The Choosing Wisely initiative and laboratory test stewardship

Introduction

One undeniable hallmark of medical care in the US is that it is expensive. According to 2016 data from the Organization for Economic Cooperation and Development, the US spends substantially more per capita than any of the 35 countries it surveys, beating out the second place country (Switzerland) by nearly 25% [1]. Put another way, other wealthy countries spend half as much per person on healthcare as the US does, on average. However, the US has fewer physician consultations per capita than most comparable countries [2]. This implies that when a US patient visits their doctor, what occurs at that visit must, on average, cost substantially more than it would in most other wealthy nations, either due to increased utilization of services or pharmaceuticals, or increased prices for those services, increased administrative costs or some combination of these and other factors. More frustratingly, the US does not seem to be purchasing much value with these expenditures, as healthcare outcomes of the US lag well behind many developed nations [3].

When confronted with the large cumulative expenditure on healthcare in the US, one might be tempted to ignore the role laboratory tests play in driving costs because the overall percentage of healthcare dollars going to laboratory testing sits at around 4% of US national healthcare expenditure [4]. However, those in the laboratory profession recognize the fallacy of this argument almost immediately, because while laboratory tests are collectively, and often individually, quite inexpensive, they provide a substantial fraction of the objective data that is used in medical decision-making, and they therefore likely play a substantial role in driving other non-laboratory costs [5]. A clear example of this is a prostate-specific antigen (PSA) test, which is individually inexpensive (the US governmental insurance program Medicare’s reimbursement in the author’s region at the time of this writing is USD $22.71 for a single PSA test), but which could lead to a biopsy, prostatectomy, chemotherapy or radiation therapy, the cumulative costs of which may exceed the initial test cost by over 1000-fold. Alas, quantifying this magnification effect is challenging. While it has been stated, quoted and cited for more than a decade that laboratory tests influence approximately 70% of medical decisions, this notion, often proffered as fact, is actually a surmise made by a single clinical pathologist that has taken on a life of its own, despite the fact that it has no basis in any rigorous study or evidence (Dr. Rod Forsman, personal communication and this [6]). While the true average impact of all laboratory tests on all medical decisions is thus unknown, as nearly a third of outpatient visits, approximately half of emergency department visits and essentially all inpatient visits are associated with laboratory testing, it must be the case that laboratory testing has an effect on healthcare decisions that is far larger than its proportion of expenditure [7].

The relative lack of evidence of laboratory testing’s role in overall health care overutilization and concomitant over-expenditure notwithstanding, it seems entirely rational to surmise, along with Dr. Forsman, that laboratory tests have an outsized effect in driving medical decision making, and also to assume that optimizing laboratory test utilization will be beneficial to patients and our economy. It is with these ideas in mind that several of the specialty societies engaged in Choosing Wisely in 2012–2013 began to develop recommendations, many of which dealt specifically with the topic of laboratory test overutilization. Since that time, many other international Choosing Wisely efforts have arisen [8], as indicated in Table 1, adding to these US recommendations.

The Choosing Wisely process

Significant thought was put into the process by which the Choosing Wisely initiative would solicit or encourage recommendations to be made by the participating specialty societies [9]. The effort was based on two precepts: Complexity Theory, in which a process was allowed to develop organically so that an “emergent design” arose in a context of minimum constraints, and Self-Determination Theory, in which the autonomy of each individual special society was supported so that they could work in their own preferred way and only join the overall initiative when ready. Although it would have been very difficult to document, it is likely that some specialty societies that otherwise might have resisted participation in the Choosing Wisely initiative chose to join eventually because so many other societies had joined that their absence would be noticed, akin to “peer pressure”. With these mechanisms in place, the individual specialty societies began to generate recommendations based on only a few rules: recommendations should pertain to tests and treatments that are costly and/or used frequently, recommendations should be evidence-based, recommendations should pertain to decisions that are within the control of the specialty and recommendations should be developed and approved using a transparent process.

As one can imagine, some of the later criticisms of Choosing Wisely (to be discussed below) stem directly from this rather loose structure. The process Choosing Wisely allows for recommendation generation is much less rigorous than what is in place in essentially any process that generates official medical practice guidelines, which is one reason that Choosing Wisely generates “recommendations” rather than “guidelines”.

