Contact email

adalal1@bwh.harvard.edu

Overview

This project utilized an automated email workflow for notifying providers of patients’ post-discharge test results in order to improve rates of actionable post-discharge test result follow-up. 

Department

Department of Medicine

Division

Division of General Internal Medicine

Collaborators

Adam Schaffer, MD, MPH
Esteban F. Gershanik, MD, MPH, MMSc
Ranganath Papanna, MD
Katyuska Eibensteiner
Nyryan V. Nolido, MA
Cathy S. Yoon, MS1
Deborah Williams, MHA
Stuart R. Lipsitz, ScD
Christopher L. Roy, MD
Jeffrey L. Schnipper, MD, MPH

Status/Stage of Development

Planning

Practice Setting

Academic Medical Center

National/Policy Context

  • In a 2003 study published in the Annals of Internal Medicine, the authors found that 59% of preventable adverse events during the discharge process involved communication challenges. 
    • There is a need for increased communication between hospital-based and outpatient providers during the discharge process to improve patient safety. 
  • A 2011 systematic review found that among hospitals nationwide, between 20 and 60 percent of inpatients did not receive follow-up regarding test results. 
    • Delayed or absent follow-up of important test results can cause patient harm due to incorrect treatment, disease progression, and misdiagnosis among several other factors. 
  • Despite the widespread adoption of electronic health records, EHRs do not uniformly send notification of post-discharge test results to discharging physicians or outpatient providers.

Local/Organizational Context

  • A prior study from the Brigham and Women’s Hospital determined that 41% of patients leave the hospital with outstanding tests that have pending results, many of which were actionable (e.g., positive cultures, malignant cells on biopsy specimen). Physicians were often unaware of these actionable results.
  • In a previous study, the authors created email notifications that alert the discharging attending and primary care physicians of results from Tests Pending at Discharge (TPADs), but they now seek to determine whether these notifications affect the proportion of actionable TPADs that receive follow up.  They demonstrated that the automated notification system improved awareness of TPAD results for both attendings and primary care physicians.

Patient Population Served and Payor Information

  • The average age of the patients whose post-discharge test results were involved in the intervention was 60.2 years. 
    • 45.5% of patients who received the intervention were male and 54.5% were female, and 65% of patients were Caucasian and 32.5% were “non-Caucasian”.
  • 56.1% of patients who received the intervention had a median income less than $47,000 and 40.7% had a median income of greater than $47,000. 
  • 33.3% of patients who received the intervention had private insurance and 59.4% of patients had public or government-sponsored insurance.

Leadership

  • Anuj K Dalal, MD (Principal Investigator)
  • Katyuska Eibensteiner (Research Assistant)
  • Nyryan Nolido (Project Manager)

Project Research + Planning

  • In our prior study we describe the technical and user requirements of the automatic email notification system. Specifically, the workflows to update and maintain the identify of the inpatient attending at discharge and primary care physician were actively maintained and updated, enabling us to send emails to facilitate closed loop communication of these results to key physicians responsible for follow-up. There was technical minimal work required to run the “job” to identify patients with tests pending at discharge.

Tools or Products Developed

  • Automated email alert: When a post-discharge test result returns, an email is automatically sent to the institutional email addresses of the discharging physician and primary care physician (only if they were in-network) describing the test result and assigning responsibility for follow-up.

Tech Involved
  • Desktop computer
  • Electronic medical record
  • Email
Team Members Involved
  • Physicians
  • Software Engineers/ IT
  • Support Staff
Workflow Steps

  • A patient is discharged from the hospital’s general medicine or cardiology service with a Test Pending at Discharge (TPAD). 
  • When the TPAD result returns, the physician who discharged the patient from the hospital receives an automated email message to his or her institutional email address. 
    • The email subject line reads “Important Post-Discharge Test Result”. 
    • Inside the email, the discharging physician is assigned responsibility for the follow-up. 
    • The results of the test are also contained within the body of the email. 
    • If the patient’s primary care provider (PCP) is in-network (the physician belongs to a practice that is part of the health plan’s covered network), he or she is carbon copied (cc’d) on the email. 
    • If the PCP is not in-network, the discharging physician can elect to contact the primary care provider and is able to document this action in the EHR. 
      • Note: out-of-network PCPs were mailed discharge summaries within 48 hours of discharge. 
  • When the discharging physician wants to communicate with the PCP, they can “reply all” to the email notification, provide a summary of the patient’s situation, and assign responsibility for further action to the PCP, if desired. 

