A Framework for Integrating Telehealth Equitably across the cancer care continuum.

The COVID-19 pandemic placed a spotlight on the potential to dramatically increase the use of telehealth across the cancer care continuum, but whether and how telehealth can be implemented in practice in ways that reduce, rather than exacerbate, inequities are largely unknown. To help fill this critical gap in research and practice, we developed the Framework for Integrating Telehealth Equitably (FITE), a process and evaluation model designed to help guide equitable integration of telehealth into practice. In this manuscript, we present FITE and showcase how investigators across the National Cancer Institute's Telehealth Research Centers of Excellence are applying the framework in different ways to advance digital and health equity. By highlighting multilevel determinants of digital equity that span further than access alone, FITE highlights the complex and differential ways structural determinants restrict or enable digital equity at the individual and community level. As such, achieving digital equity will require strategies designed to not only support individual behavior but also change the broader context to ensure all patients and communities have the choice, opportunity, and resources to use telehealth across the cancer care continuum.

Identifying Nonpatient Authors of Patient Portal Secure Messages in Oncology: A Proof-of-Concept Demonstration of Natural Language Processing Methods.

PURPOSE: Patient portal secure messages are not always authored by the patient account holder. Understanding who authored the message is particularly important in an oncology setting where symptom reporting is crucial to patient treatment. Natural language processing has the potential to detect messages not authored by the patient automatically. METHODS: Patient portal secure messages from the Memorial Sloan Kettering Cancer Center were retrieved and manually annotated as a predicted unregistered proxy (ie, not written by the patient) or a presumed patient. After randomly splitting the annotated messages into training and test sets in a 70:30 ratio, a bag-of-words approach was used to extract features and then a Least Absolute Shrinkage and Selection Operator (LASSO) model was trained and used for classification. RESULTS: Portal secure messages (n = 2,000) were randomly selected from unique patient accounts and manually annotated. We excluded 335 messages from the data set as the annotators could not determine if they were written by a patient or proxy. Using the remaining 1,665 messages, a LASSO model was developed that achieved an area under the curve of 0.932 and an area under the precision recall curve of 0.748. The sensitivity and specificity related to classifying true-positive cases (predicted unregistered proxy-authored messages) and true negatives (presumed patient-authored messages) were 0.681 and 0.960, respectively. CONCLUSION: Our work demonstrates the feasibility of using unstructured, heterogenous patient portal secure messages to determine portal secure message authorship. Identifying patient authorship in real time can improve patient portal account security and can be used to improve the quality of the information extracted from the patient portal, such as patient-reported outcomes.

Managing Critical Patient-Reported Outcome Measures in Oncology Settings: System Development and Retrospective Study.

BACKGROUND: Remote monitoring programs based on the collection of patient-reported outcome (PRO) data are being increasingly adopted in oncology practices. Although PROs are a great source of patient data, the management of critical PRO data is not discussed in detail in the literature. OBJECTIVE: This first-of-its-kind study aimed to design, describe, and evaluate a closed-loop alerting and communication system focused on managing PRO-related alerts in cancer care. METHODS: We designed and developed a novel solution using an agile software development methodology by incrementally building new capabilities. We evaluated these new features using participatory design and the Fit between Individuals, Task, and Technology framework. RESULTS: A total of 8 questionnaires were implemented using alerting features, resulting in an alert rate of 7.82% (36,838/470,841) with 13.28% (10,965/82,544) of the patients triggering at least one alert. Alerts were reviewed by 501 staff members spanning across 191 care teams. All the alerts were reviewed with a median response time of 1 hour (SD 185 hours) during standard business hours. The most severe (red) alerts were documented 56.83% (2592/4561) of the time, whereas unlabeled alerts were documented 27.68% (1298/4689) of the time, signaling clinician concordance with the alert thresholds. CONCLUSIONS: A PRO-based alert and communication system has some initial benefits in reviewing clinically meaningful PRO data in a reasonable amount of time. We have discussed key system design considerations, workflow integration, and the mitigation of potential impact on the burden of care teams. The introduction of a PRO-based alert and communication system provides a reliable mechanism for care teams to review and respond to patient symptoms quickly. The system was standardized across many different oncology settings, demonstrating system flexibility. Future studies should focus on formally evaluating system usability through qualitative methods.

