In the wake of the HITECH Act of 2009, office-based physician adoption of electronic medical records (EMRs) has nearly doubled as providers have taken advantage of meaningful use incentives. While the trend toward implementation of systems to maintain, manipulate, and share data has been palpable, there is no cohesive code of ethics addressing the issues related to the use of aggregated data. Informatics is by its nature multidisciplinary, and these varied stakeholders are governed by value systems that differ in significant ways.
The resulting IT systems and use cases present ethical challenges including threats to patient autonomy and shared decision-making, the provider-patient relationship, and the Big Data that is leveraged to inform evidence-based medicine. Despite their many benefits, both actual and aspirational, health information technology (HIT) and data science offer no panacea for the ills of our beleaguered healthcare system. In some cases the tech serves to exacerbate old problems, in others new problems spring from the socio-technical sandbox, and ultimately it is the patient who bears the greatest burden.
Many of the issues that we face with the use of HIT result from a battle that has been raging over the soul of healthcare – the clash of values between the business of medicine and the care of the patient. Business is winning the battle – it has been for some time – and it has shaped the development and implementation of information technology in the U.S. healthcare system. Health IT, in current incarnations, tends to look through the vulnerable toward a favorable bottom line.
Depending on who you ask, you will get a different answer as to whether HIT serves the triple aim of efficiency, cost-effectiveness, and improved outcomes in healthcare, but it matters not. The triple aim creates transactions where once there were relationships. And, as frustrations mount so do profits, and the values and dollars have been shifting accordingly. To wit: the base pay of insurance and hospital executives and administrators often outpaces those of surgeons and hospitalists, while medical societies report en masse to the National Coordinator for Health Information Technology that HIT certification and meaningful use are detrimental to patient safety, security, usability and interoperability. Since stakeholders are guided by different values and professional duties, a cohesive set of ethical guidelines is needed to inform practice for all professionals involved. The following four points provide some of the reasons why.
1. Patients are more than data points
Aggregated data of treatment outcomes do not necessarily reflect the needs of an individual patient and her experience with her health or healthcare provider. The record is a disembodied representation of a patient, who is deconstructed byte by byte without effectively being represented as a whole person. The lament from physicians has been consistent – EHRs and their myriad structured data points do not tell us much about what is happening with a particular patient, and clinical usability leaves much to be desired. The health, wellbeing, and dignity of an individual are not found in the aggregate – they are discovered at the bedside along the course of a provider-patient relationship. While there is an emphasis on patient satisfaction and follow-up to services received, aggregated data does not shape the individual patient experience or reflect the bedside interactions of an individual. Reducing a large population of individuals to a set of data points may create a scenario in which the reality of each individual is lost in the mire of the aggregate.
In the pursuit of cost-effective treatments, the patient narrative has been all but forsaken. Instead, we hear calls for inclusion of the data narrative which, in a certain sense, is an assault on patient autonomy and the shared decision-making process. It turns clinical paternalism on its head and exchanges goals of care and efficacy for cost-effectiveness. The fact is that improved outcomes and cost-containment are not mutually inclusive, and the data narrative may create adversarial relationships between patients and providers. With the business of health care guiding the use of HIT, the value systems are no longer aligned. The move toward a patient-centric, shared decision-making model ushered out the days of the doctor knows best – only to usher in the current meme of the data knows best.
2. Patients are more than consumers of treatment
Patients are vulnerable individuals as they pass through our health system. Many have medical and emotional needs and face challenges managing their own care and treatment in a consumer role. Consumers are well informed and typically consider cost as a part of their decision-making process outside of healthcare. Providers, on the other hand, are notorious for their general lack of transparency when it comes to pricing and hidden costs of care. While patients should be informed and educated, requiring a proactive consumer places many patients in a role that causes stress and unnecessary hardship when most patients want to address a specific medical needs and return to daily life. Educating the patient is the role of the healthcare provider, and information needs to be presented in a way that is meaningful to the patient’s well-being instead of placing patients in the role of an information-seeking consumer.
But there is a larger issue here with this high tech, low touch consumer model—we sometimes employ information technology for its newness and without solid evidence to confirm its efficacy. Use of technology for the sake of using new technology is pure folly in a clinical environment, and it can leave patients in the cold feeling frustrated and afraid. Health information technologists like to think that they work in the state of the art, but that is not so because they have not had the proper apprenticeship. With technology, the state of the art is the newfangled; in healthcare, Hippocrates reminds us through the millennia that the art is long. Clinically speaking, the notion that HIT is state of the art is at once laughable and lamentable. The art of clinical practice comes from honing practice over time to meet the many and varied needs of patients, yet there is precious little training on how to incorporate these technologies into clinical encounters. That piece, the art, has been an afterthought in the world of health IT. We have been far more focused on value and revenue than what is meaningful to patients and their treating physicians. In a technological system, the personal touch cannot be taken for granted, and many patients would still prefer a person to an iPad or portal. Now, get the machine that goes, “Bing!”
