Difference between revisions of "Ethics in Informatics"
|Line 86:||Line 86:|
Submitted by Mahima Vijayaraghavan
Submitted by Mahima Vijayaraghavan
Latest revision as of 19:46, 8 November 2020
As technology becomes more intertwined with medicine, there remain major ethical concerns with its use and pervasiveness in healthcare (physicians, nursing, pharmaceutical etc.)
Healthcare delivery and physicians ascribe to an ethical model based on four principles :
1. Respect for autonomy - allow individuals the capacity to choose on their own
2. Non-Maleficence - refrain for harm
3. Beneficence - provide care that benefits the patient and prevent harm
4. Justice – equitable resource allocation, and fairness in decision making.
Informatics ethical principles center around similar principles. This article serves as a broad introduction to the history of ethics in informatics and specific fields of informatics and their respective ethical dilemmas.
History of Ethics in Informatics
In the early 20th century, Norbert Wiener published his book ‘The Human use of Human Beings,’ which introduced ethical principles and the relationship between humans and machines. Beginning in the late 1970s and 1980s with the advent of computer science, ethics was introduced in parallel by informaticians such as Charles Oppenheim, Pauker and Kzolovits. In 1986, Mason introduced the acronym of ‘PAPA’ – Privacy, Accuracy, Property, and Accessibility  . These principles have evolved and ultimately in 2016 the International Medical Informatics Association generated a code of Ethics that outlines major principles of informatics for professionals . Despite this code, as technology evolves, the guidelines remain under review and require continual revision. Consortiums such as AHIMA, and AMIA’s Ethics, Legal and Social Issues working group aims to advocate for ethical design of technology and advocate for policies surrounding such. In the UK, the UK Council for Health Informatics Professions outlines codes of ethics for its members. Technology growth far outpaces our capacity to ensure that each new development is ethically appropriate and serves our patients well; therefore, vigilance is required.
Key principles of Ethics in Informatics include but are not limited to:
1. Confidentiality and privacy
2. Data ownership
3. Data distribution
4. Appropriate selection of informatics tools based on clinical settings
5. Distribution and access to informatics tools
6. Algorithm and system evaluation and validation
Fields of Informatics and specific Ethical Issues
Artificial Intelligence (AI), Machine Learning, Natural Language Processing (NLP)
AI is an emerging technology in medicine. At the start of 2019, funding for AI companies in the imaging field exceeded 1.2 billion; this will likely continue to grow over time. Given such funding as well as incentive to grow this technology there remain concerns regarding its regulation as well as ethics of the use of AI in clinical medicine.
Principally, the ethical underpinning of algorithms used in AI and Machine Learning remain a topic of debate. The data sets used to generate these algorithms as well as the structure of the machine learning algorithms inherently contain biases as they are generated by humans, for human use. It is critical to note the difference between biases that are neutral or beneficial to patients, versus those that subsequently generate undesirable impacts on patients' wellbeing as well as access to care. The former are often beneficial for AI and Machine learning algorithms to generate improved outcomes and data mining. However, the latter introduces nuances and can generate poor health outcomes for groups of already under-represented and marginalized individuals. In addition, there are varied types of biases in AI; notably historical, representation, measurement, aggregation, as well as evaluation and algorithmic bias. Natural Language Processing too has inherent biases which may, at times, be beneficial for research as there are major differences in how gender impacts our language processing. Given the major distinction in desirable versus undesirable biases, consortiums are needed to vigilantly assess, analyze and identify these biases as well as develop a definition for fairness and clear standards. This has begun to occur in specific field; to guide principles for the use of AI in radiology, multiple governing bodies have generated a consensus statement . More broadly, AI-WATCH is an initiative of the European Commission to monitor the impact of AI on medical care as a whole.
