Over the past decade, “decision aids” – medical information in the form of software, pamphlets, and other media – have become popular tools to help patients make better choices about various medical treatments. Decision aids have recently gained relative popularity among physicians because of their potential to improve quality and control costs. To expand the use of decision aids, Section 936 of the Affordable Care Act presents guidelines for funding, developing, certifying, and implementing decision aids in the US healthcare system. However, while there are many proven benefits of decision aids, their cost saving potential has led to the entry of for-profit decision aid developers and insurers into this lucrative new market. Indeed, the content and presentation of information in decision aids can have profound effects on patient treatment choices with resulting health consequences. Given the risks of an unregulated decision aid market, the federal government has a responsibility to protect patients by a) requiring developers to be non-profit, and b) promoting the establishment of a nongovernmental organization of physicians and decision scientists that studies and creates unbiased decision aids.
An 86 year-old patient is diagnosed with terminal cancer. One day, the patient suffers a mild stroke, falls down, and breaks her hip. She is admitted to a hospital, where a doctor says she has three to nine months to live. If she receives a hip-replacement, there is a high chance her weak heart will not hold out for the surgery. If she doesn’t receive the replacement, she will painfully suffer through her remaining days. This is precisely the dilemma President Obama faced when his terminally ill grandmother, Madelyn Payne Dunham, broke her hip two weeks before she passed away in 2008. When asked how to approach difficult medical decisions, Obama said, “…there is going to have to be a conversation that is guided by doctors, scientists, ethicists” and ultimately, patients.[i]
This “conversation” is the core theme of a new approach of delivering medical care known as shared decision-making. Section 936 of the Affordable Care Act (ACA) defines shared decision-making as:
A collaborative processes between patients, caregivers or authorized representatives, and clinicians that engages the patient, caregiver or authorized representative in decision-making, provides patients, caregivers or authorized representatives with information about trade-offs among treatment options, and facilitates the incorporation of patient preferences and values into the medical plan.[ii]
In short, shared decision-making is a discussion between a patient and a physician meant to find the “right” treatment option when the relative risks, benefits, and costs of multiple medical treatments are uncertain. It is an integral part of health care reform because it redefines the patient-doctor relationship, emphasizes informed consumer choice, and has the potential to provide higher quality care at lower costs.
To facilitate shared decision-making, “decision aids” – medical information in the form of software, pamphlets, and other media – have become popular tools to help patients make better choices about various medical treatments. Over the past decade, decision aids have gained relative popularity among physicians because of their potential to improve quality and control costs.[iii] To expand the use of decision aids, Section 936 of the ACA presents guidelines for funding, developing, certifying, and implementing decision aids in the US healthcare system. 2 However, while there are many proven benefits of decision aids, their cost saving potential has led to the entry of for-profit decision aid developers and insurers into this lucrative new market. Indeed, the content and presentation of information in decision aids can have profound effects on patient treatment choices with resulting health consequences.
Given the risks of an unregulated decision aid market, the federal government has a responsibility to protect patients by a) requiring developers to be non-profit, and b) promoting the establishment of a nongovernmental organization of physicians and decision scientists that studies and creates unbiased decision aids.
Decision Aids: What are they, and how do they work?
Section 936 of the Affordable Care Act defines a patient decision aid as:
An educational tool that helps patients, caregivers or authorized representatives understand and communicate their beliefs and preferences related to their treatment options, and to decide with their health care provider what treatments are best for them based on their treatment options, scientific evidence, circumstances, beliefs, and preferences (Ibid.).
Decision aids are used in instances of preference-sensitive care, which involves “tradeoffs that affect a patient’s quality and/or length of life.”[iv] For instance, various life-saving breast cancer treatments involve the risk of breast loss, and prostate cancer procedures often pose the risk of incontinence, requiring patients to weigh expected quality of life over length of life.
