Opportunities to improve clinical summaries for patients at hospital discharge

The impact of discharge letter content on unplanned hospital readmissions within 30 and 90 days in older adults with chronic illness - a mixed methods study

BACKGROUND

Care transitions are high-risk processes, especially for people with complex or chronic illness. Discharge letters are an opportunity to provide written information to improve patients' self-management after discharge. The aim of this study is to determine the impact of discharge letter content on unplanned hospital readmissions and self-rated quality of care transitions among patients 60 years of age or older with chronic illness.

METHODS

The study had a convergent mixed methods design. Patients with chronic obstructive pulmonary disease or congestive heart failure were recruited from two hospitals in Region Stockholm if they were living at home and Swedish-speaking. Patients with dementia or cognitive impairment, or a "do not resuscitate" statement in their medical record were excluded. Discharge letters from 136 patients recruited to a randomised controlled trial were coded using an assessment matrix and deductive content analysis. The assessment matrix was based on a literature review performed to identify key elements in discharge letters that facilitate a safe care transition to home. The coded key elements were transformed into a quantitative variable of "SAFE-D score". Bivariate correlations between SAFE-D score and quality of care transition as well as unplanned readmissions within 30 and 90 days were calculated. Lastly, a multivariable Cox proportional hazards model was used to investigate associations between SAFE-D score and time to readmission.

RESULTS

All discharge letters contained at least five of eleven key elements. In less than two per cent of the discharge letters, all eleven key elements were present. Neither SAFE-D score, nor single key elements correlated with 30-day or 90-day readmission rate. SAFE-D score was not associated with time to readmission when adjusted for a range of patient characteristics and self-rated quality of care transitions.

CONCLUSIONS

While written summaries play a role, they may not be sufficient on their own to ensure safe care transitions and effective self-care management post-discharge.

TRIAL REGISTRATION

Clinical Trials. giv, NCT02823795, 01/09/2016.

Improving Resident Hospital Discharge Communication by Changing Electronic Health Record Templates to Enhance Primary Care Provider Satisfaction

BACKGROUND AND OBJECTIVES

Despite use of standardized electronic health record templates, the structure of discharge summaries may hinder communication from inpatient settings to primary care providers (PCPs). We developed an enhanced electronic discharge summary template to improve PCP satisfaction with written discharge summaries targeting diagnoses, medication reconciliation, laboratory test results, specialist follow-up, and recommendations.

METHODS

Resident template usage was measured using statistical process control charts. PCP reviewers' discharge summary satisfaction was surveyed using 5-point Likert scales analyzed using the Mann-Whitney U test. Residents were surveyed for satisfaction.

RESULTS

Resident template usage increased from 61% initially to 72% of discharge summaries at 6 months. The PCP reviewers reported increased satisfaction for summaries using the template compared with those without (4.3 vs 3.9, P = .003). Surveyed residents desired template inclusion in the default electronic discharge summary (93%).

CONCLUSIONS

This system-level resident-initiated quality improvement initiative created a novel discharge summary template that achieved widespread usage among residents and significantly increased outpatient PCP satisfaction.

AI-Assisted Summarization of Radiologic Reports: Evaluating GPT3davinci, BARTcnn, LongT5booksum, LEDbooksum, LEDlegal, and LEDclinical

BACKGROUND AND PURPOSE

The review of clinical reports is an essential part of monitoring disease progression. Synthesizing multiple imaging reports is also important for clinical decisions. It is critical to aggregate information quickly and accurately. Machine learning natural language processing (NLP) models hold promise to address an unmet need for report summarization.

MATERIALS AND METHODS

We evaluated NLP methods to summarize longitudinal aneurysm reports. A total of 137 clinical reports and 100 PubMed case reports were used in this study. Models were 1) compared against expert-generated summary using longitudinal imaging notes collected in our institute and 2) compared using publicly accessible PubMed case reports. Five AI models were used to summarize the clinical reports, and a sixth model, the online GPT3davinci NLP large language model (LLM), was added for the summarization of PubMed case reports. We assessed the summary quality through comparison with expert summaries using quantitative metrics and quality reviews by experts.

