To assess the role of value-based payment (VBP) in improving fidelity and patient outcomes in community implementation of an evidence-based mental health intervention, the Collaborative Care Model (CCM). 

 Retrospective study based on a natural experiment. 

 We used the clinical tracking data of 1806 adult patients enrolled in a large implementation of the CCM in community health clinics in Washington state. VBP was initiated in year 2 of the program, creating a natural experiment. We compared implementation fidelity (measured by 3 process-of-care elements of the CCM) between patient-months exposed to VBP and patient-months not exposed to VBP. A series of regressions were estimated to check robustness of findings. We estimated a Cox proportional hazard model to assess the effect of VBP on time to achieving clinically significant improvement in depression (measured based on changes in depression symptom scores over time).     

Estimated marginal effects of VBP on fidelity ranged from 9% to 30% of the level of fidelity had there been no exposure to VBP (P <.05 for every fidelity measure). Improvement in fidelity in response to VBP was greater among providers with a larger patient panel and among providers with a lower level of fidelity at baseline. Exposure to VBP was associated with an adjusted hazard ratio of 1.45 (95% confidence interval, 1.04-2.03) for achieving clinically significant improvement in depression.

VBP improved fidelity to key elements of the CCM, both directly incentivized and not explicitly incentivized by the VBP, and improved patient depression outcomes.

Value-based payment (VBP), when embedded in implementation initiatives, has the potential to improve fidelity to evidence-based care and patient outcomes.

VBP may improve fidelity to evidence-based care by closely mapping quality targets with key elements of the evidence-based model. 

Managed care organizations may consider using VBP as a strategy to enhance the effectiveness of implementation of evidence-based care among provider networks.

Additional support may be needed to help providers with a smaller patient panel to adequately respond to and benefit from VBP; innovations of VBP design, such as a multi-tiered system, may be needed to provide sufficient incentives to providers with higher performance at baseline.

Despite extraordinary increases in medical knowledge, healthcare in the United States frequently falls short of evidence-based standards.

 Implementation fidelity is defined as the degree to which interventions or programs are implemented as intended by the program developers

; and poor implementation fidelity is one explanation of why the promise of evidence-based medicine remains unfulfilled.

 Substantial variation among providers exists in the intensity of implementation and degree of fidelity to the evidence base,

 and current financial incentives in the US healthcare system contribute to a poor “business case” to adopt evidence-based practices in an effective manner.

 Targeted financial incentives have the potential to improve fidelity and improve implementation effectiveness.

Value-based payment (VBP)—a form of pay for performance&mdash;incentivizes quality and outcomes of care by tying payment to providers with predefined quality or efficiency targets. VBP has been widely adopted by private and public payers, with recent examples including Medicare’s launch of the hospital VBP program in 2013

 and the HHS secretary's announcement of measurable goals and a timeline to move the healthcare system at large toward quality-based payment.

 Most existing VBP programs provide standalone financial incentives without a support system to providers for care process redesign. VBPs, embedded in evidence-based care and designed to improve implementation effectiveness, are rare

In this study, we assessed if a VBP component could improve the effectiveness of Collaborative Care Model (CCM) implementation. The CCM is a team-based approach to treating depression and other common behavioral health conditions in primary care,

 with the team including a primary care physician, care manager, and a consulting psychiatrist. The key principles of the model include systematic follow-up on patients by the care manager, measurement-based care that uses symptom-rating scales to track clinical improvements or identify patients not improving,

 where treatment is systematically adjusted or intensified by the primary care team (with input from the consulting psychiatrist) for patients not improving.

 CCM implementation has gained momentum: in its proposed rules for the 2016 Physician Fee Schedule, CMS stated an intention to modify current payment to cover the CCM.

We conducted our assessment in the context of the Mental Health Integration Program (MHIP) of Washington state.

 The MHIP is an ongoing, publicly funded implementation of CCM in a diverse network of community health clinics across Washington. Started in 2008 in the 2 most urban counties in the state, it is now statewide and has served over 35,000 individuals. The VBP component of the MHIP payment started in 2009 in response to substantial variation in quality of care and patient outcomes seen in the first year. It adopted some best practice design features of VBP

: close mapping of quality measures with key elements of the evidence-based model, substantial incentive payment size (25% of total payments to providers), and a dynamic set of quality measures and targets that are adjusted over time. Thus, a unique opportunity is offered to assess VBP's role in improving implementation effectiveness of evidence-based care in community settings.

