Precision oncology: lessons learned and challenges for the future

The decreasing cost of and increasing capacity of DNA sequencing has led to vastly increased opportunities for population-level genomic studies to discover novel genomic alterations associated with both Mendelian and complex phenotypes. To translate genomic findings clinically, a number of health care institutions have worked collaboratively or individually to initiate precision medicine programs. These precision medicine programs involve designing patient enrollment systems, tracking electronic health records, building biobank repositories, and returning results with actionable matched therapies. As cancer is a paradigm for genetic diseases and new therapies are increasingly tailored to attack genetic susceptibilities in tumors, these precision medicine programs are largely driven by the urgent need to perform genetic profiling on cancer patients in real time. Here, we review the current landscape of precision oncology and highlight challenges to be overcome and examples of benefits to patients. Furthermore, we make suggestions to optimize future precision oncology programs based upon the lessons learned from these “first generation” early adopters.

The full article can be downloaded below.  

Getting Beyond Hype Vs Hope in Precision Medicine and AI: The Life Cycle Of Technology Revolutions

Powerful new technologies have the potential to radically transform both science and society. In science, as Douglas Robertson describes in Phase Change (2003), a new technology like the microscope, the telescope, and the calculus can profoundly alter the questions we ask, and advance our ability to better understand nature. Society, visibly, can also be transformed by technology, as we’ve seen with examples ranging from the steam engine and the telegraph to automation and the internet.

The catch is, this transformation doesn’t occur overnight – far from it.  The remarkable and often maddening aspect of innovation (as I’ve discussed here, here) is the exceptionally long time it takes between the time a technology is originally invented and the time when people figure out how to use it most effectively.

In this three-part piece, I will first present a framework, developed by economist Carlota Perez, describing the life cycle of transformative technologies, and outline relevant refinements, introduced by columnist Daniel Gross. I’ll then locate our contemporary debate around the utility (or not) of precision medicine – and particularly, precision oncology – in the context of this framework; this section is richly informed by the perspective shared by key physician and physician-scientist thought leaders in this space. Finally, I’ll suggest that AI (as a proxy for the emerging excitement – and skepticism — around digital and data in health) seems to be entering the earliest stages of the technology diffusion trajectory, which may help explain both the frenzy and the confusion.

The full Forbes article can be viewed at this link.  

Explainable AI In Health Care: Gaining Context Behind A Diagnosis

Most of the available health care diagnostics that use artificial intelligence (AI) function as black boxes—meaning that results do not include any explanation of why the machine thinks a patient has a certain disease or disorder. While AI technologies are extraordinarily powerful, adoption of these algorithms in health care has been slow because doctors and regulators cannot verify their results. However, a new type of algorithm called “explainable AI” (XAI) can be easily understood by humans. As a result, all signs point to XAI being rapidly adopted across health care, making it likely that providers will actually use the associated diagnostics.

The full Forbes article can be viewed at this link.  

Healthcare executives look to bring the joy back to medicine

As health systems survey candidates for executive roles, experience, academic achievements and other measures of scholarship typically top the priority list.

Leadership qualities, “coachability” and emotional intelligence are often overlooked, particularly at academic institutions. That represents a fundamental flaw in the hiring process and the medical education system, which hardwires doctors to be self-interested and autonomous, said Dr. Peter Pisters, president of the University of Texas MD Anderson Cancer Center.

“Our search committees have been anchored in a series of biases that include an overemphasis on scholarship and an underemphasis on leadership capabilities, especially emotional intelligence,” Pisters told the audience during Modern Healthcare’s Workplace of the Future conference last month. “That has created real challenges.”

MD Anderson has brought in psychologists specializing in emotional intelligence and incorporated that metric into the selection process. It mandates implicit bias training to ensure objectivity. Like many systems, the academic medical center has also pushed for more diverse leaders and has cast a wider net to attract more applicants.

The full Modern Healthcare article can be viewed at this link.  

Slides and recording from October 10, 2019 eHI webinar.

Using ICD-10-CM codes to capture social determinants of health (SDOH) data is an incredible opportunity to identify, document, and track factors impacting health, such as employment, food insecurity, and homelessness. This webinar featured the results of a recent collaboration between eHI and UnitedHealthcare, examining why ICD-10-CM codes are not being used to their full potential. As a result, the group developed a set of tools to promote the adoption and use of these codes by provider organizations and coding professionals.

Speakers from the collaboration will explain how to best use the tools, share what their organizations are doing to address SDOH, and answer questions such as:

• What is the benefit of standardizing the capture of SDOH data?
• What is a Z code for SDOH?
• Can coding professionals use non-physician documentation to support ICD-10 CM coding for societal and environmental conditions?
• Are there guidelines for using ICD-10 codes for SDOH?