Ethics Consultation Requests After Implementation of an Electronic Health Record Order.

PURPOSE: As the role of clinical ethics consultation in health care advances, there are calls to standardize the process of consultation. The Ethics Committee at Memorial Sloan Kettering Cancer Center (MSK) hypothesized that the process of requesting an ethics consultation could be improved by instituting an electronic health record (EHR) order for consultation requests. This report summarizes the impact of adopting an EHR order for ethics consultation requests at MSK. METHODS: This retrospective review of all clinical ethics consultations requested at a tertiary cancer center from May 2017 to February 2020 spans 17 months before and after implementation of an electronic order for consultation requests. Summary statistics are presented using Pearson chi-square analyses with a significance level of 0.05. RESULTS: There was a significant increase in the total number of consultation requests placed after implementation of the EHR order (n = 165, 0.08% of total patients) compared with before (n = 108, 0.05% of total patients; P = .007). The number of consults requested by providers from inpatient (P = .02) and outpatient (P = .04) settings significantly increased. The proportion of consults placed by medical versus nonmedical providers remained unchanged (P = .32). CONCLUSION: In this large single-institution retrospective study, implementation of an EHR order for ethics consultation requests was associated with a significant increase in the number of consultation requests. Implementation of an electronic order may decrease barriers to ethics consultation in diverse practice settings. Further longitudinal, multicenter studies are needed to assess strategies to improve access to clinical ethics consultation for oncology patients.

Adoption of Patient-Generated Health Data in Oncology: A Report From the NCCN EHR Oncology Advisory Group.

BACKGROUND: Collecting, monitoring, and responding to patient-generated health data (PGHD) are associated with improved quality of life and patient satisfaction, and possibly with improved patient survival in oncology. However, the current state of adoption, types of PGHD collected, and degree of integration into electronic health records (EHRs) is unknown. METHODS: The NCCN EHR Oncology Advisory Group formed a Patient-Reported Outcomes (PRO) Workgroup to perform an assessment and provide recommendations for cancer centers, researchers, and EHR vendors to advance the collection and use of PGHD in oncology. The issues were evaluated via a survey of NCCN Member Institutions. Questions were designed to assess the current state of PGHD collection, including how, what, and where PGHD are collected. Additionally, detailed questions about governance and data integration into EHRs were asked. RESULTS: Of 28 Member Institutions surveyed, 23 responded. The collection and use of PGHD is widespread among NCCN Members Institutions (96%). Most centers (90%) embed at least some PGHD into the EHR, although challenges remain, as evidenced by 88% of respondents reporting the use of instruments not integrated. Forty-seven percent of respondents are leveraging PGHD for process automation and adherence to best evidence. Content type and integration touchpoints vary among the members, as well as governance maturity. CONCLUSIONS: The reported variability regarding PGHD suggests that it may not yet have reached its full potential for oncology care delivery. As the adoption of PGHD in oncology continues to expand, opportunities exist to enhance their utility. Among the recommendations for cancer centers is establishment of a governance process that includes patients. Researchers should consider determining which PGHD instruments confer the highest value. It is recommended that EHR vendors collaborate with cancer centers to develop solutions for the collection, interpretation, visualization, and use of PGHD.

Preferences in Oncology History Documentation Styles Among Clinical Practitioners.

PURPOSE: Clinical notes function as the de facto handoff between providers and assume great importance during unplanned medical encounters. An organized and thorough oncology history is essential in care coordination. We sought to understand reader preferences for oncology history organization by comparing between chronologic and narrative formats. METHODS: A convenience sample of 562 clinicians from 19 National Comprehensive Cancer Network Member Institutions responded to a survey comparing two formats of oncology histories, narrative and chronologic, for the same patient. Both histories were consensus-derived real-world examples. Each history was evaluated using semantic differential attributes (thorough, useful, organized, comprehensible, and succinct). Respondents choose a preference between the two styles for history gathering and as the basis of a new note. Open-ended responses were also solicited. RESULTS: Respondents preferred the chronologic over the narrative history to prepare for a visit with an unknown patient (66% preference) and as a basis for their own note preparation (77% preference) (P < .01). The chronologic summary was preferred in four of the five measured attributes (useful, organized, comprehensible, and succinct); the narrative summary was favored for thoroughness (P < .01). Open-ended responses reflected the attribute scoring and noted the utility of content describing social determinants of health in the narrative history. CONCLUSION: Respondents of this convenience sample preferred a chronologic oncology history to a concise narrative history. Further studies are needed to determine the optimal structure and content of chronologic documentation for oncology patients and the provider effort to use this format.

