Secondary Logo

Journal Logo

Scholarly Perspectives

Technology Can Augment, but Not Replace, Critical Human Skills Needed for Patient Care

Alrassi, James MD; Katsufrakis, Peter J. MD, MBA; Chandran, Latha MD, MPH

Author Information
doi: 10.1097/ACM.0000000000003733
  • Free

Abstract

Technology has enabled astonishing advances in medical practice, sometimes with associated undesirable effects. While acknowledging the immense value of technological innovation in medical practice, particularly in the realm of artificial intelligence (AI), we highlight the importance of humanistic characteristics and behaviors—critical human skills—for practitioners and the systems in which they work evolve.

Current State of Medical Practice

Changes to the practice of medicine in recent decades have yielded many improvements, but not without costs. The same can be said for societal changes that impinge on medical practice. To begin, we review some of the changes and resulting challenges to set the context for the subsequent discussion.

The introduction of the electronic medical record (EMR) in the United States has engendered benefits as well as drawbacks. EMRs have vastly changed health care documentation and delivery, but systems optimized to bill for sick care have not achieved the promised potential to manage population health or assist in critical decision making.1 The heterogeneity of EMRs makes analyzing large amounts of patient data technically challenging. Nonetheless, health care outcomes are being collected from EMR databases. These quality metrics are being tied to reimbursements, supporting the evolution toward value-based payments.2 However, the EMR has also been linked to growing physician burnout. Physicians often struggle for hours in the evening at home to complete the day’s entries. A 3-year study of family physicians that entailed the analysis of data from event-logging records showed that participants spent more than half of their days working on an EMR and almost a quarter of this work occurred after office hours.3 Another study involving 4 different specialties indicated that doctors spend 1 to 2 hours nightly on EMR-related tasks.4 Health care provider dissatisfaction with the EMR and clerical tasks correlate with high rates of burnout and depression.5 In addition to the adverse effects on health care practitioners, burnout has the potential to harm patients.6

Other system changes have also brought a mix of benefits and costs. Vertically integrated health systems are becoming more common, and mergers and acquisitions once found only in for-profit environments are now routine in health care.7,8 The role of the physician is rapidly evolving along with the practice of medicine. Patient care is more likely to be managed by a team of health care providers rather than a single physician who was previously the sole repository of patient information.9 The rise of “minute clinics” and the greater role that nurse practitioners, physician assistants, and pharmacists currently play in health delivery signal the end of the traditional primary care physician as gatekeeper to health care.10 The airline industry has shown how breaking down hierarchies can result in safer outcomes.11 In clinical medicine, breaking down professional hierarchies may shift clinical decision making to the nurse, pharmacist, and others.12 Some of these changes benefit patient care; others may come at a cost. Shift-based medical practice can fragment care by increasing the number of handoffs, thereby increasing the likelihood of medical errors.13

The democratization of information made possible by social media and the widespread use of the Internet has also had an effect on the practice of medicine. Google has become an important source for physicians, who may use the information they find online to inform their clinical decision making.14 In addition to helping physicians and other care providers, the Internet provides unfiltered information to patients and the lay public. Almost 80% of all Internet users search the web for health-related topics and almost two-thirds search for a specific disease, yet only a quarter of these users evaluate the reliability of the information source.15 Checking the source and verifying the accuracy of information is vital as many websites provide misinformation. To illustrate, in a study regarding sleep safety information on the web, less than half of the 1,300 websites evaluated had accurate recommendations.16

Physicians perceive their patients who access Internet-based health information as “confused” or “distressed.”17 The general public’s reliance on questionable information can adversely influence the patient–physician relationship and patient care. One example is the campaign surrounding antivaccine fabrications.18 Collectively, the issues cited above—the clerical demands of the EMR, physician burnout, fragmented care systems and teams, threats to provider autonomy, and the wide availability of questionable information—represent a microcosm of the societal and technological changes reflected in current and future health care systems. However, within these challenges lie also the seeds of opportunities.

In this article, we posit that technological advances must complement—not eclipse—the critical human skills that form the core and foundation of patient care. First, we summarize truly remarkable technical advances, and then we discuss the role for the physician in this technology-enabled future. Next, we explore the corresponding changes needed by the educational system and by those who prepare practitioners of the future, advocating changes and technology that allow practitioners to sharpen and advance their humanistic critical care skills. Finally, we envision the future of health care, emphasizing a system in which empathetic, humanistic physicians provide care augmented by technology.