3. Healthcare providers are more than businesses
Data science is nothing new to health care – clinical investigators have been employing quantitative research techniques for some time. What has changed is who is using the data. It’s not just clinicians and researchers anymore; many hands hold the record nowadays and for a variety of purposes. This too is nothing new, as clinical ethicist Mark Siegler lamented the decrepitude of confidentiality back in the 80s for this very reason. Hospitals and healthcare providers need to meet more than financially viable projections. They need to be places of care and support for individuals geared toward improving health on the patient’s terms. Clinicians know this, but data scientist do not, and rightly so because it is not within their disciplinary training. In the world of health information technology, it is far more important that IT people be aligned with the values of healthcare than be good business people, as Paul Crotty calls for in his blog. In his considerations of the duality of man, Crotty may find that patients are far better served by a caring rather than a business response. At the same time, clinicians need to be better at leveraging IT. It has its problems but it’s here to stay and clinicians need to take control of their self-regulating professions. Training is needed so that providers can use IT to deepen their relationship with patients rather than have it serve as an intermediary or justification to bill at higher rates.
While improved outcomes are primarily aspirational in the world of Big Data, there is a clear and well-worn path to increased revenue and profitability. For some this is a, if not the primary rationale for providers to harness data science—to analyze claims data. Financial concerns have been, by far, the most important to those implementing health information technologies. While it is imperative that providers reduce duplicative or unnecessary interventions, adding the layer of cost-effectiveness for the provider creates a conflict of interest with the patient who may be more interested in her health and well-being than her provider’s bottom line. The standard of care should be determined by clinical efficacy rather than profitability. Moreover, the notion that cost-effectiveness leads to a lower cost of care for patients is dubious. To the contrary, what we have seen is that efficiencies lead to greater profit and executive compensation.
4. Data Discriminates
Despite hubristic claims of a potential healthcare revolution, Big Data and its purveyors are deeply biased, discriminate based on discipline, exacerbate health literacy issues, and do not necessarily lead to changes in behavior for even the most basic things, such as provider hand hygiene. For starters, we can look at whose data is being aggregated. Underserved populations, by the very nature of their being underserved, have less data to aggregate. Minority groups, women, and other groups whose health issues have bene under-represented in research face the same if not more bias in treatment as research does not reflect their needs. We can also look to research bias and the sorts of studies that are funded and published or unpublished. For example, research conducted in areas of greater commercial value can be biased towards generating return on investment, as we have seen in numerous electronic health records. Then, there is a matter of publication bias because in too many cases only positive results are published resulting in gaps in knowledge. For example, industry-funded studies are far more likely to show clinical effectiveness of a drug than neutrally-funded studies.
The use of Big Data can deepen the digital divide and gaps in health literacy. Information, knowledge, and wisdom are very different things. Pouring on the data and information does not create knowledge. Information must be tailored, channeled, and delivered in a way that meets the patient where she is.
As high as that bar is for providers, it’s exponentially higher for vulnerable patients as are the stakes.
Data and information are not enough. At some point patients require some intermediary agent to help them transform the deluge of information into knowledge in a way that is meaningful to the patient on the patient’s terms. Yet, the so-called efficiencies that we enjoy allow physicians to spend less time with patients. They can bill more in less time and then move on. That’s less time for doctors to explain things (by the way, the word doctor means teacher, not healer). Even if they had they time, there is no causal connection between knowledge and behavior modification when it comes to improvements in care. As Damian Mingle points out, it was 130 years after Ignaz Semmelweis call for improvements in hygiene that CDC adopted and published hand hygiene guidelines. What Mingle misses is that hand hygiene statistics for clinicians are still abysmal—the CDC places hand hygiene adherence in hospitals somewhere between 29 percent and 48 percent. This is not a win for data, it’s a debacle. We’ve known for some time that hand washing matters, but the data is not enough to change behaviors as grave as the consequences may be.
Data science has a lot to offer medicine, but like medicine the science must be wielded with an art. The values of medicine and information technology are disparate in many ways, but they can be harmonized if the will is there. There is the distinct possibility that the best use of health information technology is to find new ways to listen to patients and enhance the doctor-patient relationship rather than dictate or placate it. Ethics, and not data science, will help to achieve that end. Big data is merely the latest instrumentality of the medical industrial complex. Without an alignment of values, we will continue to see it leveraged to put profits before patients.
This article is published as part of the IDG Contributor Network. Want to Join?