We hypothesized that MHIP VBP improved fidelity to the CCM, as measured by key process-of-care elements of the model, both directly and not explicitly incentivized by the VBP. We further hypothesized that MHIP VBP improved patient depression outcomes. Provider organizations with a larger patient panel have more at stake under a VBP scheme and greater resources to invest in quality improvement in response to VBP.

 Providers with a lower level of performance at baseline have more room and motivation to improve.

 We thus hypothesized that the effect of VBP on implementation fidelity was greater among clinics with a larger MHIP patient caseload and among clinics with a lower level of fidelity prior to VBP.

This study focused on phase 1 of the MHIP VBP. Phase 1 (covering year 2009) used 4 process-of-care targets mapping closely with the principles of the CCM (

). Provider organizations would initially receive 75% of their total payment for the CCM (ie, with 25% holdback) and receive 5% for achieving each of the 4 targets in a calendar quarter, with an additional 5% being awarded for participation. Adjustments to the VBP scheme were made in subsequent phases by raising benchmarks for existing targets, eliminating targets that had been achieved by most provider organizations, and/or adding new targets to address emerging gaps in quality, thus providing incentives for continuous improvement.

Our study population comprised patients 18 years or older who initiated care in the MHIP between January 1, 2008, and June 30, 2009, in 1 of the 35 community health clinics that started MHIP implementation in 2008—therefore, they had experience with MHIP both before and after the launch of phase 1 VBP. We restricted patient enrollment in the MHIP to June 30, 2009, to ensure that the first 6 months of care for every patient enrolled in 2009 were under the influence of phase 1 (not phase 2) of VBP. All 35 clinics were located in King and Pierce counties&mdash;the 2 most populous counties in Washington state&mdash;and were affiliated with 7 community health centers. These community health centers were the parent organizations of the clinics and were Federally Qualified Health Centers. The populations they served were primarily patients with Medicaid, other state-funded programs, and patients who were uninsured. Clinical social workers, psychologists, licensed mental health counselors, and other clinicians staffed at the clinics served as the CCM care managers in the MHIP.

Patient inclusion criteria included a baseline Patient Health Questionnaire ([PHQ-9] with a possible range of 0-27)

 score of 10 or greater—indicating clinically significant depression&mdash;and at least 1 follow-up contact with the MHIP care manager within 24 weeks of the initial contact in order to allow at least 1 chance to assess depression outcomes. Patients whose last contact with MHIP occurred within 1 week from the first contact were further excluded as they likely were determined ineligible for MHIP. The vast majority of the MHIP patients during our study period were enrollees in Washington state’s Disability Lifeline Program. These patients were temporarily disabled because of a physical or mental health condition and expected to be unemployed for 90 days or more. King County extended eligibility to additional patient populations including low-income mothers and their children, low-income older adults, uninsured, veterans, and veterans’ family members.

Our data were from the Web-based registry

 used by all MHIP participating clinics to systematically document care management activities and clinical outcomes and to assist with population management.

Three dichotomous measures at the patient-month level captured fidelity to major domains of the CCM (Table 1). “At least 1 follow-up contact with the care manager” reflects the principle of systematic follow-up; “at least 1 psychiatric consultation” reflects the principle of stepped care—the idea that treatment should be systematically changed or intensified for patients not responding to initial treatment. An important mechanism by which stepped care is operationalized in CCM is through consultation with a mental health specialist (usually a psychiatrist) for potential treatment changes. These 2 measures were closely related to the 2 quality targets in phase 1 of MHIP VBP (Table 1). “At least 1 PHQ-9 assessment” reflects the principle of measurement-based care, whereby treatment teams use symptom rating scales to systematically track clinical improvements or lack thereof. This measure was not explicitly incentivized in MHIP VBP. Data on current medications (also documented in the MHIP registry) are not available for research at this point. We therefore had no fidelity measure mapping the fourth VBP measure, documentation of current psychiatric medication in registry for 75% of cases (Table 1). Each fidelity measure was assessed for each 4-week interval starting from the patient’s initial contact with the MHIP care manager, up to 24 weeks or until the patient’s last contact with the care manager, whichever occurred first.

We tested hypotheses about 2 potential modifiers of the effects of VBP on fidelity: size of the MHIP patient panel, measured by cumulative number of patients treated at the clinic prior to phase 1 VBP (ie, in 2008); and clinic-level fidelity at baseline, measured by the average count of follow-up contacts, psychiatric consultation, or PHQ assessments over the first months of all patients treated at the clinic in 2008.