Current State of Technology in Health Care

The advances made in AI are transforming medical care at a remarkable speed. List 1 provides examples of AI tools and applications that have been found to be equal or superior to human experts. Specific domains where AI is highly successful include visual and radiological image interpretation and the provision of diagnostic and therapeutic aids.19–35 AI software demonstrates the potential to improve direct interactions with patients, care documentation, and population health management.36–40 Technology also enables the practice of telemedicine, a particularly valuable option in rural communities and in other areas with limited access to practitioners and/or specialty expertise.41–43 A classic example of the success of telemedicine in training rural physicians and improving patient care outcomes is Project ECHO, through which primary care physicians have provided effective specialist supported care to patients with Hepatitis C.44 In 2015, the first virtual hospital, wherein patient management is provided through virtual but regular individualized interactions with many patients longitudinally over long periods of time, opened in the United States. The virtual hospital model promotes the concept of “touchless warmth.”45 In brick and mortar hospitals, robotic helpers now deliver food and medications for patients.46

List 1

Health Care Applications of Artificial Intelligence Shown to Be Equal or Superior to Human Experts

  • Visual image interpretation
    • ○ Electrocardiogram interpretation20
    • ○ Skin cancer21
    • ○ Retinal disease22–25
  • Radiologic image interpretation
    • ○ CT scan interpretation for strokes19
    • ○ Bone age26
    • ○ Longevity29
  • Diagnostic and therapeutic aids
    • ○ Suicide risk27
    • ○ Lung cancer28
    • ○ Congenital cataracts30
    • ○ Cancer detection31,32
    • ○ Cancer management33–35
    • Abbreviation: CT, computerized tomography.

What do these recent advances mean for the future? Concerns about these and other innovations prompted the Accreditation Council for Graduate Medical Education (ACGME) to develop “alternative futures” scenario-based planning.47 Resulting insights include enhanced complexity in health care delivery and accelerated “commoditization” of health care services. In this context, the term commoditization refers to the growth of lower-cost, just-in-time alternatives to traditional providers, and the accelerated use of continuously monitored patient sensors.47 These rich data sources would augment traditional reliance upon a patient’s history, physical examination, and test results. Models currently undergoing piloting have successfully predicted disease progression and future diseases in patients via insights derived from multidomain datasets.48 These models were able to predict future disease diagnoses, sometimes as far as 10 years in the future, using commonly coded information in EMRs, such as diagnostic codes, procedural codes, lab results, and medications.48

Other technologies may augment the improvements made possible by patient sensors. Advances in Natural Language Processing (NLP) may automate EMR analysis.49 Smartphones will become important tools to develop and deploy personalized medicine.50 Just as Google Maps creates a spatial representation of Earth from the level of street to continent, the large volume of data collected on humans will allow for the creation of a spatial representation of humans from DNA sequence to organ system.51 This “personal geographic information system” pertaining to a person’s health may shift care navigation from physician to patient.51

The Role of the Future Practitioner of Medicine

The fundamental changes enabled by advancing technology cause us to ask, and attempt to answer, the question What is the role for the physician in this technology-enabled future? Practitioners must evolve to accept, adapt, and/or embrace the technologies that enhance human capabilities while not relinquishing those activities that require human insight and understanding. Patients will undoubtedly continue to want trust, respect, care and connection, empathy, and effective communication, including active listening, from their care providers.52 Effective care depends on accurate diagnosing, skillfully performing necessary procedures, and prescribing appropriate therapy. Relieving human suffering depends on communicating empathically and compassionately and allowing for the safe expression of human emotions and concerns. “Critical human skills” are ones that we believe cannot be easily replaced by technology in the foreseeable future, such as communication with patients and peers, safe hand-offs of patients, interprofessional teamwork, and situational judgment. Moreover, the personal characteristics of empathy, humility, compassion, emotional intelligence, and passion for continued learning are also critical human skills. Emotional intelligence, as it relates to awareness of one’s own emotions and how to engage with patients’ emotions, will remain a vital skill for future practitioners, especially in team-based care.53 Likewise, empathy, which is clearly associated with improvements in clinically meaningful outcomes such as adherence to recommendations and overall stress reduction among patients,54 will continue to be a required physician competency.