Outcomes

  • Percentage of actionable TPADs that received follow-up: the percentage of “actionable” (actionability was determined by blinded physicians in the study) TPADs that received follow-up in the intervention group was 60.7% compared to 56.3% in the usual care group. This result was not statistically significant. 
    • In a subgroup analysis of in-network providers only, however, the intervention group had a statistically significantly higher proportion of follow-up action taken (70% vs 50%).
  • Impact on readmissions: While there were fewer readmissions in the intervention group compared to usual care (22.8% vs 31.1%), this result was not statistically significant. 
  • Impact on time-to-action: The median time to taking follow-up action in the intervention group was 9 days compared to 14 in the usual care, which was statistically significant. 
  • Type of follow-up performed: Of follow-up activities performed by physicians, the most common actions were changing, updating, or confirming a diagnosis as well as relaying results to patients.

Future Outcomes

  • Evaluating the impact of notifying patients, in addition to discharge attending and primary care physicians

 

Benefits

  • Shorter time to follow-up actions: The intervention group physicians completed follow-up actions faster than the control group physicians, which has potential benefit for patients who will receive important updates and changes to care sooner. 
  • Improved communication between physicians: By allowing for a “reply all” feature on the emails sent to hospital physicians and PCPs, the authors facilitated a new, clear line of communication between these providers. Hopefully, this will spark conversations and improve continuity of care for patients. 
  • Potential to improve follow-up: Within analysis was restricted to in-network PCPs, the proportion of patients who received follow-up actions based on their TPAD results was significantly higher in the intervention group. 
  • This suggests that should the intervention be improved such that out-of-network PCPs can be included on emails, the email system may improve follow-up to patients and lead to better patient outcomes.

Intervention-Specific Challenges

  • Inability to include out-of-network PCPs on emails: Since ⅔ of patients in the study had out-of-network PCPs, the intervention had a weaker effect and prevented the tool from improving communication between physicians. 
      • Disjointed transition of care between health systems can lead to fragmented care for patients, and the authors did not address this problem in the intervention. 
  • Lack of interoperability of EHRs: The out-of-network PCPs had to receive mailed copies of discharge reports up to 48 hours after discharge, which hinders efficient and rapid transfer of care between hospital physicians and PCPs, even if the emails prompted quick action.
  • Co-interventions rolled out simultaneously: While the study occurred, two other, similar interventions were being introduced in the hospital, including pathology and radiology test result alert systems. This made detection of differences between intervention and control groups more difficult, or may have contributed to alert fatigue in providers that confounded study results. 
    • This raises the larger issue of “alert fatigue”: as quality improvement efforts become widespread within health systems and often include added EHR notifications, physicians may be less likely to engage with the interventions. 

Sources

  • Callen J, Georgiou A, Li J, et al The safety implications of missed test results for hospitalised patients: a systematic review. BMJ Quality & Safety 2011;20:194-199.
  • Dalal AK, Schaffer A, Gershanik EF, Papanna R, Eibensteiner K, Nolido NV, Yoon CS, Williams D, Lipsitz SR, Roy CL, Schnipper JL. The Impact of Automated Notification on Follow-up of Actionable Tests Pending at Discharge: a Cluster-Randomized Controlled Trial. J Gen Intern Med. 2018 Jul;33(7):1043-1051. doi: 10.1007/s11606-018-4393-y. Epub 2018 Mar 12. PubMed PMID: 29532297; PubMed Central PMCID: PMC6025668.
  • Dalal AK, Schnipper JL, Poon EG, Williams DH, Rossi-Roh K, Macleay A, Liang CL, Nolido N, Budris J, Bates DW, Roy CL. Design and implementation of an automated email notification system for results of tests pending at discharge. J Am Med Inform Assoc. 2012 Jul-Aug;19(4):523-8. doi: 10.1136/amiajnl-2011-000615. Epub 2012 Jan 19. PubMed PMID: 22268214; PubMed Central PMCID: PMC3384118.
  • Forster AJ, Murff HJ, Peterson JF, Gandhi TK, Bates DW. The Incidence and Severity of Adverse Events Affecting Patients after Discharge from the Hospital. Ann Intern Med. 2003;138:161–167. doi: 10.7326/0003-4819-138-3-200302040-00007
  • Roy CL, Poon EG, Karson AS, Ladak-Merchant Z, Johnson RE, Maviglia SM, et al. Patient Safety Concerns Arising from Test Results That Return after Hospital Discharge. Ann Intern Med. 2005;143:121–128. doi: 10.7326/0003-4819-143-2-200507190-00011

Innovators
  • Anuj Dalal, MD
  • Jeff Schnipper, MD MPH
Talk to the Innovators