To facilitate these tradeoffs, computer-based decision aids allow patients to input their unique preferences to match them with specific treatment plans. For instance, if a patient with prostate cancer wants to choose between undergoing surgery or localized radiation treatment, they could use the online decision aid pictured in Figure 1 to make a decision consistent with their personal preferences. Such interactive web-based tools are becoming a common medium for decision aids, but similar information can be conveyed in print form as well as phone-based information sessions between patients and decision aid consultants.
Ideally, decision aids should function as unbiased sources of information, but they have the potential to psychologically influence patient behavior by the application of heuristics and cognitive biases. Heuristics are common human cognitive errors that influence judgment and decision-making. In the case of decision aids, many heuristics are susceptible to potential manipulation. One of the most common ones is called ordering bias – the “skewing of results caused by the order in which information is presented.”[v] Evidence suggests that patients perceive information more favorably when it is presented first. For instance, a recent study involving decision aids for breast cancer patients found that those who first learned about the risks of a drug treatment favored the drug more than patients who first learned about the benefits.[vi] A decision aid developer could use this information to steer patients towards or away from the drug.
Similarly, a study involving decision aids for prostate cancer patients compared the ratios of negative to positive words (i.e. risks vs. benefits) in decision aids developed by the government, academia, and for-profit insurers. The study found that for-profit decision aid developers used significantly more negative words to describe the expensive treatment of surgery and more positive words to describe the less expensive treatment of watchful waiting.[vii] Figure 2 suggests that the government and private insurers may nudge patients towards cheaper treatments using scewed wording.
Decision aid developers also varied the lengths of treatment descriptions to give the impression that ones with shorter explanations are less worth considering. In this way, the organization and presentation of information in decision aids can influence patients’ medical choices by appealing to their cognitive biases.
The Impact of Decision Aids on the Cost and Quality of Care
Despite the potential exploitation of decision aids by for-profit developers, decision aids have been shown to improve the quality of medical care. Although “quality” care can be defined in many ways, decision aids have been shown to improve quality when it comes to informing patients about treatment options, conveying risks and benefits, improving satisfaction with medical choices, and reducing decisional conflict.[viii] For instance, a recent study of breast cancer patients in Ontario found that patients who used decision aids before receiving surgery were more educated about alternative treatment options, and as a result, experienced less decisional conflict than those who did not use decision aids.[ix] A separate study of patients with ischemic heart disease showed that patients were significantly more satisfied with their care when they used decision aids than those who had not.[x] The use of decision aids also improves patient participation in the decision-making process, as evidenced by a recent report on patients using decision aids for benign prostatic hypertrophy.[xi] Although there is agreement that decision aids can improve the quality of care under these measures, their impact on cost reduction remains unsettled.
Increasing evidence also suggests that decision aids can reduce costs for specific health conditions. For example, a groundbreaking report by Group Health Cooperative in Seattle, deemed the “largest observational study to date of the implementation of patient decision aids,” examined the effect of decision aids on the rates of knee and hip replacement surgeries.[xii] They found a 26% reduction in hip replacements, a 38% reduction in knee replacements, and cost savings between 12% and 21% over the six months decision aids were used for such elective surgeries. 12 Similarly, a study of women with menorrhagia in England found that patients who received information packets, videos, and medical consultations incurred significantly lower costs than those who did not.[xiii]
Clinical evidence suggests that decision aids generally promote non-interventional treatment options over surgical ones by about 20%. 8 Because surgical procedures are often more expensive, reducing surgeries reduces costs. In this way, decision aids involving specific conditions, which often involve surgical options, have the potential to dramatically reduce medical costs. In fact, the use of decision aids for 11 procedures involving preference-sensitive care could bring about healthcare cost savings of $9 billion over the next decade. 8 However, the question remains what potential negative consequences can result from non-surgical interventions when measured in terms of quality of life. For example, reducing hip-replacements may control costs, but at the expense of pain and suffering borne by candidates for the surgery who do not receive treatment.