Choosing Wisely’s laboratory test-focused recommendations

While the American Society for Clinical Pathology (ASCP), an organization focused entirely on pathology and laboratory testing, has been involved in the Choosing Wisely initiative since 2013 and has made 20 recommendations (Table 2), as of this writing 71 recommendations relating to laboratory tests have been made by other specialty societies as well (Table 3).

The laboratory test recommendations made to date are mostly non-controversial, and to those involved in the field of laboratory test stewardship in any capacity, they mostly appear to be the proverbial “low-hanging fruit”. For example, certain recommendations, such as the ASCP’s recommendation to avoid frozen sections if the results will not affect immediate management, are little more than common sense, and the Society of Hospital Medicine and Critical Care Societies’ collaborative recommendations to avoid daily routine inpatient laboratory testing address a problem that almost everyone in medical practice agrees is wasteful. One might question the utility of recommendations like this, as they appear only to be admonitions to remember the obvious, and as calls for enhanced vigilance regarding laboratory test utilization are notoriously ineffective. Additionally, the basis of these recommendations lies not in evidence but rather in logic, as no one would design a clinical trial to assess the efficacy of needless frozen sections or routine tests ordered without any indication. Nonetheless, these recommendations may still have value in supporting an appeal to a higher authority when a contentious situation arises, i.e. a pathologist may benefit in crafting a hospital policy discouraging needless frozen sections or routine daily laboratory testing by citing Choosing Wisely.

Other laboratory test recommendations, like ASCP’s recommendation to avoid routine vitamin D screening, routine preoperative testing workups and thrombophilia workups during acute clotting events address issues that are significant but controversial to some clinicians. Selected practitioners do prefer to order these tests quite commonly, even though the evidence for their use may be weak or nonexistent. In making these recommendations, the ASCP has indicated that its experts believe that stopping these testing practices is safe, and likely beneficial to patients. Thus, these types of recommendations often find utility as the bases of large system-wide initiatives in health care systems.

Although unusual, some recommendations duplicate or conflict with other recommendations. This is unusual because the recommendation generation process involves communication with other professional societies and review of other existing recommendations, so professional societies can be aware of conflicts and adjudicate them prior to publication. PSA testing, mentioned previously, is one such area, as the American Academy of Family Physicians recommends against routine use of PSA as a cancer screening test and the American Society of Clinical Oncology recommends against screening asymptomatic men with less than 10 years of life expectancy, but the American Urological Society uses positive language in recommending, “Offer PSA screening for detecting prostate cancer only after engaging in shared decision making”. These are not entirely contradictory recommendations, but they certainly reflect different interpretations of the published literature and cost-benefit analyses of PSA testing, and they likely also reflect biases and maybe even financial motives of different groups of physicians, as urologists are likely to see a financial return from screening strategies that drive referrals. That no formal process exists to resolve such inconsistencies is a drawback of the Choosing Wisely approach.

Finally, another class of recommendations worth considering includes those that deal with proprietary tests, such as ASCP’s recommendations dealing with methylated septin 9 and nuclear magnetic resonance lipid assays. These recommendations, by virtue of their push to decrease test utilization, could conceivably adversely affect the profitability of specific laboratories or holders of intellectual property. Other recommendations, such as the one discouraging large immunoglobulin E (IgE) allergy testing panels made by the American Academy of Allergy, Asthma and Immunology (AAAAI), ostensibly threaten to reduce the utilization of tests that can generate substantial revenue for some clinical laboratories. The Choosing Wisely initiative has no plan to deal with these situations, where the recommendations of a specialty body would be expected to penalize one or a very few individuals or laboratories financially. However, as the initiative moves into the future and novel diagnostic tests continue to be developed at a fast pace, one can expect this thorny issue to arise frequently, and formal recognition of this issue may become important.

Overdiagnosis and Choosing Wisely

One of the critical concepts underlying the Choosing Wisely initiative is the idea of overdiagnosis. Welch et al. has authored numerous studies documenting the problem of overdiagnosis [10, 11], which he defines as the diagnosis of a disease that will never cause symptoms or death during a patient’s lifetime. An example of overdiagnosis is the small, indolent prostatic carcinoma discovered through PSA testing and concomitant biopsy in a 78-year-old male with a 4-year life expectancy due to coronary artery disease. While the patient truly has cancer, it should be apparent that the diagnosis of cancer does not help this individual in any way, as he is likely to die of some other cause before the cancer, even untreated, would likely cause any symptoms or death. Many physicians would advise against giving this patient a PSA test in the first place, so as to avoid this overdiagnosis and all of the expensive and potentially dangerous follow-up that the overdiagnosis could engender. This is the underlying logic of the Society for Post-Acute and Long-Term Care Medicine’s Choosing Wisely recommendation to avoid not just prostate cancer screening, but also breast and colorectal cancer screening in individuals with life expectancies shorter than 10 years.