RESULTS

In clinical summarization, BARTcnn had the best performance (BERTscore = 0.8371), followed by LongT5Booksum and LEDlegal. In the analysis using PubMed case reports, GPT3davinci demonstrated the best performance, followed by models BARTcnn and then LEDbooksum (BERTscore = 0.894, 0.872, and 0.867, respectively).

CONCLUSIONS

AI NLP summarization models demonstrated great potential in summarizing longitudinal aneurysm reports, though none yet reached the level of quality for clinical usage. We found the online GPT LLM outperformed the others; however, the BARTcnn model is potentially more useful because it can be implemented on-site. Future work to improve summarization, address other types of neuroimaging reports, and develop structured reports may allow NLP models to ease clinical workflow.

The Chinese Version of the Patient Education Materials Assessment Tool for Printable Materials: Translation, Adaptation, and Validation Study

BACKGROUND

Providing people with understandable and actionable health information can considerably promote healthy behaviors and outcomes. To this end, some valid and reliable scales assessing the patient-friendliness of health education materials, like the PEMAT-P (Patient Education Materials Assessment Tool for printable materials), have been well developed in English-speaking countries. However, the English version of the PEMAT-P has not been translated and adapted into simplified Chinese and validated in mainland China.

OBJECTIVE

This study sought to translate the PEMAT-P tool into a simplified Chinese (Mandarin) version (C-PEMAT-P, a Chinese version of the Patient Education Materials Assessment Tool for printable materials) and verify its validity and reliability for assessing the comprehensibility and actionability of health education resources written in simplified Chinese. As a result, the validated C-PEMAT-P could be used to guide health researchers and educators to design more comprehensible and actionable materials for more tailored and targeted health education and interventions.

METHODS

We translated the PEMAT-P into simplified Chinese in the following three steps: (1) forward-translating the PEMAT-P into simplified Chinese, (2) back-translating the simplified Chinese version into English, and (3) testing translation equivalence linguistically and culturally by examining the original English version of the PEMAT-P and the back-translated English version of the tool. Any discrepancies between the original English tool and the back-translated English tool were resolved through a panel discussion among the research team of all authors to produce a revised forward-translated Chinese version (C-PEMAT-P). We then evaluated the clarity of construction and wording as well as the content relevance of the C-PEMAT-P using a 4-point ordinal scale to determine its content validity. After that, 2 native Chinese speakers (health educators) used the C-PEMAT-P to rate 15 health education handouts concerning air pollution and health to validate their reliability. We calculated the Cohen coefficient and Cronbach α to determine the interrater agreement and internal consistency of the C-PEMAT-P, respectively.

RESULTS

We finalized the translated Chinese tool after discussing the differences between the 2 English versions (original and back-translated) of the PEMAT-P, producing the final Chinese version of the PEMAT-P (C-PEMAT-P). The content validity index of the C-PEMAT-P version was 0.969, the Cohen coefficient for the interrater scoring agreement was 0.928, and the Cronbach α for internal consistency was .897. These values indicated the high validity and reliability of the C-PEMAT-P.

CONCLUSIONS

The C-PEMAT-P has been proven valid and reliable. It is the first Chinese scale for assessing the comprehensibility and actionability of Chinese health education materials. It can be used as an assessment tool to evaluate health education materials currently available and a guide to help health researchers and educators design more comprehensible and actionable materials for more tailored and targeted health education and interventions.

Patient and clinician perceptions of the trauma and acute care surgery hospitalization discharge transition of care: a qualitative study

OBJECTIVES

Trauma and acute care surgery (TACS) patients face complex barriers associated with hospitalization discharge that hinder successful recovery. We sought to better understand the challenges in the discharge transition of care, which might suggest interventions that would optimize it.

METHODS

We conducted a qualitative study of patient and clinician perceptions about the hospital discharge process at an urban level 1 trauma center. We performed semi-structured interviews that we recorded, transcribed, coded both deductively and inductively, and analyzed thematically. We enrolled patients and clinicians until we achieved data saturation.