Speakers:

Caraline Coats, MHSA, Vice President Bold Goal and Population Health Strategy, Humana
Caraline Coats is Vice President of Humana’s Bold Goal and Population Health Strategy, leading Humana’s mission to help improve the health of the communities it serves by making it easier for people to achieve their best health. Coats has been with Humana for over twelve years. She started as a Regional Director of Medicare Operations in Arizona and relocated to Florida, where she became the Vice President of Network Management and subsequently, the Regional Vice President of Network Management for the East Region. In her role before joining the Bold Goal team, Coats served   as Vice President of Humana’s Value-Based Strategies, leading the organizational advancement of innovative payment models that enable Humana to support providers as population health managers in value-based care relationships. Prior to Humana, Caraline was Vice President of Operations with a hospitalist company and Assistant Vice President of Managed Care for IASIS Healthcare in the Arizona and Nevada regions. She credits her understanding and experience working directly with physicians and hospitals for the opportunities she has had with Humana. Caraline holds an undergraduate degree in biology and a Masters in Health Services Administration from the University of Michigan. She and her husband have two sons, Michael (5 years old) and Nicholas (4 years old). Outside of Humana, Caraline spends her time with family and enjoys running.

Nelly Leon-Chisen, RHIA, Director, Coding and Classification, American Hospital Association
Nelly Leon-Chisen, RHIA, is the Director of the Coding and Classification at the American Hospital Association where she is responsible for leading the Central Office on ICD-10-CM/PCS and HCPCS. The Central Office, in cooperation with the National Center for Health Statistics (NCHS), the Centers for Medicare and Medicaid Services (CMS) and the American Health Information Management Association (AHIMA), serves as the authoritative source on ICD-10-CM/PCS. She represents the AHA as one of the four Cooperating Parties responsible for the development of the ICD-10-CM and ICD-10-PCS Official Guidelines for Coding and Reporting. She is also the executive editor of the AHA Coding Clinic publications. She is the author of the ICD-10-CM and ICD-10-PCS Coding Handbook published by AHA Press. She has over 30 years of experience in the health information management field including consulting, teaching, technical and management experience in hospital health information management departments. She has lectured extensively on coding, DRG and data quality issues throughout the United States, Europe, Asia and Latin America. She is a Past President of the Chicago Area Health Information Management Association and a recipient of the Professional Achievement Award from the Illinois Health Information Management Association. 

Sheila Shapiro, Senior Vice President, Strategic Community Partnerships, UnitedHealthcare
Sheila Shapiro joined UnitedHealthcare Community Plan of Arizona in 2009. Since joining the organization, she has held the positions Chief Operations Officer, Arizona Community and State, Plan President Washington Community &State, National Vice President of myConnections and National Vice President of Population Health and Clinical Innovation. Sheila has over 30 years of experience in the health care industry. In her current position as Senior Vice President, National Strategic Partnerships with United Healthcare, Sheila is responsible for the development, advancement and implementation of the industry-leading sustainable model to standardize and use non-traditional data elements and innovative strategic partnerships to improve health outcomes at scale. Prior to joining UnitedHealthcare, Shapiro held executive positions with Blue Cross Blue Shield of Montana, Molina and Premera Blue Cross. She has led a broad range of operations and strategic objectives including claims, customer service and financial operations. Shapiro earned a Master of Arts in management from the University of Phoenix, and her bachelor’s degree is from Arizona State University.  She also holds a financial management certificate from the Wharton School of Business.  Shapiro is an Arizona Women in Business honoree and has served as vice chair on the board of directors for the March of Dimes.  

Social Determinants of Health and ICD-10-CM Coding Resources

Download this information as 2-page document at the bottom of your screen.

• What is an ICD-10-CM code?
  International Classification of Diseases, Tenth Revision, Clinical Modification coding, known as ICD-10-CM coding, is a system used by clinicians to classify and record all diagnoses and symptoms for care within the United States. Codes are based on the International Classification of Diseases, which is published by the World Health Organization (WHO), using unique alphanumeric codes to identify known diseases and other health problems. ICD-10-CM codes provide a level of detail that is necessary for storing and retrieving diagnostic information, compiling national mortality and morbidity statistics, and processing health insurance claims.

   

• What is an ICD-10-CM Z code for Social Determinants of Health (SDOH)?
  ICD-10-CM codes include a category called Z codes, which are used to describe experiences, circumstances, or problems that affect patient health, but are not considered a specific disease or injury. Z codes identify patients facing socioeconomic and psychosocial circumstances that may influence their health status and contact with health services. Currently, codes included in categories Z55-Z65 document patients’ SDOH in a standardized manner.