In an Invited Commentary in Academic Medicine, Dr. Claiborne Johnston, focusing on medicine as the art of caring rather than as knowledge management, states that “the skills of caring are also associated with improved patient outcomes.”55 Practitioners appropriately note that time is required to connect empathically, listen without interrupting, and ensure effective communication. By automating or making more efficient the current tasks required of a physician, AI has the potential to give practitioners and their patients the gift of time. In one study, for every extra minute that the doctor spent with the patient, the readmission risk fell by 8%.56 A longer visit promotes better communication, increased trust, and importantly, improved outcomes due to a better therapeutic alliance between the patient and the care provider.57

Medical Educators Must Address Technology and Humanism

Just as future practitioners must change, corresponding changes are needed both in the educational system and in those who prepare practitioners of the future. As cognitive skills are effectively and efficiently delegated to technology, medical educators must enhance their training of the noncognitive, critical human skills that learners will need to relate to patients and relieve their suffering. Medical educators must focus on developing not only appropriate curricular elements and pedagogy but also accurate, reliable, and valid means to assess the critical human skills needed for success in the future practice of medicine. Progress in this realm must march alongside progress in health care delivery and reflect available technological advances. As others have noted, predictions about the future are fraught with error, yet they often stimulate innovation.58 Technological advances sometimes yield uneven curricular innovations,59,60 but the need to wholly evolve medical education is undeniable.

Deep neural networks and the tools they engender hold great potential in increasing the use of AI algorithms in many aspects of medical care61—but only if used appropriately. Technology-augmented medical practice will require physicians to be adept at applying appropriate point-of-care tools to aid in clinical decision making and patient care. Wartman and Combs emphasize teaching skills related to the “effective integration and utilization of information” from a variety of sources.62 They advocate a reboot of medical education, training future doctors to use decision support software and to manage sensors and robots in hospitals, in homes, or even within patients’ bodies. The curriculum of medical schools must evolve.

Even with excellent AI support in clinical care, physicians must be able to discern when to apply AI-directed management to care for a patient and when to deviate from the AI algorithm. Until AI systems become fully reliable, physicians will require the skills to seek, evaluate, and use information independently. For the foreseeable future, physicians will likely still need the skills to know when and how to search primary research literature, thus students may benefit from explicit instruction in seeking and applying evidence-based knowledge.63 A recent literature review has revealed that the majority of health sciences students have limited skills in locating, evaluating, and effectively using the health information they find on the Internet.64 Pedagogy in critical research skills is still developing, but is often limited to a single institution, and reports provide insufficient information to enable others to fully evaluate or replicate the intervention.65 Learners’ skills retention is also limited; to illustrate, one study of entry-level physicians has shown that the majority had not retained high-level search skills and lacked skills in identifying and applying the best evidence.66

The critical human skills cited above will remain essential to medical practice. In a 2019 Invited Commentary in Academic Medicine entitled “Humanism in Medicine: What does it mean and why is it more important than ever?”, George Thibault laments that “the increasing reliance on technology as a substitute for human interaction … [leads] to a general decline in patient and professional satisfaction.”67 Medical educators must strive to cultivate humanistic skills in future physicians through opportunities for deliberate practice, feedback, and course correction throughout a learner’s whole medical education journey to prepare them for successful practice in the future.

Entrustable professional activities (EPAs) provide a framework for trainees to track their skill progression over multiple encounters, thereby driving learning and improvement along a developmental path. Trainees know what they are being assessed on and how their progress aligns with the expected trajectory of learning.68 In the era of technology-augmented medical practice, additional EPAs may be required of all medical trainees: the ability to query big data, a clear awareness of the power of bioinformatics, knowledge of the strengths and limitations of AI algorithms, ability to evaluate patient-derived biometric data, and skills in basic computer programming. Just as the development of current EPAs relies on individuals with expertise in medicine and education, the development of new EPAs will likely rely on experts in AI and information science.

