The use of low-value care is more than just a small problem in the United States, the Institute of Medicine indicated an estimate of $765 billion wasted health care expenditures in 2013 (Figure 1). Solutions to measure, identify, and eliminate low-value care, however, are challenging and complex. To date, most efforts aimed at reducing low-value care, such as the Choosing Wisely Initiative, have been limited to areas where there is high degree of consensus that the care rendered is low value.

In order to get a better understanding of the complexities of low-value care, we conducted interviews with 13 people representing a breadth of perspectives including clinicians from professional medical societies, payers, employer benefit design professionals from large organizations, and expert health economists. The interviews were funded by the National Pharmaceutical Council and conducted by the Altarum Institute. We focused on overuse/overtreatment, failures of care delivery and coordination, and pricing failures. We proposed working definitions of low-value care and explored how participants considered defining and measuring low-value care.

Based upon these interviews, there was a profound lack of consensus on how to incorporate clinical nuance, patient preferences and priorities, and cost-benefit tradeoffs into provider and consumer-facing initiatives to reduce low-value care.

One employee benefits executive commenting on competing risks stated, “I think this starts to tread into some really difficult, again, ethical territory. I’m not sure how we would judge based on a person’s current health status. If its benefit’s decision to say somebody in this situation doesn’t get access to the service, I don’t think we would ever do that.”

However, a clinical professional from oncology was in favor of using competing risks in decision making and commented “A treatment that might be considered as appropriate treatment for a patient with few comorbidities might be considered to be a very poor choice for a patient who’s got significant competing risk.”

Other conflicting findings include:

  • There is a healthy skepticism of consensus methods, such as Choosing Wisely, in identifying low-value care beyond obvious “low-hanging fruit.” There was not clear consensus on how to use components such as patient preferences, health-related quality of life, or competing risks and risk-benefit tradeoffs (e.g., at what age to initiate and terminate certain preventive screenings) to define and measure low-value care.
  • There was little consensus on the validity, practical application, and priority for using cost-effectiveness analysis to inform coverage and pricing decisions.
  • Although there was agreement that price and cost should be included, particularly unjustified price variation, (e.g., price at freestanding versus facility-adjoined surgical or infusion therapy center) in efforts to reduce low-value care, there was not clear agreement on how that should be done.

Where Can We Agree?

Our panelists concurred on elements of low-value care that need immediate attention. These areas have a role for administrators and clinicians to begin reducing low-value care (Figure 2).

Medical Errors

Medical errors are of utmost priority and obvious low-value care that should be addressed separately and immediately. Many contended little has been done to substantially and systematically reduce medical errors while others suggest some progress. Participants suggested that both the challenges and solutions for eliminating medical errors (e.g., improving care processes and reducing care fragmentation) were sufficiently distinct from that needed to reduce overuse that it requires separate prioritization and initiatives. Administrators and clinicians should continue to focus attention and resources on the reduction of medical errors.

Pricing Failures

Pricing failures presented a general concurrence that unjustified price variation warranted further examination. Participants generally believed that given reliable price and quality information, plus aligned incentives, challenges in and of themselves, clinicians and patients could effectively address many of these issues where prices inexplicably vary. Where prices are, in some judgments, just simply and consistently too high and create low value, separate initiatives are necessary.

We suggest administrators look to the removal of regulatory barriers to value-based contracting between biopharmaceutical companies and payers. Additionally, administrators and clinicians should look to the use of defined contribution benefit design and centers of excellence contracting to channel patients to clinicians who provide high-quality care and are willing to discount their prices in exchange for the higher volume of patients.


Overuse and overtreatment could generally be divided into care that nearly always causes more harm than benefit independent of patient, price, timing, or provider and care that may be appropriate in certain circumstances, but not in others. Overuse/Overtreatment is sufficient in prevalence and cost that immediate efforts to better understand what does and does not work to measure and reduce this low-value care are imperative. Concurrently, initiatives to apply elements of clinical nuance to definitions and measures should commence to further identify and reduce overuse.