RESULTS

We interviewed 10 patients and 10 clinicians. Most patients (70%) were male, and the mean age was 57±16 years. Clinicians included attending surgeons, residents, nurse practitioners, nurses, and case managers. Three themes emerged. (1) Communication (patient-clinician and clinician-clinician): clinicians understood that the discharge process malfunctions when communication with patients is not clear. Many patients discussed confusion about their discharge plan. Clinicians lamented that poorly written discharge summaries are an inadequate means of communication between inpatient and outpatient clinicians. (2) Discharge teaching and written instructions: patients appreciated discharge teaching but found written discharge instructions to be overwhelming and unhelpful. Clinicians preferred spending more time teaching patients and understood that written instructions contain too much jargon. (3) Outpatient care coordination: patients and clinicians identified difficulties with coordinating ongoing outpatient care. Both identified the patient's primary care physician and insurance coverage as important determinants of the outpatient experience.

CONCLUSION

TACS patients face numerous challenges at hospitalization discharge. Clinicians struggle to effectively help their patients with this stressful transition. Future interventions should focus on improving communication with patients, active communication with a patient's primary care physician, repurposing, and standardizing the discharge summary to serve primarily as a means of care coordination, and assisting the patient with navigating the transition.

LEVEL OF EVIDENCE

III-descriptive, exploratory study.

Leveraging Electronic Health Records to Address Breast Cancer Disparities

Purpose of Review

Breast cancer is the most commonly diagnosed cancer in women, and the leading cause of cancer death. However, racial and ethnic minority groups, as well as rural and underserved populations, face disparities that limit their access to specialty care for breast cancer. To address these disparities, health care providers can leverage an electronic health record (EHR).

Recent Findings

Few studies have evaluated the potential benefits of using EHRs to address breast cancer disparities, and none of them outlines a standard approach for this effort. However, these studies outline that EHRs can be used to identify and notify patients at risk for breast cancer. These systems can also automate referrals and scheduling for screening and genetic testing, as well as recruit eligible patients for clinical trials. EHRs can also provide educational materials to reduce risks associated with modifiable risk factors, such as physical activity, obesity, and smoking. These systems can also support telemedicine visits and centralize inter-institutional communication to improve treatment adherence and the quality of care.

Summary

EHRs have tremendous potential to increase accessibility and communication for patients with breast cancer by augmenting patient engagement, improving communication between patients and providers, and strengthening communication among providers. These efforts can reduce breast cancer disparities by increasing breast cancer screening, improving treatment adherence, expanding access to specialty care, and promoting risk-reducing habits among racial and ethnic minority groups and other underserved populations.

Development and Implementation of a Multidisciplinary Electronic Discharge Readiness Tool: User-Centered Design Approach

BACKGROUND

Typical solutions for improving discharge planning often rely on one-way communication mechanisms, static data entry into the electronic health record (EHR), or in-person meetings. Lack of timely and effective communication can adversely affect patients and their care teams.

OBJECTIVE

Applying robust user-centered design strategies, we aimed to design an innovative EHR-based discharge readiness communication tool (the Discharge Today tool) to enable care teams to communicate any barriers to discharge, the status of patient discharge readiness, and patient discharge needs in real time across hospital settings.

METHODS

We employed multiple user-centered design strategies, including exploration of the current state for documenting discharge readiness and directing discharge planning, iterative low-fidelity prototypes, multidisciplinary stakeholder meetings, a brainwriting premortem exercise, and preproduction user testing. We iteratively collected feedback from users via meetings and surveys.

RESULTS

We conducted 28 meetings with 20 different stakeholder groups. From these stakeholder meetings, we developed 14 low-fidelity prototypes prior to deploying the Discharge Today tool for our pilot study. During the pilot study, stakeholders requested 46 modifications, of which 25 (54%) were successfully executed. We found that most providers who responded to the survey reported that the tool either saved time or did not change the amount of time required to complete their discharge workflow (21/24, 88%). Responses to open-ended questions offered both positive feedback and opportunities for improvement in the domains of efficiency, integration into workflow, avoidance of redundancies, expedited communication, and patient-centeredness.