As technology-augmented medical practice enables the what and why of clinical decisions, and as technology outperforms the human brain for an increasing number of tasks, current assessment strategies that focus on known facts and clinical decision-making skills must also evolve. Assessment of trainees must measure their ability to use technology and query big data. Assessment experts need to develop new and valid methods to measure various aspects of critical human skills; fields outside of medicine may provide useful guidance.69

AI also promises to transform not only what is assessed but also how learners are assessed. NLP is already poised to guide scoring of medical licensing examinations.70 Assessment must become a seamless, longitudinal, and continuous process that is fully integrated into the clinical practice of medicine. Occasional, summative, point-in-time assessments must be supplemented by valid ongoing feedback to focus and guide learners’ development. The same technologies that enable interpretation of the medical record to guide patient care can be employed to assess practitioners. Multiple data points, collected at every clinical encounter over time, can create a clear picture of both the technical and the critical human skills of an emerging practitioner. NLP technology offers an option to automatically track daily clinical activities via EMR documentation. Thus, trainees can be tracked and assessed for specific skills and competencies through their authentic workplace documentation.

AI also holds promise for preclinical teaching and assessment. Rudimentary AI systems already exist that respond to the learner’s actions by providing spaced repetition of content weighted to strengthen demonstrated areas of deficit.71 One might imagine an NLP-enabled app that, through use of a phone’s microphone, could unobtrusively capture and analyze routine clinical and educational interactions. An NLP-enabled app that passively captures and analyzes spoken language could provide individualized feedback about the quality of a person’s performance in their interactions—whether a student working with peers on a case-based problem or a doctor interacting with patients and other members of the health care team. Such an app may be able to assess not just the clarity and accuracy of the content of the interaction, but even, potentially, the emotional tone.

Recognizing the primacy of critical human skills in medical care, the medical education community has worked to establish valid and reliable means to assess students’, residents’, and faculty members’ competence in skills such as clear communication, empathy, and teamwork. A methodological review of the assessment of humanism in medical students (that involved 202 different assessments reported in 155 articles) indicated that assessing humanism is often limited and reliant on a single quantitative measure.72 Likewise, the testing of emotional intelligence and empathy is still in its preliminary stages.73,74

The National Board of Medical Examiners (NBME) and collaborators have developed validated assessment tools to assess professionalism and other critical skills of pediatric residents.75 The NBME partnered with the American Board of Pediatrics and the Association of Pediatric Program Directors to form the Pediatric Milestones Assessment Collaborative (PMAC). The members of the collaborative have developed a system of technology-enabled multisource assessment to collect observations of residents, which, in turn, inform judgments about their achievement of ACGME milestones. PMAC has shown early evidence of effective longitudinal assessment of several noncognitive skills, and data collected have been used by the clinical competence committees of some residency programs to guide education.76 Different from many other residency assessments, the PMAC system uses the assessor’s responses to tailor assessment in real time, delivering different queries based on data provided to the assessor, thus improving reliability and reducing administrative burden. The principles and processes developed during this work may generalize to other specialties and to other stages along the education/training/practice continuum.

Challenges for the Future

While success is not guaranteed, the advances and opportunities described above create the potential for dramatic improvements both in patient outcomes and in patients’ and providers’ experience of care delivered. This final section explores some possible scenarios for the future.

Concerns over whether the machine will replace the human remain. A dystopian view of future medical care envisions a world where patients interact directly with health care appliances, describing their symptoms to an NLP interface, providing objective information via biosensors, and submitting to specimen collection as needed to inform the algorithms that will develop their diagnosis and treatment. Financial resources may determine both the range of available treatments and the ease with which individuals can access services.

This need not—and should not—be our future. In AI Super Powers: China, Silicon Valley and the New World Order, Kai-Fu Lee writes that AI “disseminates world class medical knowledge equally throughout highly unequal societies, and lets all doctors and nurses focus on the human tasks that no machine can do: making patients feel cared for and consoling them when the diagnosis isn’t bright.”77 We posit that ideal future medical practice will embody more integrated delivery of care, with the machine enabling optimal critical decision making while the human provides compassion, respect, listening, and emotional support to patients. How physicians care matters to patients and their outcomes55—and to physicians and their well-being. Modifying existing curricular and assessment frameworks now will help prevent the growth of a future workforce of disgruntled, dissatisfied physicians who provide technology-dependent health care mechanically without truly caring for their patients or about their outcomes.

