Guiding Principles to Address the Impact of Algorithm Bias on Racial and Ethnic Disparities in Health and Health Care

Importance

Health care algorithms are used for diagnosis, treatment, prognosis, risk stratification, and allocation of resources. Bias in the development and use of algorithms can lead to worse outcomes for racial and ethnic minoritized groups and other historically marginalized populations such as individuals with lower income.

Objective

To provide a conceptual framework and guiding principles for mitigating and preventing bias in health care algorithms to promote health and health care equity.

Evidence Review

The Agency for Healthcare Research and Quality and the National Institute for Minority Health and Health Disparities convened a diverse panel of experts to review evidence, hear from stakeholders, and receive community feedback.

Findings

The panel developed a conceptual framework to apply guiding principles across an algorithm's life cycle, centering health and health care equity for patients and communities as the goal, within the wider context of structural racism and discrimination. Multiple stakeholders can mitigate and prevent bias at each phase of the algorithm life cycle, including problem formulation (phase 1); data selection, assessment, and management (phase 2); algorithm development, training, and validation (phase 3); deployment and integration of algorithms in intended settings (phase 4); and algorithm monitoring, maintenance, updating, or deimplementation (phase 5). Five principles should guide these efforts: (1) promote health and health care equity during all phases of the health care algorithm life cycle; (2) ensure health care algorithms and their use are transparent and explainable; (3) authentically engage patients and communities during all phases of the health care algorithm life cycle and earn trustworthiness; (4) explicitly identify health care algorithmic fairness issues and trade-offs; and (5) establish accountability for equity and fairness in outcomes from health care algorithms.

Conclusions and Relevance

Multiple stakeholders must partner to create systems, processes, regulations, incentives, standards, and policies to mitigate and prevent algorithmic bias. Reforms should implement guiding principles that support promotion of health and health care equity in all phases of the algorithm life cycle as well as transparency and explainability, authentic community engagement and ethical partnerships, explicit identification of fairness issues and trade-offs, and accountability for equity and fairness.

Using an online quiz-based reinforcement system to teach healthcare quality and patient safety and care transitions at the University of California

Quality issue

Implementing quality improvement (QI) education during clinical training is challenging due to time constraints and inadequate faculty development in these areas.

Initial assessment

Quiz-based reinforcement systems show promise in fostering active engagement, collaboration, healthy competition and real-time formative feedback, although further research on their effectiveness is required.

Choice of solution

An online quiz-based reinforcement system to increase resident and faculty knowledge in QI, patient safety and care transitions.

Implementation

Experts in QI and educational assessment at the 5 University of California medical campuses developed a course comprised of 3 quizzes on Introduction to QI, Patient Safety and Care Transitions. Each quiz contained 20 questions and utilized an online educational quiz-based reinforcement system that leveraged spaced learning.

Evaluation

Approximately 500 learners completed the course (completion rate 66-86%). Knowledge acquisition scores for all quizzes increased after completion: Introduction to QI (35-73%), Patient Safety (58-95%), and Care Transitions (66-90%). Learners reported that the quiz-based system was an effective teaching modality and preferred this type of education to classroom-based lectures. Suggestions for improvement included reducing frequency of presentation of questions and utilizing more questions that test learners on application of knowledge instead of knowledge acquisition.

Lessons learned

A multi-campus online quiz-based reinforcement system to train residents in QI, patient safety and care transitions was feasible, acceptable, and increased knowledge. The course may be best utilized to supplement classroom-based and experiential curricula, along with increased attention to optimizing frequency of presentation of questions and enhancing application skills.

Leveraging lean principles in creating a comprehensive quality program: The UCLA health readmission reduction initiative

INTRODUCTION

UCLA Health embarked to transform care by integrating lean methodology in a key clinical project, Readmission Reduction Initiative (RRI).

METHODS

The first step focused on assembling a leadership team to articulate system-wide priorities for quality improvement. The lean principle of creating a culture of change and accountability was established by: 1) engaging stakeholders, 2) managing the process with performance accountability, and, 3) delivering patient-centered care. The RRI utilized three major lean principles: 1) A3, 2) root cause analyses, 3) value stream mapping.

RESULTS

Baseline readmission rate at UCLA from 9/2010-12/2011 illustrated a mean of 12.1%. After the start of the RRI program, for the period of 1/2012-6/2013, the readmission rate decreased to 11.3% (p<0.05).