CONCLUSIONS

Survey data suggest that this electronic discharge readiness tool has been successfully adopted by providers and clinical staff. Frequent stakeholder engagement and iterative user-centered design were critical to the successful implementation of this tool.

When text simplification is not enough: could a graph-based visualization facilitate consumers' comprehension of dietary supplement information?

Objective

Dietary supplements are widely used. However, dietary supplements are not always safe. For example, an estimated 23 000 emergency room visits every year in the United States were attributed to adverse events related to dietary supplement use. With the rapid development of the Internet, consumers usually seek health information including dietary supplement information online. To help consumers access quality online dietary supplement information, we have identified trustworthy dietary supplement information sources and built an evidence–based knowledge base of dietary supplement information—the integrated DIetary Supplement Knowledge base (iDISK) that integrates and standardizes dietary supplement related information across these different sources. However, as information in iDISK was collected from scientific sources, the complex medical jargon is a barrier for consumers’ comprehension. The objective of this study is to assess how different approaches to simplify and represent dietary supplement information from iDISK will affect lay consumers’ comprehension.

Materials and Methods

Using a crowdsourcing platform, we recruited participants to read dietary supplement information in 4 different representations from iDISK: (1) original text, (2) syntactic and lexical text simplification (TS), (3) manual TS, and (4) a graph–based visualization. We then assessed how the different simplification and representation strategies affected consumers’ comprehension of dietary supplement information in terms of accuracy and response time to a set of comprehension questions.

Results

With responses from 690 qualified participants, our experiments confirmed that the manual approach, as expected, had the best performance for both accuracy and response time to the comprehension questions, while the graph–based approach ranked the second outperforming other representations. In some cases, the graph–based representation outperformed the manual approach in terms of response time.

Conclusions

A hybrid approach that combines text and graph–based representations might be needed to accommodate consumers’ different information needs and information seeking behavior.

A STEEEP Hill to Climb: A Scoping Review of Assessments of Individual Hospitalist Performance

BACKGROUND

Although ensuring high-quality care requires assessment of individual hospitalist performance, current assessment approaches lack consistency and coherence. The Institute of Medicine's STEEEP framework for quality healthcare conceptualizes quality through domains of "Safe," "Timely," "Effective," "Efficient," "Equitable," and "Patient Centered." This framework may be applicable to assessing individual hospitalists.

OBJECTIVE

This scoping review sought to identify studies that describe variation in individual hospitalist performance and to code this data to the domains of the STEEEP framework.

METHODS

Via a systematic search of peer-reviewed literature that assessed the performance of individual hospitalists in the Medline database, we identified studies that described measurement of individual hospitalist performance. Forty-two studies were included in the final review and coded into one or more domains of the STEEEP framework.

RESULTS

Studies in the Safe domain focused on transitions of care, both at discharge and within the hospital. Many studies were coded to more than one domain, especially Timely, Effective, and Efficient. Examples include adherence to evidence-based guidelines or Choosing Wisely recommendations. The Patient Centered domain was most frequently coded, but approaches were heterogeneous. No included studies addressed the domain Equitable.

CONCLUSIONS

Applying the STEEEP framework to the published literature on assessment of individual hospitalist performance revealed strengths and weaknesses. Areas of strength were assessments of transitions of care and application of consensus guidelines. Other areas, such as equity and some components of safe practice, need development. All domains would benefit from more practical approaches. These findings should stimulate future work on feasibility of multidimensional assessment approaches.

Attitudes of physicians towards target groups and content of the discharge summary: a cross-sectional analysis in Styria, Austria

OBJECTIVES

The discharge summary (DS) represents one of the most important instruments to ensure a safe patient discharge from the hospital. They sometimes have poor quality in content and often include medical jargon, which the patient and their relatives cannot easily understand. Therefore, many risks for patient safety exist. This study investigated the questions for whom the DS is and which contents are necessary to ensure a safe treatment.