A clinically significant improvement in depression was defined as achieving a follow-up PHQ-9 score under 10 or achieving at least 50% reduction in PHQ-9

 within 24 weeks of initial care manager contact.

The rolling enrollment of patients in the MHIP throughout our study period created a natural experiment in the sense that exposure of patient episodes of care to VBP resembled random assignment. We compared the fidelity outcomes for patient-months exposed to VBP with patient-months not exposed to VBP. We estimated a multi-level linear probability model for each fidelity outcome with a random intercept at the patient level to account for clustering of months of the same patient. The key independent variable was a dichotomous indicator of VBP exposure, defined as 1 if the index patient-month started after January 1, 2009—when phase 1 VBP took effect&mdash;and 0 otherwise. Because care management activities were more intensive in the early months of a patient’s treatment episode,

 our adjusted analysis controlled for a set of dummy variables indicating whether the index month was the patient’s first through sixth month in MHIP. We conducted several sensitivity analyses. We restricted the sample to Disability Lifeline Program enrollees, which accounted for 85% of the entire sample. We also estimated the linear probability model with a fixed effect for each patient, first with the entire sample (ie, regardless of whether a patient had exposure to VBP), then restricting to patients who contributed months both pre- and post VBP during the first 24 weeks of care. The latter approach allowed us to examine the effects of VBP on treatment fidelity within the same patient, but was conducted with a much smaller sample size (about one-third of the original sample), and thus subject to lower precision in estimates. We also estimated the logistic version of each model and compared implications (eg, marginal effect of VBP) based on both sets of analyses.

Hypotheses regarding the modifying effects were tested by adding interaction terms between the VBP indicator and the hypothesized modifier to the models. We included an additional interaction between VBP and the quadratic term of the modifier (eg, cumulative caseload squared) to allow for nonlinear effects. The VBP incentives were directed at the community health centers (also known as “implementation sites” in MHIP)—7 in total in our study sample; each site had multiple clinics. We conducted sensitivity analysis where hypothesized modifiers (caseload and baseline fidelity) were measured at the site level.

To assess the association between VBP and depression outcomes, we estimated an extended Cox proportional hazard model of time to improvement in depression, censored at 24 weeks after the initial assessment/contact, or the patient’s last contact with the MHIP care manager, whichever occurred first. The key independent variable was an indicator of VBP exposure (ie, 1 after January 1, 2009, and 0 otherwise). This indicator varied over time (switched from 0 to 1) for patients who enrolled in MHIP in 2008, but whose observation period ended in 2009. We conducted tests of the proportional hazard assumption based on the Schoenfeld residuals.

Adjusted analyses controlled for baseline patient age and gender, baseline PHQ-9 scores and comorbid behavioral health conditions, MHIP eligibility categories, and clinic fixed effects (to control for between-clinic differences in quality of care). To control for possible clinic learning over time, we also included the number of months the clinic had been participating in MHIP at the time the index patient was enrolled in MHIP and its quadratic and cubic terms.

 presents descriptive statistics of baseline patient characteristics for the entire sample (n = 1806) and by whether a patient had at least 1 month of exposure to VBP within 24 weeks since their first contact with MHIP care managers. Patients with no exposure to VBP were more likely to be enrollees in the Disability Lifeline program than patients with at least 1 month of exposure (96.2% vs 79.7%). Partly because of this difference, patients with no exposure were more likely to be aged between 40 and 59 years and more likely to have a PHQ-9 score of 20 or higher (indicating severe depression symptoms) compared with patients without any exposure. Prevalence of comorbid behavioral conditions was comparable between the 2 cohorts, except that rates of anxiety and bipolar disorders were slightly lower among the no-exposure group than the exposed group.

For the fidelity outcomes, results of multivariate analyses were presented in the form of predicted probabilities with and without exposure to VBP and the marginal effect of VBP (

). (The mixed-effects linear probability models and their logistic counterparts generated very similar results; results reported hereafter and in Table 3 were based on the linear probability models.) Based on analyses conducted with the entire sample, the effect of VBP on the probability of at least 1 follow-up contact, psychiatric consultation, and PHQ

9 assessment in a month was an increase of 0.05 (95% confidence interval (CI), 0.00-0.10; 

<.05), respectively. The magnitude of the increase was about 9%, 30%, and 15% of the respective fidelity outcome had there been no exposure to VBP. Analysis restricting to Disability Lifeline enrollees (85% of the unrestricted sample) generated very similar results, with slightly smaller marginal effects of VBP for follow-up contacts and PHQ assessments, but slightly greater marginal effect for psychiatric consultation (Table 3).

