinician and leading patient safety expert at Columbia University. Displaying photographs of the patient in the electronic health record (EHR) could potentially reduce wrong patient errors. “Our brains process images quicker and more effortlessly than words, making it much easier for a clinician to know if they are in the wrong record when they see a photo,” says Dr. Adelman.

Displaying photos in the EHR

Dr. Adelman led a study to see whether displaying patient photos—and how they were presented—in EHRs prevented orders on the wrong patients. The team wanted to assess the effect of displaying patient photos in a passive way, using a photo in a banner, called “Storyboard” in Epic, where there is no specific action a provider needs to take. This compares to photos presented in an interruptive popup, where a patient photo is displayed in an ID verification alert, and providers must click to verify they are in the correct patient’s chart. The researchers tested three interventions versus a no-photo group: 1) Storyboard with photo, 2) the popup alert with photo, and 3) both Storyboard with photo and popup alert with photo.

Wrong patient orders were identified by using the Wrong-Patient Retract-and-Reorder Measure (WP-RAR), a validated, reliable, and automated method for identifying wrong-patient orders developed by Dr. Adelman. The WP-RAR measure identifies orders placed for a patient and retracted within 10 minutes, then reordered by the same clinician for a different patient within the next 10 minutes. These are called near-miss errors, which are self-caught by the clinician before they reach the patient and potentially cause harm.

A key piece to making the photo verification work? The availability of patient photos in the EHR. At the start of the study, only about 40 percent of patients had photos available. Over 3 months, the research team worked with patient services, information services, and operations to increase the number of patient photos. This included developing a job aid, training staff on how and when to take photos, and activating settings on kiosks to enable patients to take their own photo during registration.

“We need rigorous evidence to demonstrate a safety practice, which can motivate hospitals to adopt these practices. We hope this research can help accelerate the adoption of photos in EHRs.” – Dr. Jason Adelman

Photos improve patient safety

The research demonstrated that all three combinations of photo display significantly reduced orders on the wrong patients compared with no photos displayed. Notably, there were no significant differences between the two different photo displays, Storyboard versus popup; both showed improved action by the provider. Displaying the patient’s photo in a noninterruptive banner in the EHR worked as well as the popup alert. The findings are significant, according to Dr. Adelman: “Photos in Storyboard passively work so well that adding a pop-up has a small increase, but not statistically significant and likely not worth the burden to alert fatigue.” Given the burden of alert fatigue on clinicians, he advocates the use of passive intervention to maximize benefit and minimize burden on clinicians.

Findings from the study support the widespread implementation of patient photos by health systems and the continued development and expansion of features to display patient photos by EHR vendors. To support uptake of photo use, Dr. Adelman and his team are developing a toolkit that summarizes their experience and includes synthesized materials developed during implementation. The kit will include various methods on how to incorporate patient photos in different technical environments. An addendum provides specifications on how to capture a quality patient photo, checklists and templates, sample policies and workflows, and training and patient education materials.

This Research is Featured in

Principal Investigator

Adelman, Jason Stuart

Organization

Columbia University Health Sciences

Research Profile

Funded Amount

$1,577,033

Research Type

Completed

This Research is Featured in

Related Stories




Leveraging Health IT to Test Solutions That Are Replicable, Scalable, and Improve Patient Safety

Impact

“Studies like that by Adelman and colleagues point the way to the creation of a digital learning health care system, in which the results of the interactions between clinicians (and, increasingly, patients and families) and the EHR are analyzed to help guide the strategies that lead to the highest value and most satisfying care. Having spent tens of billions of dollars digitating the health care system, it is essential to take advantage of the unique capacity of digital tools to allow clinicians and health care systems to learn from every click.” -editorial by Drs. Wachter, Murray, and Adler-Milstein4

Wrong-patient errors can affect any patient in any healthcare setting for a variety of reasons.

Dr. Jason Adelman, named as one of 50 experts leading the field of patient safety in 2018 by Becker’s Hospital Review,2 has led multiple AHRQ Health IT-funded research efforts centered on health IT safety. The research of Dr. Adelman, Executive Director of Patient Safety Research at Columbia University Irving Medical Center/NewYork-Presbyterian Hospital, is widespread and being replicated by other organizations. Dr. Adelman is part of the ECRI Patient Safety Collaborative, where he is advising several hospitals on how to implement the Wrong-Patient Retract-and-Reorder (RAR) Measure that he developed to evaluate the frequency of wrong-patient errors that occur through CPOE systems. When a clinician places an order, then cancels the order and places the same order for a different patient within the next 10 minutes, the measure flags it as a wrong-patient RAR event. While capturing the instance of retracting and reordering mistakes does not correct or prevent errors, it provides the facility with information to discover error trends along with opportunities to intercept processes that lead to such errors. Using such data as evidence of the CPOE’s potential impact on patient safety, decision makers can factor RAR measures in choosing one health IT design over another.

Having multiple EHRs open simultaneously does not increase wrong-patient orders.

“Placing orders on the wrong patient should never happen. Yet, human error is very common in the healthcare environment. Healthcare is inherently complex and heavily reliant on people rather than technology to protect patients from harm. It will likely take a multi-pronged health IT approach to prevent these types of errors.”
- Dr. Jason Adelman

Dr. Adelman’s AHRQ Health IT-funded research found that restricting clinicians to having one EHR record open at a time did not significantly reduce the rate of wrong-patient order errors compared with allowing up to four records to be open concurrently. His work did not support the ONC and Joint Commission recommendation that EHR systems should only allow one record to be displayed at a time. In the study, published in the May 14, 2019 issue of JAMA,3 using the RAR measure, Dr. Adelman and his team compared the risk of wrong-patient orders while accessing one versus four records open in a variety of clinical settings, including hospitals, EDs, and outpatient facilities. While no differences in wrong-patient orders were observed between those clinician groups, there was considerable variation in the frequency of errors in different clinical settings. The rate of wrong-patient order errors was lowest in outpatient settings, where physicians may care for one patient at a time. The highest rates, meanwhile, were seen in inpatient critical care and obstetrics units, which reflected differences in workflows and number of patients being cared for simultaneously, researchers noted. The research offers insights for healthcare systems that are trying to balance patient safety with the needs of busy clinicians who need tools for efficient workflow.

The right patient, right dose, right medication, right route, and right frequency.

The Health IT Program has also funded Dr. Adelman to extend research on RAR, to validate and evaluate the reliability of RAR using a different EHR system, so it can be implemented and may advance interventions to prevent these serious and complicated safety issues. The intent is to develop health IT measures that identify deviations from the Five Rights of Medication Safety: right patient, right dose, right medication, right route, and right frequency.

Use of photographs to prevent errors.

An additional Adelman-led, AHRQ-funded study is assessing the effectiveness of using patient photographs as an additional identifier in the EHR system to avoid wrong-patient errors when using CPOE systems. The research team will conduct a randomized controlled trial of wrong-patient error rates between systems with patient photos and without. Ultimately, the team plans to develop a toolkit with guidance to help other health systems in implementing patient photos in EHR systems.

2 https://www.beckershospitalreview.com/lists/50-experts-leading-the-field-of-patient-safety-2018.html

3 https://jamanetwork.com/journals/jama/fullarticle/2733207

4 https://jamanetwork.com/journals/jama/article-abstract/2733189