In the era of technology-augmented care and the leveling of medical knowledge, medical school faculty—having so far admitted students based on their brilliance and cognitive abilities—will now need to select and train students with high empathy, communication skills, and emotional intelligence who are likely to blossom into the best healers capable of alleviating human suffering. The challenge for medical educators and regulatory bodies will be to draw from multiple data sources to develop an assessment mosaic: a rich, multifaceted picture of the individual. This mosaic could reliably provide data to discern the competence of learners—upon matriculation and throughout their training—in the critical human skills that will continue to play an important role in patient care, even in a technology-integrated system. Authentic, longitudinal, multisource assessments that are workplace-based and embedded in real patient care encounters will have to augment, and potentially replace, current simulated environments of assessment.

If medicine can successfully harness the potential of AI while simultaneously fostering the humanistic development of caring physicians, we will create a future that differs dramatically from the dystopia described above. In our bright future, patients are attended by a physician in a comfortable, nurturing environment that may be a health care facility or their own home; diagnosis and treatment decisions are facilitated by high-fidelity tele-present technology. Either the physician or the patient can invoke the participation of an AI avatar at any point during the encounter to solicit information, advice, or an opinion that is guided and informed by knowledge of both the patient’s full health record and current medical literature. The avatar remains silent unless invited into the interaction, or unless anything in the patient–doctor interaction—such as prescribing a contraindicated treatment—triggers a need for intervention. The AI system, first, automatically organizes the subjective and objective information exchanged during the visit into a written record for shared review by both the physician and the patient, and then drafts diagnoses and treatments for the physician’s review and approval.

Medical licensing authorities are already grappling with technology-enabled changes in the practice of medicine such as those related to telemedicine and licensure regulations.78 They will have to grapple with the evolution of standards of minimal competence to practice in the new era of technology-integrated medical practice. Technology that enables and supports patients and providers to move to high-tech, high-touch care would be welcomed by health care providers and patients alike. Technology-enhanced medicine practiced by humanistic, technology-trained providers will enable patients to receive high-quality compassionate care while allowing the provider to flourish by finding meaning and joy in this essentially human activity of medical care.