The Use of Decision Aids by For-profit Companies
The cost saving potential of decision aids has sparked demand for their production and use. Naturally, a new market has evolved to meet this demand composed of for-profit developers, insurers, government agencies, academics, physicians, and others. While academics and care providers have little incentive to control patients’ choices through decision aids, those who pay for healthcare stand to gain by encouraging cheaper treatments. For instance, consider the following statement on a heart surgery decision aid: “For at least 70% of people who have heart bypass surgery, the survival rate is no better than if they had chosen to take medication alone.”[xiv] The biased tone clearly discourages surgery in favor of less costly medication treatment. If instead the statement was, “For at least 30% of people who have heart bypass surgery, the survival rate is better than if they had chosen to take medication alone,” patients would more likely favor bypass surgery. In this case, alternate phrasing reflects the intentions and biases of the author.
A 2012 report published by the Beazley Institute for Health Law & Policy found that “Decision support tools created by for-profit corporations or other interested parties may dramatically increase the risk of exposing patients to bias and misinformation, particularly given the fact that no legal mechanism currently exists to protect against such harms.”[xv] Although Section 936 of the ACA calls for decision aid certification standards, it mandates that the Department of Health and Human Services partner with an unspecified “entity” to design minimum requirements. 2 Based on the language in the ACA, there are no restrictions barring this “entity” from affiliations with the private insurance industry, for-profit decision aid developers, the pharmaceutical industry, or other special interest groups. Given the amount of lobbying involved in the ACA, it is likely that a biased “entity” could influence that certification process and further deregulate the “wild-west” market of for-profit decision aid developers.
Regulation of the Decision Aids Market
To regulate the ability of for-profit decision aid developers from influencing patient decisions, their financial incentives must be removed by requiring them to be non-profit. By removing their profit motives, developers can focus on creating informative products, not biased ones directed towards cost-savings. A recent report from the American Medical Association (AMA) highlighted the importance of developer objectivity when it stated, “…The independence of groups creating these tools and the use of quality control measures is especially important.”[xvi] Such groups must remain independent in two ways: a) detached from direct financial incentives, and b) unaffiliated with those who seek to use decision aids for cost-reduction.
As described in Figure 2, payers of healthcare coverage are automatically incentivized to reduce costs, including both for-profit insurers and the government. 7 Due to their vested financial interests, neither for-profits nor government agencies should be entrusted to develop unbiased decision aids. Therefore, as suggested by the AMA, only an independent advisory board is qualified to study and develop decision aids for clinical use.
Even President Obama acknowledged the necessity of such a board when he said, “…you have to have some independent group that can give you guidance. It’s not determinative, but I think has to be able to give you some guidance.” 1 Although this “independent group” is missing from the ACA, a coalition of doctors, decision scientists, and medical professionals would be best suited to study, evaluate, and develop unbiased decision aids to be used in the modern US healthcare system. Such a group would lack profit motive, be independent from government and private sector interests (such as cost reduction), and provide tools to facilitate shared decision-making. Additionally, doctors would be best positioned to identify practical patient needs and structural inefficiencies with decision aids in clinical practice. In turn, they will be more qualified to provide feedback about improvements to decision aids than profit-motivated developers who do not work directly with patients.
Once formed, this independent group should serve as the “entity” designated in the ACA to partner with the Department of Health and Human Services and develop certification standards for decision aids. These standards should address controls for bias and set minimum requirements for quality of outcomes when decision aids are used in clinical settings.
The Affordable Care Act takes steps towards achieving full transparency and accountability when it comes to creating unbiased decision aids, but falls short of fully regulating the burgeoning decision aid market. President Obama must follow through on his intention to create an “independent group” and amend his keystone piece of healthcare legislation to protect patients from manipulation by financially motivated interest groups.
1 Leonhardt, David. “After the Great Recession.” New York Times Magazine 28 Apr. 2009: 1-6. Print.
2 United States. Cong. House. The Patient Protection and Affordable Care Act. 111th Cong., 2nd sess. H. H. R. 3590. (enacted). Print.