Sensitivity analysis of the fidelity outcomes by controlling for patient fixed effects produced similar or slightly greater point estimates of the marginal effects of VBP (eAppendix Table [

]). However, fixed-effects analysis, restricted to patients who received care both before and after VBP, had a much-reduced sample size (359 patients compared with 1806 in the unrestricted fixed-effects analysis). Marginal effects of VBP were not statistically significant for follow-up contacts or psychiatric consultation, but remained strong for PHQ assessments (marginal effect of VBP: 0.09; 95% CI, 0.02-0.16).

Our analysis indicated that both the size of the MHIP caseload at a clinic and the level of fidelity prior to VBP modified the effect of the VBP. As shown in eAppendix Figure A, for follow-up contacts and PHQ assessments, the marginal effect of VBP increased with the number of patients treated at the clinic prior to VBP. For follow-up contacts, the marginal effect of VBP did not achieve statistical significance until the number of patients treated at clinic in 2008 was 100 or more (top 25% of clinics); and for PHQ assessments, not until at least 140 (top 10% of clinics). Caseload did not seem to modify the VBP effect for psychiatric consultation. On the other hand, for each fidelity measure, the marginal effect of VBP decreased with the level of fidelity at the clinic prior to the start of VBP (eAppendix Figure B). For example, the effect of VBP on follow-up contacts was significantly greater than 0 only among clinics whose first-month follow-up contacts in 2008 averaged below 0.8 (accounting for 75% of all clinics). Sensitivity analysis defining modifiers at the implementation site level produced largely consistent findings. One exception was that, for psychiatric consultation, the VBP effect seemed to decrease with the size of MHIP caseload at the site level, whereas there was no clear modifying effect of caseload defined at the clinic level.

Consistent with results for the fidelity outcomes, exposure to VBP was associated with an adjusted hazard ratio (HR) of 1.45 (95% CI, 1.04-2.03) for achieving clinically significant improvement in depression, indicating that exposure to VBP was associated with a shorter time to improvement. This result held when we restricted the sample to Disability Lifeline patients (adjusted HR, 1.47; 95% CI, 1.03-2.12).

With data from a statewide implementation of the CCM in community health clinics, we found that a VBP program embedded in community-based implementation improved fidelity to several key process-of-care elements of the evidence-based model, both directly incentivized and not explicitly incentivized by the VBP. Consistent with our hypotheses, we also found stronger responses to VBP among provider organizations that cared for a larger number of patients and among organizations with a lower level of initial fidelity. Finally, we found that VBP led to better patient outcomes indicated by a shorter time to clinically significant improvement in depressive symptoms. 

Our VBP effect findings contrast with the limited evidence supporting the effectiveness of existing VBP programs.

 Several reasons may underlie the differences. First, the MHIP paired financial incentives with chronic care quality improvement and capacity-building efforts. An expert team at the University of Washington provided training to care managers from all participating community health clinics and made archived training materials available online, and consulting psychiatrists were arranged in a contractual relationship to work with all participating clinics. Implementation of a clinical tracking system—a crucial tool to enable population health management and case tracking&mdash;was a precondition for receiving funding for MHIP and was achieved at all clinics. Meanwhile, existing VBP contracts typically provide no support system for quality improvement. Second, the MHIP VBP targeted several key elements of a single evidence-based care model, focusing improvement efforts and sending strong signals and clear directions to provider organizations on what to improve. Existing VBP programs typically contain a large number of quality targets that may not be clinically meaningful, dissipating incentives and failing to engage clinicians.

Monthly PHQ assessments—the measure not explicitly incentivized under MHIP VBP&mdash;improved by 15% in response to VBP. Because PHQ assessment was conducted at follow-up contacts with the care manager, incentivizing systematic follow-up may have had the “spillover effects” of incentivizing these assessments. A direct implication is that, for fidelity/process-of-care measures that complement one another, designers of VBP programs may consider recognizing some, but not all of them, as VBP targets. Keeping the target set parsimonious (and thus not diluting incentives) may not need to come at the price of forsaking important quality goals.