   

• How does standardizing the capture of SDOH data codes benefit population health?
  Traditionally, data recorded during a patient visit directly relates to a patient’s health but does not incorporate outside factors that can impact well-being. SDOH data captures information at a level traditional health data sources cannot, and ICD-10-CM Z codes can record this information, giving deeper insights into factors impacting health, such as employment, food insecurity, and housing. Standardizing SDOH would assist in identifying, documenting, and tracking additional markers of health, beyond the physical, and would permit clinicians, hospitals, and health plans to share the information through medical records and insurance claims data.

   

• Are there guidelines for using ICD-10-CM codes for SDOH?
  ICD-10-CM diagnosis codes have been adopted under the Health Insurance Portability and Accountability Act (HIPAA) for all healthcare settings. Guidelines for Z codes are included in the Centers for Medicare & Medicaid Services (CMS) ICD-10-CM Official Guidelines for Coding and Reporting for FY 2020. https://www.cdc.gov/nchs/data/icd/10cmguidelines-FY2020_final.pdf

   

• Why should providers, non-physician healthcare providers, and coders use ICD-10-CM Z codes for SDOH?
  Utilizing Z codes for SDOH enables hospitals and health systems to better track patient needs and identify solutions to improve the health of their communities. The extraction of SDOH data from the Electronic Health Record (EHR) for clinical, operational, and research purposes can facilitate tracking, identification, and referrals to social and governmental services. Rather than a new system or new tool to capture SDOH, leveraging existing ICD-10-CM codes offers an opportunity to expand on the existing system. This practical application brings SDOH into a clinician’s workflow and becomes a part of the patient’s electronic medical record and claims history.

   

• What are the limitations of ICD-10-CM Z codes for SDOH?
  Currently, Z codes for SDOH capture some, but not all, domains of SDOH. Stakeholder groups have requested that the ICD-10 Coordination and Maintenance Committee expand the codes to represent more granular information that would inform more precise, effective, and efficient social interventions, such as “barrier situations” which prevent consumers from obtaining medications and routine and preventive care. Although coding for SDOH is not mandated, when there is documentation of SDOH in the patient’s notes, it is still possible to use Z codes in the same manner that medical coding is done. Coding professionals may not know to scan for SDOH or may be hesitant to use the codes. Additionally, if a code has not been developed for a specific SDOH issue, the issue will not be coded and will not be included in the patient’s overall plan of care, nor as part of the claim submission process, unless it is recorded as narrative text.

   

• Are coding professionals allowed to use non-physician documentation to support ICD-10-CM coding for societal and environmental conditions?
  Yes, coding professionals at hospitals and health systems can report these codes based on documentation by all clinicians involved in the care of patients, such as case managers, discharge planners, social workers and nurses. In early 2018, the American Hospital Association’s (AHA) Coding Clinic published guidance that allows the reporting of SDOH ICD-10 codes based on non-physician documentation. The ICD-10-CM Cooperating Parties approved the advice, with the change effective February 2018.

   

• Where can I find more resources and initiatives around SDOH data and ICD-10-CM Coding?
  • American Hospital Association (AHA) Information sheet on ICD-10 Coding for Social Determinants of Health
  • "Health Disparities: Social Determinants of Health)” (Free online education module by the American Medical Association (AMA)
  • AMA’s Integrated Health Model Initiative (IHMI), Resource Link & Link 2
  • AMA and UnitedHealthcare’s collaboration to support the creation of new ICD-10 codes related to SDOH
  • Centers for Disease Control (CDC) Social Determinants of Health Webpage
  • CMS 2020 ICD-10-CM Files
  • CMS ICD-10-CM Official Guidelines for Coding and Reporting for FY 2020
  • CMS’ Accountable Health Communities Health-Related Social Needs Screening Tool & FAQ
  • Healthy People 2020’s Social Determinants of Health Webpage
  • Humana Food Insecurity and Loneliness Toolkits
  • The National Association of Community Health Centers (NACHC) Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences (PRAPARE)
  • American Academy of Family Physicians (AAFP)’s EveryONE Project Toolkit
  • SIREN’s (Social Interventions Research and Evaluation Network) Gravity Project Resource Link & Link 2
  • NCQA Population Health Management Resource Guide

In Summer of 2019, eHealth Initiative and UnitedHealthcare’s National Strategic Partnerships Division convened a collaborative meeting of leaders from payer organizations and other stakeholder groups to address the use of ICD-10-CM codes for capturing social determinants of health (SDOH) data. This meeting marked a significant milestone in the shift to value-based care. Despite the competitive nature of healthcare, the private sector is working together to address factors pertinent to patient care and well-being in a sustainable, scalable manner. The group discussed the need for better education of provider and billing coders on the value of collecting and using SDOH data and identified strategies to accomplish this task:

• Develop a consistent and unified approach to communicate and assist providers and coders in utilizing existing ICD-10-CM codes for SDOH.
• Formulate a strategy and unified approach for providers and coders to assist with adoption and utilization of the proposed SDOH codes once approved.