A Randomized Controlled Trial Evaluating Electronic Outpatient Symptom Monitoring After Ambulatory Cancer Surgery.

OBJECTIVE: We implemented routine daily electronic monitoring of patient-reported outcomes (PROs) for 10 days after discharge after ambulatory cancer surgery, with alerts to clinical staff for worrying symptoms. We sought to determine whether enhancing this monitoring by adding immediate automated normative feedback to patients regarding expected symptoms would further improve the patient experience. SUMMARY OF BACKGROUND DATA: PRO monitoring reduces symptom severity in cancer patients. In ambulatory cancer surgery, it reduces potentially avoidable urgent care center (UCC) visits, defined as those UCC visits without readmission. METHODS: Patients undergoing ambulatory cancer surgery (n = 2624) were randomized to receive standard PRO monitoring or enhanced feedback. The primary study outcome was UCC visits without readmission within 30 days; secondary outcomes included patient anxiety and nursing utilization. RESULTS: There was no significant difference in the risk of a potentially avoidable UCC visit [1.0% higher in enhanced feedback, 95% confidence interval (CI) -0.2-3.1%; P = 0.12]. There were similarly no significant differences in UCC visits with readmission or readmission overall (P = 0.4 for both). Patients randomized to enhanced feedback demonstrated a quicker reduction in anxiety (P < 0.001) and required 14% (95% CI 8-19%; P < 0.001) and 10% (95% CI 5-16%, P < 0.001) fewer nursing calls over 10 and 30 days postoperatively. CONCLUSIONS: Providing patients with feedback about symptom severity during recovery from ambulatory cancer surgery reduces anxiety and nursing workload without affecting UCC visits or readmissions. These results support wider incorporation of normative feedback in systems for routine PRO monitoring.

Interfaces for collecting data from patients: 10 golden rules.

Memorial Sloan Kettering Cancer Center has more than a decade's experience creating online interfaces for obtaining data from patients as part of routine clinical care. We have developed a set of "golden rules" for design of these interfaces. Many relate to the knowledge imbalance between professional staff (whether medical or informatics) and patients, who are often old and sick and have limited knowledge of technology. Others relate to the clinical nature of the encounter: data cannot be taken from patients as part of clinical care unless there is a plan to act on whatever information is prepared. We also note that the plethora of marketing questionnaires makes patients suspicious of surveys: patient trust is hard to gain and easy to lose. Addition of these golden rules to standard approaches to interface design will maximize our ability to obtain data from patients and thus improve communication between patients and clinicians.

Ambulatory cancer care electronic symptom self-reporting (ACCESS) for surgical patients: a randomised controlled trial protocol.

INTRODUCTION: An increasing proportion of cancer surgeries are ambulatory procedures requiring a stay of 1 day or less in the hospital. Providing patients and their caregivers with ongoing, real-time support after discharge aids delivery of high-quality postoperative care in this new healthcare environment. Despite abundant evidence that patient self-reporting of symptoms improves quality of care, the most effective way to monitor and manage this self-reported information is not known. METHODS AND ANALYSIS: This is a two-armed randomised, controlled trial evaluating two approaches to the management of patient-reported data: (1) team monitoring, symptom monitoring by the clinical team, with nursing outreach if symptoms exceed normal limits, and (2) enhanced feedback, real-time feedback to patients about expected symptom severity, with patient-activated care as needed.Patients with breast, gynaecologic, urologic, and head and neck cancer undergoing ambulatory cancer surgery (n=2750) complete an electronic survey for up to 30 days after surgery that includes items from a validated instrument developed by the National Cancer Institute, the Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE). Information provided to patients in the Enhanced Feedback group is procedure-specific and based on updated PRO-CTCAE data from previous patients. Qualitative interviews are also performed. The primary study outcomes assess unplanned emergency department visits and symptom-triggered interventions (eg, nursing calls and pain management referrals) within 30 days, and secondary outcomes assess the patient and caregiver experience (ie, patient engagement, patient anxiety and caregiver burden). ETHICS AND DISSEMINATION: This study is approved by the Institutional Review Board at Memorial Sloan Kettering Cancer Center. The relationships between the study team and stakeholders will be leveraged to disseminate study findings. Findings will be relevant in designing future coordinated care models targeting improved healthcare quality and patient experience. TRIAL REGISTRATION NUMBER: NCT03178045.