References

1. Jensen PB, Jensen LJ, Brunak S. Mining electronic health records: Towards better research applications and clinical care. Nat Rev Genet. 2012;13:395–405.
2. Heller RE 3rd, Coley BD, Simoneaux SF, et al. Quality measures and pediatric radiology: Suggestions for the transition to value-based payment. Pediatr Radiol. 2017;47:776–782.
3. Arndt BG, Beasley JW, Watkinson MD, et al. Tethered to the EHR: Primary care physician workload assessment using EHR event log data and time-motion observations. Ann Fam Med. 2017;15:419–426.
4. Sinsky C, Colligan L, Li L, et al. Allocation of physician time in ambulatory practice: A time and motion study in 4 specialties. Ann Intern Med. 2016;165:753–760.
5. Shanafelt TD, Dyrbye LN, Sinsky C, et al. Relationship between clerical burden and characteristics of the electronic environment with physician burnout and professional satisfaction. Mayo Clin Proc. 2016;91:836–848.
6. Panagioti M, Geraghty K, Johnson J, et al. Association between physician burnout and patient safety, professionalism, and patient satisfaction: A systematic review and meta-analysis. JAMA Intern Med. 2018;178:1317–1330.
7. Frakt AB, Garthwaite C. The CVS-Aetna Merger: Another large bet on the changing U.S. health care landscape. Ann Intern Med. 2018;168:511–512.
8. Bazzoli GJ, Dynan L, Burns LR, Yap C. Two decades of organizational change in health care: What have we learned? Med Care Res Rev. 2004;61:247–331.
9. Carter BL, Rogers M, Daly J, Zheng S, James PA. The potency of team-based care interventions for hypertension: A meta-analysis. Arch Intern Med. 2009;169:1748–1755.
10. Laws M, Scott MK. The emergence of retail-based clinics in the United States: Early observations. Health Aff (Millwood). 2008;27:1293–1298.
11. Helmreich RL, Merritt AC, Wilhelm JA. The evolution of crew resource management training in commercial aviation. Int J Aviat Psychol. 1999;9:19–32.
12. Gupte G, Vimalananda V, Simon SR, DeVito K, Clark J, Orlander JD. Disruptive innovation: Implementation of electronic consultations in a veterans affairs health care system. JMIR Med Inform. 2016;4:e6.
13. Riesenberg LA, Leitzsch J, Massucci JL, et al. Residents’ and attending physicians’ handoffs: A systematic review of the literature. Acad Med. 2009;84:1775–1787.
14. Tang H, Ng JHK. Googling for a diagnosis—Use of Google as a diagnostic aid: Internet based study. BMJ. 2006;333:1143–1145.
15. Silver L, Huang C. Social activities, information seeking on subjects like health and education top list of mobile activities.Pew Research Center: Internet& Tech. https://www.pewresearch.org/internet/2019/08/22/social-activities-information-seeking-on-subjects-like-health-and-education-top-the-list-of-mobile-activities. Published August 22, 2019. Accessed August 24, 2020.
16. Chung M, Oden RP, Joyner BL, Sims A, Moon RY. Safe infant sleep recommendations on the Internet: Let’s Google it. J Pediatr. 2012;161:1080–1084.
17. Ahmad F, Hudak PL, Bercovitz K, Hollenberg E, Levinson W. Are physicians ready for patients with Internet-based health information? J Med Internet Res. 2006;8:e22.
18. Smith N, Graham T. Mapping the anti-vaccination movement on Facebook. Inf Commun Soc. 2019;22:1310–1327.
19. U.S. Food and Drug Administration. FDA permits marketing of clinical decision support software for alerting providers of a potential stroke in patients. https://www.fda.gov/news-events/press-announcements/fda-permits-marketing-clinical-decision-support-software-alerting-providers-potential-stroke. Published February 13, 2018. Accessed August 24, 2020.
20. Haider A, Fazel-Rezai R. Heart signal abnormality detection using artificial neural networks. J Med Dev. 2014;8:020915.
21. Esteva A, Kuprel B, Novoa RA, et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature. 2017;542:115–118.
22. De Fauw J, Ledsam JR, Romera-Paredes B, et al. Clinically applicable deep learning for diagnosis and referral in retinal disease. Nat Med. 2018;24:1342–1350.
23. De Fauw J, Keane P, Tomasev N, et al. Automated analysis of retinal imaging using machine learning techniques for computer vision. F1000Res. 2016;5:1573.
24. Gulshan V, Peng L, Coram M, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA. 2016;316:2402–2410.
25. Brown JM, Campbell JP, Beers A, et al.; Imaging and Informatics in Retinopathy of Prematurity (i-ROP) Research Consortium. Automated diagnosis of plus disease in retinopathy of prematurity using deep convolutional neural networks. JAMA Ophthalmol. 2018;136:803–810.
26. Lee H, Tajmir S, Lee J, et al. Fully automated deep learning system for bone age assessment. J Digit Imaging. 2017;30:427–441.