CONCLUSION

To impact readmissions, solutions must evolve from smaller service- and location-based interventions into strategies with broader approach. As elucidated, a systematic clinical approach grounded in lean methodologies is a viable solution to this complex problem.

From discharge to readmission: Understanding the process from the patient perspective

BACKGROUND

Patient engagement is critical in delivering high-quality care. However, literature investigating patient perspectives on readmissions is lacking.

OBJECTIVES

To understand patients' beliefs and attitudes about 30-day readmissions and to elucidate areas for improvement aimed at reducing readmissions.

DESIGN

In person survey.

SETTING

Academic medical center and affiliated community hospital.

PATIENTS

Patients with 30-day readmissions to medicine and cardiology services.

MEASUREMENTS

Patient readiness, attitudes toward readmissions, discharge instructions, ambulatory resources, and follow-up care.

RESULTS

Of 479 eligible patients approached for interviews, 230 (48%) were interviewed. Of these, 28% reported not feeling ready for discharge, and this correlated with inadequate symptom resolution, poor pain control, and concerns about self-care. Sixty-five percent remembered reviewing discharge paperwork, but over 22% could not identify critical information on this paperwork. Eighty-five percent reported having a primary doctor; however, only 56% of patients who received a contact number on discharge called a physician before returning to the hospital. One-third of patients knew where to obtain same-day care outside of the emergency room. Lastly, patients reported feeling more relieved than burdened upon readmission (7.7 [standard deviation {SD} 2.8) vs 5.9 [SD 3.4]; P < 0.001, scale of 1-10).

CONCLUSIONS

By engaging readmitted patients we have illuminated areas for future interventions, including better symptom management and self-care planning before discharge, more clarity in discharge instructions, promoting awareness of outpatient resources, and improved alignment of patient and provider attitudes about readmissions. As the United States strives to reduce readmissions, attending to the patient perspective is critical in informing appropriate avenues for quality improvement. Journal of Hospital Medicine 2016;11:407-412. © 2016 Society of Hospital Medicine.

Tracking and sustaining improvement initiatives: leveraging quality dashboards to lead change in a neurosurgical department

Increasingly, hospitals and physicians are becoming acquainted with business intelligence strategies and tools to improve quality of care. In 2007, the University of California Los Angeles (UCLA) Department of Neurosurgery created a quality dashboard to help manage process measures and outcomes and ultimately to enhance clinical performance and patient care. At that time, the dashboard was in a platform that required data to be entered manually. It was then reviewed monthly to allow the department to make informed decisions. In 2009, the department leadership worked with the UCLA Medical Center to align mutual quality-improvement priorities. The content of the dashboard was redesigned to include 3 areas of priorities: quality and safety, patient satisfaction, and efficiency and use. Throughout time, the neurosurgery quality dashboard has been recognized for its clarity and its success in helping management direct improvement strategies and monitor impact. We describe the creation and design of the neurosurgery quality dashboard at UCLA, summarize the evolution of its assembly process, and illustrate how it can be used as a powerful tool of improvement and change. The potential challenges and future directions of this business intelligence tool are also discussed.

Analysis of 30-day readmissions among neurosurgical patients: surgical complication avoidance as key to quality improvement

OBJECT.: In terms of measuring quality of care and hospital performance, an outcome of increasing interest is the 30-day readmission rate. Recent health care policy making has highlighted the necessity of understanding the factors that influence readmission. To elucidate the rate, reason, and predictors of readmissions at a tertiary/quaternary neurosurgical service, the authors studied 30-day readmissions for the Department of Neurosurgery at two University of California, Los Angeles (UCLA), hospitals.

METHODS

Over a 3-year period, the authors retrospectively identified adult and pediatric patients who had been discharged from the UCLA Medical Center after having undergone a major neurosurgical procedure and being readmitted within 30 days. Data were obtained on demographics, follow-up findings, diagnosis and reason for readmission, major operations performed, and length of stay during index admission and readmission. Reasons for readmission were broadly categorized into surgical, medical diagnosis/complication, problem associated with the original diagnosis, neurological decompensation, pain management, and miscellaneous. For further characterization, subgroup analysis and in-depth chart review were performed.

RESULTS

Over the study period, 365 (6.9%) of 5569 patients were readmitted within 30 days. The most common diagnosis at index admission was brain tumor (102 patients), followed by CSF shunt malfunction (63 patients). The most common reason for readmission was surgical complication (50.1%). Among those with surgical complications, the largest subgroup consisted of patients with CSF shunt-related problems (77 patients). The second and third largest subgroups were surgical site infection and CSF leakage (41 and 31 patients, respectively). Medical diagnosis/complication was the second most frequent (27.9%) reason for readmission.