Attendees agreed that the best strategy to communicate and promote the adoption of ICD-10-CM codes for SDOH to various stakeholder audiences was to:

• Develop this document as well as two-page communication tools for various audiences, including providers and coders. The coder tool, Transforming health care: Why including SDOH codes on claims is critical and provider tool, Using SDOH coding to transform health outcomes are available for use.
• Promote the use of the communication tools at various payer organizations through a high-level communication plan that outlines dissemination to stakeholder groups.

For a brief overview on the topic, check out our eHI Explains ICD-10-CM Coding for SDOH. Links to the provider and coder documents are forthcoming.

Missed Appointments, Missed Opportunities: Tackling The Patient No-Show Problem

I was asked last week at a healthcare industry roundtable about the single-biggest problem I’m trying to solve in my organization. My answer was surprisingly mundane, but relatable to anyone who has worked in a medical office: helping to address the ever-patient vexing problem of predicting and addressing patient no-shows.

Much has been made about the economic effects of patient “no-shows” on the health care system. One study found that no-shows cost the U.S. health care system more than $150 billion a year and individual physicians an average of $200 per unused time slot. After all, whether or not patients show up, healthcare organizations and medical practices still have to pay their staffs and cover expenses like rent and the cost of equipment.

But above and beyond the economic implications, no-shows have a direct impact on individuals’ health. When patients miss appointments, continuity of care is interrupted. Medication efficacy can’t be monitored regularly. Preventive services and screenings can’t be delivered in a timely manner. Acute illnesses are more likely to go untreated and become chronic conditions with complications. In short, missing an appointment can be severely detrimental to one’s health.

The full Forbes article can be viewed at this link.  

Medical device surveillance with electronic health records

Post-market medical device surveillance is a challenge facing manufacturers, regulatory agencies, and health care providers. Electronic health records are valuable sources of real-world evidence for assessing device safety and tracking device-related patient outcomes over time. However, distilling this evidence remains challenging, as information is fractured across clinical notes and structured records. Modern machine learning methods for machine reading promise to unlock increasingly complex information from text, but face barriers due to their reliance on large and expensive hand-labeled training sets. To address these challenges, we developed and validated state-of-the-art deep learning methods that identify patient outcomes from clinical notes without requiring hand-labeled training data. Using hip replacements—one of the most common implantable devices—as a test case, our methods accurately extracted implant details and reports of complications and pain from electronic health records with up to 96.3% precision, 98.5% recall, and 97.4% F1, improved classification performance by 12.8–53.9% over rule-based methods, and detected over six times as many complication events compared to using structured data alone. Using these additional events to assess complication-free survivorship of different implant systems, we found significant variation between implants, including for risk of revision surgery, which could not be detected using coded data alone. Patients with revision surgeries had more hip pain mentions in the post-hip replacement, pre-revision period compared to patients with no evidence of revision surgery (mean hip pain mentions 4.97 vs. 3.23; t = 5.14; p < 0.001). Some implant models were associated with higher or lower rates of hip pain mentions. Our methods complement existing surveillance mechanisms by requiring orders of magnitude less hand-labeled training data, offering a scalable solution for national medical device surveillance using electronic health records.

The full article can be downloaded below.  

Rural hospital closings reach crisis stage, leaving millions without nearby health care

Presidential candidates and other politicians have talked about the rural health crisis in the U.S., but they are not telling rural Americans anything new. Rural Americans know all too well what it feels like to have no hospital and emergency care when they break a leg, go into early labor, or have progressive chronic diseases, such as diabetes and congestive heart failure.

More than 20% of our nation’s rural hospitals, or 430 hospitals across 43 states, are near collapse. This is despite the fact that rural hospitals are not only crucial for health care but also survival of their small rural communities. Since 2010, 113 rural hospitals across the country have closed, with 18% being in Texas, where we live.

About 41% of rural hospitals nationally operate at a negative margin, meaning they lose more money than they earn from operations. Texas and Mississippi had the highest number of economically vulnerable facilities, according to a national health care finance report in 2016.

As rural health researchers, we’re well aware of the scope of rural hospitals woes’, which span the entire country. Struggling rural hospitals reflect some of the problems with the U.S. health care system overall, in that the poor often struggle to have access to care and there are few obvious solutions to controlling rising costs.

If 20% of America lives in a rural county, why is the nation so slow to address rural health disparities?

The full Salon article can be viewed at this link.