Arsoniadis, Elliot2021-02-222021-02-222020-12https://hdl.handle.net/11299/218738University of Minnesota Ph.D. dissertation. 2020. Major: Health Informatics. Advisor: Genevieve Melton-Meaux. 1 computer file (PDF); 122 pages.Handoff is the process by which the care of a patient is transferred from the responsibility of one provider (or team of providers) to another. Handoff that occurs between physicians in training, or resident physicians, has become the focus of numerous quality and safety initiatives, especially since the introduction of work-hour restrictions for resident physicians that have increased the number of handoffs taking place during a hospitalization. The Handoff Note is the (traditionally paper) artifact accompanying the Handoff Process, and in its most basic form consists of the names of patients being transferred from the care of one clinician to another. Other data is often included, such as demographic data, room number, and a brief summary. Increasingly, Handoff Notes are becoming integrated into the Electronic Health Record (EHR). However, there remains no universally accepted standard for EHR-Integrated Handoff Note content, nor standards for what content is appropriate for automatic entry from the EHR versus manual entry by providers. Further, there have been few efforts to elucidate resident physician preferences for Handoff Note content, structure, or format, despite being the principal users of these tools. Although many data elements can be automatically entered into the Handoff Note by the EHR, certain key elements will likely remain manually entered by clinicians as narrative text. Analysis of these free text elements in Handoff Notes may reveal important insights for informaticians, safety and quality experts, and those involved in graduate medical education. In the first part of this study we look to key stakeholders for standards surrounding optimal Handoff Note content, and after choosing one with greatest buy-in, compare it to content contained in individual EHR-Integrated Handoff Notes described in the literature. The chosen standard covered 67% of Content Headings described in the literature, and thirteen more unique Content Headings were found in the literature to add to this standard. Using these findings from the literature and guided by prior semi-structured interviews with resident physicians, we performed a large-scale survey on resident physician preferences for Handoff Note content, structure, and format. We found that some of the most important and trustworthy elements of the Handoff Note were narrative text data manually entered by other clinicians, including “Plan”, “Illness Severity”, and “Patient Summary”. Based on these insights, we then designed and implemented an EHR-Integrated Handoff Note within our institution. Years after implementation we found that nearly all “primary” service teams and many “consult” service teams continued to utilize the Handoff Note, including resident physicians is such different specialties as medicine, pediatrics, behavioral health, obstetrics/gynecology, neurology, surgery, and critical care. Analysis of the narrative text portions of the Handoff Note showed that “Patient Summary” and “To Do” text boxes were updated 1.0 and 1.6 times per day, respectively. The majority of these updates occurred between 12 pm – 5:59 pm, likely indicating preparation of the Handoff Note for the evening Handoff Process. However, many changes also take place between 6 am – 11:59 am, indicating possible use of the Handoff Note to aid team rounding activities. We also analyzed narrative text for errors, using progress notes and other data from the EHR as gold standard. We found at least one error in 65% of Handoff Notes. The majority of errors were related to the omission of key data, rather than the entry of incorrect data. Increased errors were found with increasing hospital day, as well as with authors in early stages (medical students, PGY-1 physicians) and later stages (>PGY-4 or attending physician) of training. While the integration of the Handoff Note within the EHR and the automatic entry of many data elements into the note will prove useful, manual entry of certain narrative text continues to be critical. Future work on the parts of informatics and usability experts should focus on ways to make composing and updating these notes easier and encourage accuracy and frequency of updates. Members of the graduate medical education community should also make it a priority to formalize training surrounding accurate and complete Handoff Note composition as an important adjunct to existing training surrounding the Handoff Process. These tools have the potential to greatly improve patient safety and quality of care.enClinical InformaticsElectronic Health RecordHandoffHealth InformaticsPatient SafetyThesis or Dissertation