Clinics with a larger patient caseload responded to VBP more in 2 of 3 fidelity measures considered, suggesting that smaller clinics may perceive insufficient incentives because of the limited scope of their VBP payment

 and/or may lack the resources to make systematic changes to care in response to VBP. To ensure that provider organizations of all sizes (and their patients) benefit from VBP, implementation initiatives may consider pooling resources, for example, by establishing learning collaboratives and providing coaching and consultation to organizations in need. Consistent with our hypothesis and findings of previous studies,

 lower baseline fidelity was associated with greater improvement in fidelity in response to VBP, thus reducing the variation in fidelity/quality among provider organizations implementing the evidence-based model. Although a desirable outcome, it also reveals the fact that, with a single performance target for all providers, high-performers may not be adequately motivated to improve further even though there is still room for improvement. One option would be to have 2 sets of thresholds to be applied to provider organizations with different initial levels of performance.

 This option, however, adds to the complexity of the VBP and may be perceived unfair or unacceptable, especially by high-performing provider organizations.

We assessed the effects of VBP in a natural experiment, not a randomized trial. Although we controlled for important differences in patient baseline characteristics, differences among clinics that do not change over time (with clinic fixed effects), and proxies for provider learning over time, these controls were not perfect. However, a series of sensitivity analyses demonstrated the robustness of our findings. The fidelity and patient outcomes we examined were subject to data availability and usability; we were not able to examine antidepressant management as an important fidelity outcome at this point, but intend to do so in the future when data become available. We used data from the implementation of a specific care management model in a single state, which potentially limited the generalizability of our findings to other evidence-based care approaches or other geographic areas. This limitation, however, is mitigated by the fact that the CCM is highly consistent with the Chronic Care Model

 and that the MHIP involved a large and diverse set of provider organizations.

 Department of Healthcare Policy and Research (YB, AAE), and Department of Psychiatry (YB), Weill Cornell Medical College, New York, NY; Department of Health Care Policy, Harvard Medical School (TGM), Boston, MA; Department of Psychiatry and Behavioral Science, School of Medicine, University of Washington (YFC, JU), Seattle, WA; Department for Health Management and Policy, School of Public Health, University of Michigan (AMR), Ann Arbor, MI; Department of Psychiatry and the Dartmouth Institute of Health Policy and Clinical Practice, Geisel School of Medicine at Dartmouth University (MLB), Lebanon, NH; Department of Psychiatry, Columbia University Medical Center (HAP), New York, NY; New York-Presbyterian Hospital (HAP), New York, NY; Community Health Plan of Washington (EH), Seattle, WA.

 This study is supported by the National Institute of Mental Health (1R01MH104200; YB, TGM, YFC, AAE, AMR, MLB, HAP, JU). 

 Drs Chan and Unützer received salary support from Community Health Plan of Washington for training, clinical consultation, and quality improvement efforts related to the Mental Health Integration Program. Ms Hafer is employed by Community Health Plan of Washington, the sponsor of the care management program studied in the paper. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article. 

Concept and design (YB, TGM, YC, AMR, MLB, EH, JU); acquisition of data (YB, YC, AAE, EH, JU); analysis and interpretation of data (YB, TGM, YC, AMR, MLB, HAP, EH, JU); drafting of the manuscript (YB, TGM, AMR, MLB, EH, JU); critical revision of the manuscript for important intellectual content (YB, TGM, YC, MLB, HAP, EH, JU); statistical analysis (YB, TGM); provision of patients or study materials (EH, JU); obtaining funding (YB, JU); administrative, technical, or logistic support (AAE); and supervision (YB, JU).

 Yuhua Bao, PhD, Weill Cornell Medical College, 402 E 67th St, New York, NY 10065. E-mail: yub2003@med.cornell.edu.

Our study provided strong evidence that a VBP component adopting best practices of VBP design and being embedded in an implementation initiative is effective in improving fidelity to key elements of the evidence-based model, both directly and not directly incentivized by the VBP, and, in turn, improving patient outcomes. 

1. Blumenthal D, McGinnis JM. Measuring Vital Signs: an IOM report on core metrics for health and health care progress. 

 2015;313(19):1901-1902. doi: 10.1001/jama.2015.4862.

2. Dusenbury L, Brannigan R, Falco M, Hansen WB. A review of research on fidelity of implementation: implications for drug abuse prevention in school settings. 