DESIGN

Cross-sectional analysis.

SETTING

Styria, Austria.

PARTICIPANTS

3948 internal and external physicians were consulted.

INTERVENTIONS

An online survey consisting of 24 questions was conducted. The survey was distributed to physicians working in the province of Styria, Austria, in 2018 over a period of 6 months.

MAIN OUTCOMES AND MEASURES

Attitudes of internal and external physicians in terms of target group, content and health literacy.

RESULTS

In total, 1060 physicians participated in the survey. The DS is considered as a communication tool among physicians (97.9%) and the patients are also indicated as addressees (73.5%). Furthermore, there is a high level of agreement that understandable information in the DS leads to fewer questions of the patients (67.9%).

CONCLUSION

In conclusion, the DS is not only seen as a document for the further treating physician but is also relevant for the patient. Incorporating the patient into their treatment at all levels may possibly strengthen the individual health literacy of the patient and their caring relatives.

Challenges optimizing the after visit summary

BACKGROUND

The after visit summary (AVS) is a paper or electronic document given to patients after a medical appointment, which is intended to summarize patients' health and guide future care, including self-management tasks.

OBJECTIVE

To describe experiences of health systems implementing a redesigned outpatient AVS in commercially available electronic health record (EHR) systems to inform future optimization.

MATERIALS AND METHODS

We conducted semi-structured interviews with information technology and clinical leaders at 12 hospital and community-based healthcare institutions across the continental United States focusing on the process of AVS redesign and implementation. We also report our experience implementing a redesigned AVS in the Epic EHR at the Mount Sinai Hospital in New York City, NY.

RESULTS

Health systems experienced many challenges implementing the redesigned AVS. While many IT leaders noted that the redesigned AVS is easier to understand and the document is better organized, they claim the effort is time-consuming, Epic system upgrades render AVS modifications non-functional, and primary care and specialty practices have different needs in regards to content and formatting. Our team was able to modify the document by changing the order of print groups, modifying the font size, bolding section headers, and inserting page breaks. Similar to other health systems, our team found that it is difficult to achieve some desired features due to limitations in the EHR platform.

CONCLUSION

Health IT leaders view the AVS as a valuable source of information for patients. However, limitations to AVS modifications in EHR systems present challenges to optimizing the tool. EHR vendors should incorporate learning from healthcare systems innovation efforts and consider building more flexibility into their product development.

A Heuristic Evaluation to Assess Use of After Visit Summaries for Supporting Continuity of Care

BACKGROUND

Outpatient providers often do not receive discharge summaries from acute care providers prior to follow-up visits. These outpatient providers may use the after-visit summaries (AVS) that are given to patients to obtain clinical information. It is unclear how effectively AVS support care coordination between clinicians.

OBJECTIVES

Goals for this effort include: (1) developing usability heuristics that may be applied both for assessment and to guide generation of medical documents in general, (2) conducting a heuristic evaluation to assess the use of AVS for communication between clinicians, and (3) providing recommendations for generating AVS that effectively support both patient/caregiver use and care coordination.

METHODS

We created a 17-item heuristic evaluation instrument for assessing usability of medical documents. Eight experts used the instrument to assess each of four simulated AVS. The simulations were created using examples from two hospitals and two pediatric patient cases developed by the National Institute of Standards and Technology.

RESULTS

Experts identified 224 unique usability problems ranging in severity from mild to catastrophic. Content issues (e.g., missing medical history, marital status of a 2-year-old) were rated as most severe, but widespread formatting and structural problems (e.g., inconsistent indentation, fonts, and headings; confusing ordering of information) were so distracting that they significantly reduced readers' ability to efficiently use the documents. Overall, issues in the AVS from Hospital 2 were more severe than those in the AVS from Hospital 1.

CONCLUSION

The new instrument allowed for quick, inexpensive evaluations of AVS. Usability issues such as unnecessary information, poor organization, missing information, and inconsistent formatting make it hard for patients, caregivers, and clinicians to use the AVS. The heuristics in the new instrument may be used as guidance to adapt electronic health record systems so that they generate more useful and usable medical documents.