In addition to this PSA example, several other Choosing Wisely recommendations explicitly try to prevent overdiagnosis in areas other than cancer. The AAAAI’s recommendations against large immunoglobulin G and IgE allergy panels and IgE testing without a history consistent with IgE-related food allergy is clearly rooted in the propensity of these tests to yield several questionable positive results, or overdiagnoses, that have no clinical consequence. For example, if a patient who enjoys strawberries is being evaluated for a nut allergy and an IgE against a strawberry antigen is discovered due to large IgE panel testing, it is unlikely that this patient has a serious risk for anaphylaxis from strawberries. Neither the inconvenience of strawberry avoidance nor the expense of any therapeutic intervention aimed at inducing tolerance to strawberries would be warranted. While this seems logical, it has been, in the author’s experience, rather challenging to communicate this risk of testing to ordering providers and laboratories successfully, especially when financial incentives support the decision to test.

Efficacy of Choosing Wisely

At its inception, the Choosing Wisely initiative promised many benefits. It had the potential to increase awareness, amongst patients and providers alike, of the problem of resource overutilization in medicine. It also held the promise to reduce, directly or indirectly, some of the practices that were called out in its many recommendations. Finally, one might have hoped that it would have some positive effect on the so-called triple aim of healthcare, which is to improve the patient experience of care, improve the health of populations and reduce the cost of healthcare. Whether or not Choosing Wisely has done any of these things has been debatable, in large part due to the fact that the initiative does not include explicit aims focused on measurement or monitoring of efficacy. Nonetheless, the peer-reviewed literature now contains several studies of Choosing Wisely’s efficacy in targeted areas.

A 2015 study demonstrated that services provided to members of a large US insurer followed trends that were partially in accord with Choosing Wisely recommendations, namely those concerning imaging for headache and cardiac imaging, indicating that a temporally related, although questionably causal, relationship could be observed between Choosing Wisely and ordering practices. However, other services that Choosing Wisely recommended against actually increased, such as human papilloma virus testing in young women, and other services recommended against remained common and unchanged in frequency, such as preoperative chest X-rays [12], so the positive effect observed in the two imaging categories may have been a statistical fluke or an isolated effect that was not applicable to other areas. Another study in 2015, however, showed a much more positive effect of Choosing Wisely. In this study [13, 14], inspired by the Canadian and US Choosing Wisely campaigns, the testing requisition for the Canadian province of Alberta was changed so that vitamin D testing was not permitted without the ordering physician providing an indication of medical necessity. This intervention led to a 92% reduction in vitamin D testing and an estimated $4 million USD in projected annual monetary savings. The contrasting results of these two studies indicate that passive adoption of Choosing Wisely recommendations is weak and maybe nonexistent, but that targeted implementations of recommendations can be highly effective.

A 2017 US study [15] assessed multiple interventions undertaken at a large military hospital aimed at reducing unnecessary laboratory testing, primarily related to Choosing Wisely recommendations targeting daily routine testing. Interventions included modifications of repetitive inpatient orders for routine tests and displaying laboratory test costs (US Medicare reimbursement rates) to ordering providers, both of which were found to lead to sustained reductions in routine test ordering, by 19.4% and 15.3%, respectively. This study demonstrates the power of using multiple, linked interventions coupled with computerized order set modifications to effect lasting change in ordering behaviors, and also evidence that a targeted effort to implement Choosing Wisely recommendations can succeed.

Another study, from the Netherlands [16], similarly focused on Choosing Wisely recommendations by choosing to implement processes inspired by the American College of Rheumatology’s antinuclear antibody (ANA) testing recommendation. Through a combination of education and provision of non-individualized feedback on test ordering patterns, the authors reported a significant decrease in ANA test orders from the targeted rheumatologists, significant decreases in repeat ANA test orders and decreased variation in test ordering practices between rheumatologists at one site in the study. While small (the study only included 29 total rheumatologists) and only focused on patients who actually received ANA tests, it demonstrates that education paired with feedback can be a successful strategy for reducing test utilization. One weakness of this study is that it was not able to demonstrate that the reduction in ANA test volumes was entirely due to reductions in unnecessary tests.