3. Chin MH, Cook S, Drum ML, et al; Midwest Cluster Health Disparities Collaborative. Improving diabetes care in midwest community health centers with the health disparities collaborative. 

4. Landon BE, Hicks LS, O’Malley AJ, et al. Improving the management of chronic disease at community health centers. 

5. Rubenstein LV, Jackson-Triche M, Unützer J, et al. Evidence-based care for depression in managed primary care practices. 

Articles

POLICY

Value-based payment improved fidelity to key elements of the Collaborative Care Model&mdash;an evidence-based mental health intervention&mdash;and improved patient depression outcomes in Washington state.

Value-Based Payment in Implementing Evidence-Based Care: The Mental Health Integration Program in Washington State

In this article, we consider the problem of financing highly effective and cost-effective prescription drugs within a value-based pricing system. Precisely because these drugs are highly effective, their value-based prices may be quite expensive; and moreover, the value-based price of a cure ought to be set high enough to create incentives for innovation, otherwise these beneficial therapies may be underdeveloped. However, in our fragmented health insurance system, where patients move frequently between payers, these payers generally lack the incentives to pay value-based prices for cures because they cannot ensure that they will reap the long-term economic benefits. Therefore, we argue that there is a need for mechanisms to spread the burden of financing of cures across payers to maximize patient access and the public good. We suggest that risk adjustment, reinsurance, and risk corridors are familiar policy options that merit consideration to address the problem and create incentives for value-based pricing. 

	Exorbitant prescription drug prices are rightly being scrutinized, and several promising proposals to lower prices have been advanced. For the small subset of prescription drugs that are cost-effective cures, however, the policy challenge is not simply to lower prices.

		Prices of these extremely effective therapies ought to be value based and high enough to incentivize development.

		Even value-based prices may be too expensive for private payers that cannot realize long-term economic benefits of cost-effective cures because of patient churn.

		Despite their flaws, the “3Rs”— risk adjustment, reinsurance, and risk corridors&mdash;are familiar policy options that merit consideration to address this problem.

The pricing of prescription drugs has recently received considerable attention—and rightly so. This scrutiny follows widely publicized decisions by drug manufacturers to charge exorbitant amounts for their products; seemingly excessive drug prices, however, are not all created equal. Rather, the legal, regulatory, and market structures enabling high prices may differ from product to product&mdash;and likewise, the policy challenges posed by high drug prices may differ. Paying for the subset of expensive but highly effective cures poses particularly difficult challenges, including the lack of incentives for health plans to cover these expensive treatments.

Many high prices are a result of market power from patent protection, regulatory exclusivity, the usual healthcare market forces in the United States, or some combination thereof. These high prices may come about through steep mark-ups. For example, Turing Pharmaceuticals abruptly raised the price of pyrimethamine—a decades-old drug lacking competing products&mdash;from $13.50 to $750 per pill, shortly after acquiring the product. And Mylan has been scrutinized for increasing the price of a 2-pack of EpiPens from $100 in 2007 to over $600 in 2016 before the expected arrival of a generic. In other instances, manufacturers have steadily&mdash;and more gradually&mdash;raised prices for newer brand name drugs that they developed. As one example, Novartis increased the price of its cancer treatment, imatinib, from approximately $30,000 per year in 2001 to $92,000 per year in 2012. For all of these therapies, policies primarily aim to reduce prices through introducing competition or price controls.

In a small subset of cases, however, highly effective and novel therapeutics are introduced on the market with seemingly astronomical, but also arguably cost-effective, initial prices. The clearest example may be sofosbuvir-based drugs, hailed as cures for hepatitis C virus (HCV) infection and priced by Gilead Sciences, Inc, at roughly $1000 per pill, or $84,000 for a 12-week course of therapy.

 However, sofosbuvir-based drugs are not the only products that may pose this problem; for instance, novel biologic therapeutics, such as gene and stem cell therapies, may one day offer cures for previously untreatable conditions, and similarly, may come with exorbitant—yet possibly cost-effective&mdash;price tags.

Policy Challenges for Expensive, Cost-Effective Cures

For this subset of novel therapeutics that are cost-effective cures, the policy challenge is not simply to lower the drugs’ prices. Prices of such therapies ought to be high enough to create incentives for innovation, and if set too low, beneficial therapies may be underdeveloped or undersupplied. Instead, the policy challenge is to ensure that payers have the incentives and resources to finance the therapies. In the United States, achieving this goal is particularly challenging because patients rarely stay with the same payer for more than a few months or years. Because of this churn within the fragmented health insurance system, payers are not likely to realize the long-term economic benefits of cost-effective cures.