27. Walsh CG, Ribeiro JD, Franklin JC. Predicting risk of suicide attempts over time through machine learning. Clin Psychol Sci. 2017;5:457–469.
28. Yu KH, Zhang C, Berry GJ, et al. Predicting non-small cell lung cancer prognosis by fully automated microscopic pathology image features. Nat Commun. 2016;7:12474.
29. Oakden-Rayner L, Carneiro G, Bessen T, Nascimento JC, Bradley AP, Palmer LJ. Precision radiology: Predicting longevity using feature engineering and deep learning methods in a radiomics framework. Sci Rep. 2017;7:1648.
30. Long E, Lin H, Liu Z, et al. An artificial intelligence platform for the multihospital collaborative management of congenital cataracts. Nat Biomed Eng. 2017;1:24.
31. Mori Y, Kudo SE, Misawa M, et al. Real-time use of artificial intelligence in identification of diminutive polyps during colonoscopy: A prospective study. Ann Intern Med. 2018;169:357–366.
32. Yasaka K, Akai H, Abe O, Kiryu S. Deep learning with convolutional neural network for differentiation of liver masses at dynamic contrast-enhanced CT: A preliminary study. Radiology. 2018;286:887–896.
33. Coudray N, Ocampo PS, Sakellaropoulos T, et al. Classification and mutation prediction from non-small cell lung cancer histopathology images using deep learning. Nat Med. 2018;24:1559–1567.
34. Lee SI, Celik S, Logsdon BA, et al. A machine learning approach to integrate big data for precision medicine in acute myeloid leukemia. Nat Commun. 2018;9:42.
35. Doyle-Lindrud S. Watson will see you now: A supercomputer to help clinicians make informed treatment decisions. Clin J Oncol Nurs. 2015;19:31–32.
36. Muoio D. Voice-powered, in-home care platform wins Amazon Alexa diabetes competition. MobiHealthNews. https://www.mobihealthnews.com/content/voice-powered-home-care-platform-wins-amazon-alexa-diabetes-competition. PublishedOctober 16, 2017. Accessed November 5, 2020.
37. Melton GB, Hripcsak G. Automated detection of adverse events using natural language processing of discharge summaries. J Am Med Inform Assoc. 2005;12:448–457.
38. Raghupathi W, Raghupathi V. Big data analytics in healthcare: Promise and potential. Health Inf Sci Syst. 2014;2:3.
39. Semigran HL, Linder JA, Gidengil C, Mehrotra A. Evaluation of symptom checkers for self diagnosis and triage: Audit study. BMJ. 2015;351:h3480.
40. Mack H. Buoy raises $6.7 million series A for AI-powered chatbot. The Wall Street Journal. https://www.wsj.com/articles/buoy-raises-6-7-million-series-a-for-ai-powered-chatbot-1502323040. Published August 9, 2017. Accessed August 24, 2020.
41. Medicaid.gov.. Telemedicine. https://www.medicaid.gov/medicaid/benefits/telemedicine/index.html. Published 2020 Accessed August 24, 2020.
42. The computer will see you now. The Economist. https://www.economist.com/science-and-technology/2014/08/20/the-computer-will-see-you-now. Published August 20, 2014. Accessed September 2, 2020.
43. Arora S, Thornton K, Murata G, et al. Outcomes of treatment for hepatitis C virus infection by primary care providers. N Engl J Med. 2011;364:2199–2207.
44. Robert Wood Johnson Foundation. Project ECHO. https://www.rwjf.org/en/how-we-work/grants-explorer/featured-programs/project-echo.html. Published May 3, 2019. Accessed August 24, 2020.
45. Allen A. A hospital without patients.Politico. https://www.politico.com/agenda/story/2017/11/08/virtual-hospital-mercy-st-louis-000573. Published November 8, 2017. Accessed August 24, 2020.
46. Masterson K. Higher tech and higher touch.University of California, San Francisco Patient Care. https://www.ucsf.edu/news/2015/04/125386/higher-tech-and-higher-touch. Published April 23, 2015. Accessed August 24, 2020.
47. Nasca TJ, Thomas CW. Medicine in 2035: Selected insights from ACGME’s scenario planning. J Grad Med Educ. 2015;7:139–142.
48. Miotto R, Li L, Kidd BA, Dudley JT. Deep patient: An unsupervised representation to predict the future of patients from the electronic health records. Sci Rep. 2016;6:26094.
49. Devarakonda MV, Mehta N, Tsou CH, Liang JJ, Nowacki AS, Jelovsek JE. Automated problem list generation and physicians perspective from a pilot study. Int J Med Inform. 2017;105:121–129.
50. Topol EJ. The Patient Will See You Now: The Future of Medicine Is in Your Hands. 2016. New York, NY: Basic Books.
51. Topol EJ. Individualized medicine from prewomb to tomb. Cell. 2014;157:241–253.
52. Rapport F, Hibbert P, Baysari M, et al. What do patients really want? An in-depth examination of patient experience in four Australian hospitals. BMC Health Serv Res. 2019;19:38.
53. Mayer JD, Salovey P, Caruso DR. Emotional intelligence: New ability or eclectic traits? Am Psychol. 2008;63:503–517.
54. Derksen F, Bensing J, Lagro-Janssen A. Effectiveness of empathy in general practice: A systematic review. Br J Gen Pract. 2012;63:76–84.