Criticisms of Choosing Wisely

Criticisms of Choosing Wisely are rooted in the historical perspective that the initiative began only a few years after the passage of the US Affordable Care Act, also known as “Obamacare”. This was a time of great focus on healthcare economics, as well as significant disagreement between political parties in the US regarding how to solve problems in the US healthcare system. At its core, Choosing Wisely presented (and still presents) a political target for proponents and beneficiaries of the current fee-for-service based US healthcare system, as lowering healthcare costs comes with a necessary side-effect of lowering the revenue of those who perform the services and receive the fees. Other critics have warned that Choosing Wisely recommendations are difficult to follow and could interfere with the doctor-patient relationship [17], that patients may not be able to understand the recommendations [18], and that the entire initiative is about saving money and not about patient care [19]. One might also wonder how enforceable simple recommendations could ever be, and likewise question the effectiveness of recommendations that no one is required even to read, much less follow. While peer pressure may have induced several professional societies to participate in the program, it is unclear whether or not peer pressure alone might also curb test overutilization by the members of those same professional societies. While no attempt to refute any of these arguments will be made here, it is important to understand the criticisms in the context of Choosing Wisely’s origin in the US. The US’s outsized healthcare expenditures, the US citizenry’s discomfort with the thought of “less is more” and ignorance of the fact (or unwillingness to believe) that US healthcare is not actually the best in the world, the lack of a centralized healthcare model in the US, the related lack of a single central review body in the US that has the authority to assess the efficacy and utility of medical practices and the US’s current political climate that explicitly prioritizes neither evidence nor expertise in decision-making all work together against initiatives in the US that would seek to reduce healthcare services even when trusted groups say that those services are unnecessary.

Besides criticisms aimed at the purpose of the initiative, additional criticisms of Choosing Wisely have been aimed at the methodology used to create the recommendations. A study comparing Choosing Wisely US recommendations to German S3 Guidelines [20] found that of the different participants’ “Top Five” recommendations in 2015, only 131 of 412 were found to be “trustworthy” according to assessments used to grade evidence-based guidelines. Only 75 recommendations had equivalents in the German guidelines, but several others were found to have sufficient evidentiary support cited in the recommendation text. This analysis does not reach a surprising result, as Choosing Wisely never intended its recommendations to rise to the standards expected of formal clinical guidelines, but the question remains as to what clinicians and patients should do with the 281 recommendations that did not meet criteria for trustworthiness.

As to the effectiveness of Choosing Wisely, the lack of compelling evidence of passive adoption of recommendations has been cited above. The best evidence in support of Choosing Wisely would be comprehensive studies of implementation and outcomes across many sites, but while frameworks have been proposed for monitoring the success of the initiative [21], perhaps one of the most incisive criticisms of Choosing Wisely is that so little has actually been done to date to measure the overall effect of all of the campaign. Taking a cynical position, the existence of the Choosing Wisely “Champions” program within the initiative (http://www.choosingwisely.org/success-stories/champions/) may paradoxically reflect this problem, as the initiative’s focus on local success stories could partly be due to the paucity of evidence of a more global benefit. The fairest assessment of this issue, however, is probably that the initiative is still young, that local successes are still successes and that quality improvement takes time and effort. While little exists today to demonstrate a global benefit from Choosing Wisely, there have been many undeniable and meaningful positive changes effected on local scales that are clearly tied to the initiative.

Conclusions

The US Choosing Wisely initiative represents a new tactic in a decades-long struggle to control the overutilization of medical tests and procedures. Through use of a patient-focused website and an underlying model that is flexible and that encourages wide participation and commonsense recommendations, the initiative aims to support doctors, patients and healthcare systems in their efforts to reduce the quantity of unnecessary medical care delivered in the US. Success stories abound, with individuals and healthcare systems demonstrating substantial savings derived from reductions in overutilization when targeted interventions are undertaken. Other studies indicate modest or no discernible improvements in utilization, indicating that there is significant variation in the effectiveness of passive adoption of Choosing Wisely recommendations. As is often the case with quality improvement projects undertaken within an organization, the overall outcome benefits of Choosing Wisely will be difficult to demonstrate because randomized and controlled experiments are simply unfeasible. Nonetheless, it is hard to imagine how an effort targeted simply at raising awareness among physicians and patients of common sense approaches to rational utilization of services in healthcare cannot be regarded as an overall good, and indeed as another helpful addition to the laboratory test stewardship toolbox.