To be clear, regardless of the reasons for high drug prices, the growth of overall spending on therapeutic products in the last decade—and the resulting strain on governmental, private payers’, and patients’ budgets&mdash;is deeply troubling.

 As a result, numerous reforms to address drug affordability have been advanced. At the state level, legislation has been proposed that would require drug manufacturers to make transparent the costs of drug development, cap prices for public payers in a state, or permit the importation of cheaper foreign drugs.

 Other proposals advocate for change at the federal level, such as the federal government utilizing its “march-in” rights under the Bayh-Dole Act or negotiating for better drug prices for Medicare.

Among the numerous reforms to address drug affordability in the United States that have been proposed, a particularly promising one is to move to a value-based pricing system. In such a system, a drug’s price is based on its cost-effectiveness, calculated by the relative cost per unit of benefit—whether measured in quality-adjusted life-years (QALYs), future healthcare costs avoided, budgetary impact, or other metrics.

 The underlying goal is to ensure that the drug’s price is commensurate with its health benefits. Value-based pricing would help reduce drug prices where costs appear unwarranted by their relative benefits.

Nevertheless, even if value-based pricing and the other reforms are adopted, cures and highly effective therapies may continue to pose significant financing challenges. Their value-based prices may continue to be very expensive precisely because they are so beneficial.

 Again, sofosbuvir-based therapies for HCV infection provide an instructive example. These drugs offer sustained virologic response rates greater than 95% in most patients—meaning they are highly effective&mdash;and are associated with a shorter duration of treatment and fewer adverse effects than previous HCV treatments.

 Because of the clear benefits of sofosbuvir-based treatments, they have been found to be cost-effective for a majority of patients even at their current prices.

 Furthermore, therapies that treat serious conditions and are more effective and safer than existing treatments—similar to sofosbuvir-based products&mdash;are precisely the kind of novel therapies that we want to encourage, and prices should reflect that.

The value-based prices of highly effective therapies may thus be unaffordable for many individuals without financing mechanisms that spread the cost across large populations. In this way, financing cures is no different than financing other types of important, but expensive, types of healthcare, such as open-heart surgery. In the United States, the basic mechanism to spread the financial cost of expensive healthcare services and therapies is insurance. However, private health insurers are not built to account for the long-term cost-effectiveness of a given therapy because patients frequently move between payers. Payers that bear the high short-term costs of a cost-effective cure, therefore, will rarely realize the long-term economic benefits. This inability to capture the financial benefits dampens payers’ incentives to pay even the value-based price of cures.

 To the contrary, health insurers may try to avoid costly patients by structuring drug formularies to dissuade such patients from choosing the plan, thereby engaging in the type of health-based discrimination the Affordable Care Act (ACA) was designed to prevent.

The benefits of cost-effective cures are “public goods”—in other words, the payer that bears the cost cannot exclude others from realizing the economic benefits of the cure because of churn in the US insurance market.

 The challenge is how to privately finance the public good of high-priced cures. In the absence of a considered solution, the de facto response is for governmental and private payers to limit access to the therapy in a type of rationing based on severity of illness and ability to pay. Perversely, rationing such therapies may result in avoidable excess morbidity, mortality, and healthcare spending that is distributed unevenly, and perhaps unjustly. To realize the hopes for highly effective, novel therapies in the United States, it is therefore critical to explore policy options for spreading the cost of cures.

This policy challenge also arises in other contexts in which the risks of high-cost patients must be spread across disparate payers—and risk adjustment, reinsurance, and risk corridors (the “3Rs”) are actuarial solutions to this problem. Under a risk adjustment program, all payers pay into a fund that will compensate those payers that incur unusually high costs. Reinsurance is an insurance policy that insurers buy to protect against excess financial risk. Risk corridors limit both downside risk of losses and excess profits for health plans. The 3Rs have been used in the ACA and Medicare Part D, to protect insurers against adverse selection and consumers against destabilization of the insurance market and discriminatory health insurance practices.

A similar program could be implemented for the high costs associated with paying for cost-effective cures in private health plans. Incentives for value-based pricing could be incorporated into the program by limiting program coverage to those therapies that are both highly cost-effective and priced using value-based methodologies. Although the ACA’s 3Rs have faced significant political and legal challenges, and the fate of the ACA is uncertain, these programs have had some success in Medicare Part D to stabilize drug plan premiums and insurer participation, and should be considered.