CONCLUSIONS

Surgical complications seem to be a major reason for readmission at the neurosurgical practice studied. Results indicate that the outcomes that are amenable to and would have the greatest effect on quality improvement are CSF shunt-related complications, surgical site infections, and CSF leaks.

Healthcare reform from the inside: A neurosurgical clinical quality program

During the past decade, the U.S. health care system has faced increasing challenges in delivering high quality of care, ensuring patient safety, providing access to care, and maintaining manageable costs. While reform progresses at a national level, health care providers have a responsibility and obligation to advance quality and safety. In 2009, the authors implemented a department-wide Clinical Quality Program. This Program comprised of an inter-disciplinary group of providers and staff working together to ensure the highest quality of patient care. The following methodology was followed to establish the Program: (1) Identifying the Department's quality improvement (QI) and patient safety priorities based on reviewing prior performance data; (2) Aligning the Department's priorities with institutional goals to select mutually significant initiatives; (3) Finalizing the goals for improvement based on departmental priorities, existing expertise and resources; (4) Launching the Program through an inter-disciplinary retreat that emphasizes open dialogue, innovative solutions, and fostering leadership in frontline providers; (5) Sustaining the QI initiatives through proactive performance review and management of barriers; and (6) Celebrating success to empower providers to remain engaged. Several challenges are inherent to the implementation of a clinical quality program, including lack of time and expertise, and the hierarchical nature of medicine, which can create a barrier to teamwork. This Program illustrates that improvement can lead to a sustainable clinical quality program and culture change.

Promoting patient-centred care through trainee feedback: assessing residents' C-I-CARE (ARC) program

AIMS

In recent years, patient satisfaction has been integrated into residency training practices through core competency requirements as set forth by the Accreditation Council of Graduate Medical Education (ACGME). In 2006, the UCLA Health Systems established a program designed to obtain patient feedback and assess the communication abilities of resident physicians with a standard tool through the Assessing Residents' C-I-CARE (ARC) Program.

METHODS

This Program utilized a 17-item questionnaire, completed via a facilitator-administered interview, which employed polar, Likert and comment scale questions to assess physician trainees' interpersonal and communication skills.

RESULTS

From 2006 to 2010, the ARC Program provided patient feedback data to more than six clinical departments while collecting 5,634 surveys for 323 trainees. Scores for resident recognition and performance increased from the first to second year of activity by an average of 22.5%, while attending recognition scores decreased 19% over the four years. Additionally, residents and attendings in surgical specialties received higher recognition rates than those in non-surgical specialties.

CONCLUSIONS

The ARC Program provided a standard tool for attaining patient feedback through a facilitator-administered survey that assisted in the accreditation process of training programs. Furthermore, hospitals, health organizations and medical schools may find the ARC Program valuable in collecting information for quality control as well as providing an opportunity for students to become involved in the healthcare field.

Hospital quality and patient safety competencies: development, description, and recommendations for use

BACKGROUND

Hospitalists are poised to have a tremendous impact on improving the quality of care for hospitalized patients. However, many hospitalists are inadequately prepared to engage in efforts to improve quality, because medical schools and residency programs have not traditionally emphasized healthcare quality and patient safety in their curricula.

METHODS

Through a multistep process, the Society of Hospital Medicine (SHM) Quality Improvement Education (QIE) subcommittee developed the Hospital Quality and Patient Safety (HQPS) Competencies to provide a framework for developing and assessing curricula and other professional development experiences. This article describes the development, provides definitions, and makes recommendations on the use of the HQPS Competencies.

RESULTS

The 8 areas of competence include: Quality Measurement and Stakeholder Interests, Data Acquisition and Interpretation, Organizational Knowledge and Leadership Skills, Patient Safety Principles, Teamwork and Communication, Quality and Safety Improvement Methods, Health Information Systems, and Patient Centeredness. Reflecting differing levels of hospitalist involvement in healthcare quality, 3 levels of expertise within each area of competence have been established: basic, intermediate, and advanced. Standards for each competency area use carefully selected action verbs to reflect educational goals for hospitalists at each level.

CONCLUSIONS

Formal incorporation of the HQPS Competencies into professional development programs, and innovative educational initiatives and curricula, will help provide current hospitalists and the next generations of hospitalists with the needed skills to be successful.