55. Johnston SC. Anticipating and training the physician of the future: The importance of caring in an age of artificial intelligence. Acad Med. 2018;93:1105–1106.
56. Andreyeva E, David G, Song H. The effects of home health visit length on hospital readmission.National Bureau of Economic Research. https://www.nber.org/papers/w24566#:~:text=Using%20our%20instruments%20and%20controlling,readmission%20by%20approximately%208%20percent. Published April 2018. Accessed August 24, 2020.
57. Tingley K. Trying to put a value on the doctor-patient relationship.The New York Times.https://www.nytimes.com/interactive/2018/05/16/magazine/health-issue-reinvention-of-primary-care-delivery.html. Published May 16, 2018 Accessed August 24, 2020.
58. Sklar DP. Looking ahead: Futures planning for medical education. Acad Med. 2019;94:1401–1403.
59. Bahner DP, Goldman E, Way D, Royall NA, Liu YT. The state of ultrasound education in U.S. medical schools: Results of a national survey. Acad Med. 2014;89:1681–1686.
60. Royal College of Surgeons. Welcome to the future of surgery. Future of Sugery. https://futureofsurgery.rcseng.ac.uk/. 2018
61. Topol EJ. Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again. 2019. New York, NY: Basic Books.
62. Wartman SA, Combs CD. Medical education must move from the information age to the age of artificial intelligence. Acad Med. 2018;93:1107–1109.
63. Kash MJ. Teaching evidence-based medicine in the era of point-of-care databases: The case of the giant bladder stone. Med Ref Serv Q. 2016;35:230–236.
64. Haruna H, Hu X. International trends in designing electronic health information literacy for health sciences students: A systematic review of the literature. J Acad Libr. 2018;44:300–312.
65. Maggio LA, Kung JY. How are medical students trained to locate biomedical information to practice evidence-based medicine? A review of the 2007–2012 literature. J Med Libr Assoc. 2014;102:184–191.
66. Cullen R, Clark M, Esson R. Evidence-based information-seeking skills of junior doctors entering the workforce: An evaluation of the impact of information literacy training during pre-clinical years. Health Info Libr J. 2011;28:119–129.
67. Thibault GE. Humanism in medicine: What does it mean and why is it more important than ever? Acad Med. 2019;94:1074–1077.
68. Wormald BW, Schoeman S, Somasunderam A, Penn M. Assessment drives learning: An unavoidable truth? Anat Sci Educ. 2009;2:199–204.
69. Pearl RM, Fogel AL. New physicians will need business school skills. New England Journal of Medicine Catalyst. https://catalyst.nejm.org/doi/full/10.1056/CAT.17.0416?utm_source=T&utm_campaign=RHC_NEJM_PhysiciansBizSkills&utm_medium=social. Published August 2017 Accessed September 2, 2020.
70. Sarker A, Klein AZ, Mee J, Harik P, Gonzalez-Hernandez G. An interpretable natural language processing system for written medical examination assessment. J Biomed Inform. 2019;98:103268.
71. Haynes MR, Gaglani SM, Wilcox MV, Mitchell T, Deleon V, Goldberg H. Learning through osmosis: A collaborative platform for medical education. Innovations in global medical and health education.2014:2.
72. Buck E, Holden M, Szauter K. A methodological review of the assessment of humanism in medical students. Acad Med. 2015;90(11 suppl):S14–S23.
73. Hemmerdinger JM, Stoddart SD, Lilford RJ. A systematic review of tests of empathy in medicine. BMC Med Educ. 2007;7:24.
74. Srivastava K, Joshi S, Raichaudhuri A, et al. Emotional intelligence scale for medical students. Ind Psychiatry J. 2011;20:39–44.
75. Hicks PJ, Margolis M, Poynter SE, et al.; APPD LEARN-NBME Pediatrics Milestones Assessment Group. The pediatrics milestones assessment pilot: Development of workplace-based assessment content, instruments, and processes. Acad Med. 2016;91:701–709.
76. Turner TL, Bhavaraju VL, Luciw-Dubas UA, et al.; the Association of Pediatric Program Directors Longitudinal Educational Assessment Research Network–National Board of Medical Examiners Pediatrics Milestones Assessment Group. Validity evidence from ratings of pediatric interns and subinterns on a subset of pediatric milestones. Acad Med. 2017;92:809–819.
77. Lee K-F. AI Superpowers: China, Silicon Valley, and the New World Order. 2018. Boston, MA: Houghton Mifflin Harcourt.
78. Federation of State Medical Boards. FSMB Survey Identifies Telemedicine as Most Important Regulatory Topic for State Medical Boards in 2016. https://www.fsmb.org/siteassets/advocacy/news-releases/2016/annual_state_board_survey_sesults.pdf. Published December 15, 2016 Assessed August 24, 2020.
Copyright © 2020 by the Association of American Medical Colleges