Real-World Outcomes of an Automated Physician Support System for Genome-Driven Oncology.

PURPOSE: Matching patients to investigational therapies requires new tools to support physician decision making. We designed and implemented Precision Insight Support Engine (PRECISE), an automated, just-in-time, clinical-grade informatics platform to identify and dynamically track patients on the basis of molecular and clinical criteria. Real-world use of this tool was analyzed to determine whether PRECISE facilitated enrollment to early-phase, genome-driven trials. MATERIALS AND METHODS: We analyzed patients who were enrolled in genome-driven, early-phase trials using PRECISE at Memorial Sloan Kettering Cancer Center between April 2014 and January 2018. Primary end point was the proportion of enrolled patients who were successfully identified using PRECISE before enrollment. Secondary end points included time from sequencing and PRECISE identification to enrollment. Reasons for a failure to identify genomically matched patients were also explored. RESULTS: Data were analyzed from 41 therapeutic trials led by 19 principal investigators. In total, 755 patients were accrued to these studies during the period that PRECISE was used. PRECISE successfully identified 327 patients (43%) before enrollment. Patients were diagnosed with 29 tumor types and harbored alterations in 43 oncogenes, most commonly ERBB2 (21.3%), PIK3CA (14.1%), and BRAF (8.7%). Median time from sequencing to enrollment was 163 days (interquartile range, 66 to 357 days), and from PRECISE identification to enrollment 87 days (interquartile range, 37 to 180 days). Common reasons for failing to identify patients before enrollment included accrual on the basis of molecular alterations that did not match pre-established PRECISE genomic eligibility (140 [33%] of 428) and external sequencing not available for parsing (127 [30%] of 428). CONCLUSION: PRECISE identified 43% of all patients accrued to a diverse cohort of early-phase, genome-matched studies. Purpose-built informatics platforms represent a novel and potentially effective method for matching patients to molecularly selected studies.

Informing, Reassuring, or Alarming? Balancing Patient Needs in the Development of a Postsurgical Symptom Reporting System in Cancer.

After ambulatory surgeries, patients who recover at home have multiple questions about wound healing, symptoms and medication side effects, and recovery expectations. We conducted user testing and rapid application development of a newly developed symptom reporting system that supports home-based recovery by inviting patients to self-report symptoms in the days after surgery and then receive an immediate feedback report giving context for their reported symptoms. Findings showed that some participants primarily valued reassurance, whereas others prioritized receiving alerts about potential problems. Results also showed that most patients wanted feedback framed as comparing their progress to their expected progress, not to that of other patients. The final feedback report provided patients with actionable recommendations, small graphs showing their progress, and with short "gist" text interpretations. The system has been implemented, and recruitment is ongoing for a large clinical trial of its effectiveness for reducing adverse events and unnecessary emergency or urgent care visits.

Automated eligibility screening and monitoring for genotype-driven precision oncology trials.

The Information Systems Department at Memorial Sloan Kettering Cancer Center developed the DARWIN Cohort Management System (DCMS). The DCMS identifies and tracks cohorts of patients based on genotypic and clinical data. It assists researchers and treating physicians in enrolling patients to genotype-matched IRB-approved clinical trials. The DCMS sends automated, actionable, and secure email notifications to users with information about eligible or enrolled patients before their upcoming appointments. The system also captures investigators input via annotations on patient eligibility and preferences on future status updates. As of August 2015, the DCMS is tracking 159,893 patients on both clinical operations and research cohorts. 134 research cohorts have been established and track 64,473 patients. 51,192 of these have had one or more genomic tests including MSK-IMPACT, comprising the pool eligible for genotype-matched studies. This paper describes the design and evolution of this Informatics solution.