Unsupervised ensemble ranking of terms in electronic health record notes based on their importance to patients

BACKGROUND

Allowing patients to access their own electronic health record (EHR) notes through online patient portals has the potential to improve patient-centered care. However, EHR notes contain abundant medical jargon that can be difficult for patients to comprehend. One way to help patients is to reduce information overload and help them focus on medical terms that matter most to them. Targeted education can then be developed to improve patient EHR comprehension and the quality of care.

OBJECTIVE

The aim of this work was to develop FIT (Finding Important Terms for patients), an unsupervised natural language processing (NLP) system that ranks medical terms in EHR notes based on their importance to patients.

METHODS

We built FIT on a new unsupervised ensemble ranking model derived from the biased random walk algorithm to combine heterogeneous information resources for ranking candidate terms from each EHR note. Specifically, FIT integrates four single views (rankers) for term importance: patient use of medical concepts, document-level term salience, word co-occurrence based term relatedness, and topic coherence. It also incorporates partial information of term importance as conveyed by terms' unfamiliarity levels and semantic types. We evaluated FIT on 90 expert-annotated EHR notes and used the four single-view rankers as baselines. In addition, we implemented three benchmark unsupervised ensemble ranking methods as strong baselines.

RESULTS

FIT achieved 0.885 AUC-ROC for ranking candidate terms from EHR notes to identify important terms. When including term identification, the performance of FIT for identifying important terms from EHR notes was 0.813 AUC-ROC. Both performance scores significantly exceeded the corresponding scores from the four single rankers (P<0.001). FIT also outperformed the three ensemble rankers for most metrics. Its performance is relatively insensitive to its parameter.

CONCLUSIONS

FIT can automatically identify EHR terms important to patients. It may help develop future interventions to improve quality of care. By using unsupervised learning as well as a robust and flexible framework for information fusion, FIT can be readily applied to other domains and applications.

Finding Important Terms for Patients in Their Electronic Health Records: A Learning-to-Rank Approach Using Expert Annotations

BACKGROUND

Many health organizations allow patients to access their own electronic health record (EHR) notes through online patient portals as a way to enhance patient-centered care. However, EHR notes are typically long and contain abundant medical jargon that can be difficult for patients to understand. In addition, many medical terms in patients' notes are not directly related to their health care needs. One way to help patients better comprehend their own notes is to reduce information overload and help them focus on medical terms that matter most to them. Interventions can then be developed by giving them targeted education to improve their EHR comprehension and the quality of care.

OBJECTIVE

We aimed to develop a supervised natural language processing (NLP) system called Finding impOrtant medical Concepts most Useful to patientS (FOCUS) that automatically identifies and ranks medical terms in EHR notes based on their importance to the patients.

METHODS

First, we built an expert-annotated corpus. For each EHR note, 2 physicians independently identified medical terms important to the patient. Using the physicians' agreement as the gold standard, we developed and evaluated FOCUS. FOCUS first identifies candidate terms from each EHR note using MetaMap and then ranks the terms using a support vector machine-based learn-to-rank algorithm. We explored rich learning features, including distributed word representation, Unified Medical Language System semantic type, topic features, and features derived from consumer health vocabulary. We compared FOCUS with 2 strong baseline NLP systems.

RESULTS

Physicians annotated 90 EHR notes and identified a mean of 9 (SD 5) important terms per note. The Cohen's kappa annotation agreement was .51. The 10-fold cross-validation results show that FOCUS achieved an area under the receiver operating characteristic curve (AUC-ROC) of 0.940 for ranking candidate terms from EHR notes to identify important terms. When including term identification, the performance of FOCUS for identifying important terms from EHR notes was 0.866 AUC-ROC. Both performance scores significantly exceeded the corresponding baseline system scores (P<.001). Rich learning features contributed to FOCUS's performance substantially.

CONCLUSIONS

FOCUS can automatically rank terms from EHR notes based on their importance to patients. It may help develop future interventions that improve quality of care.