Clinical Decision Support
Increasingly EMRs and CPOE introduce decision support including Drug-Drug interaction warnings, drug substitutions, or alerts for impending clinical decline. Although many CDS tools are robust, there remains possibility for error. Furthermore, if there is error introduced, who bears responsibility for the poor outcome (practitioner, vendor, both)? Traditional CDS tools do not take into account the patient’s individual socioeconomical variations and preferences. In light of this, some advocate for CDS to be reframed as clinician information support systems to allow for “preference sensitive clinical decision making” . Despite broad utilization and adoption of CDS, there remain ethical questions regarding its appropriate use and implications when errors arise.
Digital Health and Telehealth
The Internet of Things and wearables pose new opportunities for analyzing large amounts of data from patients. In addition, we can obtain information on digital biomarkers based on how humans interact with this technology - often times without their explicit knowledge. However, sex and gender-based differences in the utilization of, and interaction with wearables has not been thoroughly analyzed . The access to such technologies is based on monetization of such wearables, cost incentives as well as reimbursements. Similarly there remain numerous ethical discussions surrounding appropriate use of telehealth, accessibility and comprehension, as well as technical operator issues which may compromise quality of care . Ensuring equitable access of telehealth to a broad population is essential as adoption of telehealth increases.
As the field of informatics develops and becomes more pervasive in healthcare, the need for ethical governance bodies within each field of informatics is essential. It is possible that health informatics ethics may emerge as field in its own right .
Areas for further review include but are not limited to:
Public Health and Population Health Informatics 
Pediatric Informatics and ethical use of data from minors
 W. Phillips, “Ethical controversies about proper health informatics practices,” Mo. Med., vol. 112, no. 1, pp. 53–57, 2015.
 P. Szolovits and S. G. Pauker, “H. R. Warner, Txperiences with computer-based patient moni- toring,”,” pp. 1224–1226, 1979.
 M. RO, Four Ethical Issues of the Informatic Age. 1986.
 K. W. Goodman and K. W. Goodman, “The IMIA Code of Ethics for Health Information Professionals,” Ethics, Med. Inf. Technol., pp. 152–159, 2015, doi: 10.1017/cbo9781139600330.011.
 N. M. Safdar, J. D. Banja, and C. C. Meltzer, “Ethical considerations in artificial intelligence,” Eur. J. Radiol., vol. 122, no. July 2019, pp. 2019–2021, 2020, doi: 10.1016/j.ejrad.2019.108768.
 D. Cirillo et al., “Sex and gender differences and biases in artificial intelligence for biomedicine and healthcare,” npj Digit. Med., vol. 3, no. 1, pp. 1–11, 2020, doi: 10.1038/s41746-020-0288-5.
 A. Rajkomar, M. Hardt, M. D. Howell, G. Corrado, and M. H. Chin, “Ensuring fairness in machine learning to advance health equity,” Ann. Intern. Med., vol. 169, no. 12, pp. 866–872, 2018, doi: 10.7326/M18-1990.
 J. R. Geis et al., “Ethics of Artificial Intelligence in Radiology: Summary of the Joint European and North American Multisociety Statement,” J. Am. Coll. Radiol., vol. 16, no. 11, pp. 1516–1521, 2019, doi: 10.1016/j.jacr.2019.07.028.
 V. K. Rajput, J. Dowie, and M. K. Kaltoft, “Are clinical decision support systems compatible with patient-centred care?,” Stud. Health Technol. Inform., vol. 270, pp. 532–536, 2020, doi: 10.3233/SHTI200217.
 C. J. Humbyrd, “Virtue Ethics in a Value-driven World: Ethical Telemedicine,” Clin. Orthop. Relat. Res., vol. 477, no. 12, pp. 2639–2641, 2019, doi: 10.1097/CORR.0000000000000908.
 K. Masters, “Health Informatics Ethics,” Heal. Informatics Pract. Guid., no. June, pp. 233–252, 2018, doi: 10.32425/infoed/2018.1.11.
 C. Ölvingson, J. Hallberg, T. Timpka, and K. Lindqvist, “Ethical issues in public health informatics: Implications for system design when sharing geographic information,” J. Biomed. Inform., vol. 35, no. 3, pp. 178–185, 2002, doi: 10.1016/S1532-0464(02)00527-0.
Submitted by Mahima Vijayaraghavan