Perioperative outcome and long-term mortality for heart failure patients undergoing intermediate- and high-risk noncardiac surgery: impact of left ventricular ejection fraction.

The impact of left ventricular ejection fraction (LVEF) on outcome in patients with heart failure (HF) undergoing noncardiac surgery has not been extensively evaluated. In this study, 174 patients (mean age, 75+/-12 years, 47% male, mean LVEF (47%+/-18%) underwent intermediate- or high-risk noncardiac surgery. Patients were stratified by LVEF, and adverse perioperative complications were identified and compared. Adverse perioperative events occurred in 53 patients (30.5%), including 14 (8.1%) deaths within 30 days, 26 (14.9%) myocardial infarctions, and 44 (25.3%) HF exacerbations. Among the factors associated with adverse perioperative outcomes in the first 30 days were advanced age (>80 years), diabetes, and a severely decreased LVEF (<30%). Long-term mortality was high, and Cox proportional hazards analysis demonstrated that LVEF was an independent risk factor for long-term mortality.

Use of semantic features to classify patient smoking status.

The recent i2b2 NLP Challenge smoking classification task offers a rare chance to compare different natural language processing techniques on actual clinical data. We compare the performance of a classifier which relies on semantic features generated by an unmodified version of MedLEE, a clinical NLP engine, to one using lexical features. We also compare the performance of supervised classifiers to rule-based symbolic classifiers. Our baseline supervised classifier with lexical features yields a microaveraged F-measure of 0.81. Our rule-based classifier using MedLEE semantic features is superior, with an F-measure of 0.83. Our supervised classifier trained with semantic MedLEE features is competitive with the top-performing smoking classifier in the i2b2 NLP Challenge, with microaveraged precision of 0.90, recall of 0.89, and F-measure of 0.89.

Approach to mobile information and communication for health care.

Evidence suggests that inadequate access to information and ineffective communication are proximal causes of errors and other adverse events in-patient care. Within the context of reducing these proximal causes of errors, we explore the use of novel information-based approaches to improve information access and communication in health care settings. This paper describes the approaches for and the design of extensions to a clinical information system used to improve information access and communication at the point of care using information-based handheld wireless applications. These extensions include clinical and information resources, event monitoring, and a virtual whiteboard (VWB).

Mining complex clinical data for patient safety research: a framework for event discovery.

Successfully addressing patient safety requires detecting medical events effectively. Given the volume of patients seen at medical centers, detecting events automatically from data that are already available electronically would greatly facilitate patient safety work. We have created a framework for electronic detection. Key steps include: selecting target events, assessing what information is available electronically, transforming raw data such as narrative notes into a coded format, querying the transformed data, verifying the accuracy of event detection, characterizing the events using systems and cognitive approaches, and using what is learned to improve detection.

Detecting adverse events using information technology.

CONTEXT: Although patient safety is a major problem, most health care organizations rely on spontaneous reporting, which detects only a small minority of adverse events. As a result, problems with safety have remained hidden. Chart review can detect adverse events in research settings, but it is too expensive for routine use. Information technology techniques can detect some adverse events in a timely and cost-effective way, in some cases early enough to prevent patient harm. OBJECTIVE: To review methodologies of detecting adverse events using information technology, reports of studies that used these techniques to detect adverse events, and study results for specific types of adverse events. DESIGN: Structured review. METHODOLOGY: English-language studies that reported using information technology to detect adverse events were identified using standard techniques. Only studies that contained original data were included. MAIN OUTCOME MEASURES: Adverse events, with specific focus on nosocomial infections, adverse drug events, and injurious falls. RESULTS: Tools such as event monitoring and natural language processing can inexpensively detect certain types of adverse events in clinical databases. These approaches already work well for some types of adverse events, including adverse drug events and nosocomial infections, and are in routine use in a few hospitals. In addition, it appears likely that these techniques will be adaptable in ways that allow detection of a broad array of adverse events, especially as more medical information becomes computerized. CONCLUSION: Computerized detection of adverse events will soon be practical on a widespread basis.