In addition, we propose the following actions:

  • Care that nearly always causes more harm than benefit independent of patient, price, timing, or provider. We suggest administrators and clinicians look to incorporate Choosing Wisely, USPTF, and similar lists into performance and quality measurement for alternative payment programs.
  • Care that that may be appropriate under certain circumstances. Current low-value care lists should be expanded beyond nearly universal low-value care to include items where value is circumstance dependent. Given concerns about limiting access to appropriate care, we recommend a tiered approach that takes into account both the magnitude and certainty of low-value care use. This effort would need to incorporate patient protections.

From our interviews, reducing low-value care with interventions such as using cost effectiveness analysis, limiting access to care due to competing risks, and denying low-value care where patient preferences and priorities request such care, was generally a low priority. Administrators deferred to clinicians, and clinicians generally agreed that shared decision making, done well, which included cost conversations, can address many of these grey areas and reduce low-value care sufficiently and appropriately.

Another panelist when discussing high-cost interventions said, “I know I wouldn’t be comfortable having specific dollar thresholds for any particular service, certainly for very high-cost specialty medications and other services. We just have to — where those high costs are there, we just have to do our very best at making sure they’re being used as appropriately as possible, so that there’s actually a benefit.”

Participants felt that there is an important role for consumers to play in reducing low-value care. Consumers can become activated patients and engage in well-designed, well-conducted shared decision making conversations with their clinicians to choose high-value care.

Beginning to understand where there is consensus on how to define and measure low-value care is an important step in efforts to reduce the clinical harm and the spending associated with low-value care. Our findings suggest that there is enough agreement in specific areas to begin substantial efforts to decrease low-value care, even though some disagreement remains on aspects of how to define and measure the subject. These findings are relevant to administrators, policymakers, researchers, and consumers as they highlight areas that will be useful to focus on in order to begin reducing low-value care.

Eliminating low-value care creates several opportunities to enhance the patient experience, improve quality, and lower costs. Potential uses of the immediate savings from reducing low-value care include:

  • expand access to insurance for the 33 million uninsured Americans;
  • offer enhanced coverage for established high-value care and clinical innovations;
  • provide funding for social determinants of health; and,
  • establish the means to provide the long-term care, services, and supports that frail elders will require in the coming years.

Given the large dollars associated with low-value care, and the opportunity to spend that money on more worthwhile initiatives, we argue that awaiting perfect definitions and measures cannot get in the way of implementing initiatives to reduce patient harm and wasteful spending. If we do not take immediate action to drive down low-value care, patient-centered outcomes will suffer and the likelihood of alternative approaches to slow growth in health care expenditures will be reduced.

Figure 1. Wasted health care expenditures in 2013


Figure 2. Policy Approaches To Address Low-Value Care

Medical Errors
Eliminating medical errors (e.g., improving care processes and reducing care fragmentation) is sufficiently distinct from that need to reduce overuse; therefore it requires separate prioritization and initiatives
Price Failures
Remove regulatory barriers to value based contracting between biopharmaceutical companies and payers
Use defined contribution benefit design and centers of excellence contracting to channel patients to clinicians who provide high quality of care and willing to discount their prices in exchange for the higher volume of patients. It is appropriate for services for which there exists substantial variation in price, but only limited variation in quality
Overuse/Overtreatment: Care that nearly always causes more harm than benefit independent of patient, price, timing, or provider
Incorporate Choosing Wisely, USPTF and similar lists into performance and quality measurement for alternative payment programs
Overuse/Overtreatment: Care that that may be appropriate under certain circumstances
Expand Choosing Wisely, USPTF and similar lists beyond nearly universal LVC to include items where value is circumstance dependent:

o Develop a tiered list of LVC items in which tier placement is determined both by magnitude of LVC use and certainty that an item is considered LVC

o Reduction of circumstance dependent LVC should be incentivized by using stepped care, VBID principles to encourage alternatives, close monitoring of doctors use of these questionable services coupled with direct feedback, and bonuses to doctors for keeping that use below a threshold. Incentives to reduce LVC should be tier dependent.

o Robust set of patient protections, such as appeals process and use of prior authorizations that require human review, will be used to ensure access to these services when appropriate.
Increase independent, publicly disseminated comparative effectiveness research to include more direct comparisons of alternative therapeutic and surgical interventions to include comparisons of alternative medications and devices.
Leverage large US investment in EHRs to prioritize targets for reducing LVC