 Despite their limits, they are familiar models that can help address the challenges for financing cost-effective cures. Other policies to address the disjunction between the immediate expenses and delayed benefits of high-cost cures, such as credits or financial transfers from future payers, remain theoretical but also should be explored further.

Because the value-based price for cost-effective cures may be high, new drugs and biologics that hold the promise of cures may pose challenges for a value-based pricing system—even if currently proposed reforms are enacted. In the fragmented health insurance system, where patients move frequently between plans, payers lack incentives to pay even value-based prices for cures because they will not reap the long-term benefits. However, if prices drop below value-based thresholds, incentives for innovation will diminish. We, therefore, need to explore mechanisms to spread the burden of financing of cures to maximize patient access and the public good. Value-based pricing is essential, but without interventions to spread the cost of cures, beneficial and cost-effective therapies may still be undersupplied and out of reach for many patients. 

 Georgia State University College of Law (ECFB, PJZ), Atlanta, GA.

 The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

 Concept and design (ECFB, PJZ); drafting of the manuscript (ECFB, PJZ); critical revision of the manuscript for important intellectual content (ECFB, PJZ).

 Erin C. Fuse Brown, JD, MPH, Assistant Professor, Georgia State University College of Law, 85 Park Pl NE, Atlanta, GA 30303. E-mail: efusebrown@gsu.edu.

1. Kesselheim AS, Avorn J, Sarpatwari, A. The high cost of prescription drugs in the United States: origins and prospects for reform. 

. 2016;316(8):858-871. doi: 10.1001/jama.2016.11237.

2. Chhatwal J, Kanwal F, Roberts MS, Dunn MA. Cost-effectiveness and budget impact of hepatitis C virus treatment with sofosbuvir and ledipasvir in the United States. 

 2015;162(6):397-406. doi: 10.7326/M14-1336.

3. Scott CT, DeFrancesco L. Gene therapy’s out-of-body experience. 

2016;34(6):600-607. doi: 10.1038/nbt.3592.

4. Moreno GA, Mulligan K, Huber C, Linthicum MT. Costs and spillover effects of private insurers’ coverage of hepatitis C treatment. 

5. Dusetzina SB. Share of specialty drugs in commercial plans nearly quadrupled, 2003-14. 

2016;35(7):1241-1246. doi: 10.1377/hlthaff.2015.1657.

6. Sarpatwari A, Avorn J, Kesselheim AS. State initiatives to control medical costs—can transparency legislation help? 

2016;374(24):2301-2304. doi: 10.1056/NEJMp1605100.

7. Bach PB, Pearson, SD. Payer and policy maker steps to support value-based pricing for drugs. 

. 2015;314(23):2503-2504. doi: 10.1001/jama.2015.16843.

8. Neumann PJ, Cohen JT. Measuring the value of prescription drugs. 

 2015;373(27):2595-2597. doi: 10.1056/NEJMp1512009.

9. Basu A. Financing cures in the United States. 

. 2015;15(1):1-4. doi: 10.1586/14737167.2015.990887.

10. Jacobs DB, Sommers BD. Using drugs to discriminate—adverse selection in the insurance marketplace. 

 2015;372(5):399-402. doi: 10.1056/NEJMp1411376.

11. Explaining health care reform: risk adjustment, reinsurance, and risk corridors. Kaiser Family Foundation website. http://kff.org/health-reform/issue-brief/explaining-health-care-reform-risk-adjustment-reinsurance-and-risk-corridors/. Published August 17, 2016. Accessed July 5, 2016.

12. Jost TS. Stabilizing forces: the difference premium stabilization programs make in the Affordable Care Act marketplaces and Medicare Part D. 

website. http://theactuarymagazine.org/stabilizing-forces/. Accessed September 30, 2016.&ensp;

COMMENTARY

For cost-effective cures, the drug pricing policy challenge is not simply to lower prices, but also to spread the drug costs among payers.

The Challenge of Paying for Cost-Effective Cures

Breast cancer screening guidelines and metrics are inconsistent with each other and may differ from breast screening practice patterns in primary care. This study measured breast cancer screening practice patterns in relation to common evidence-based guidelines and accountability metrics.

 Cohort study using primary data collected from a regional breast cancer screening research network between 2011 and 2014.

 Using information on women aged 30 to 89 years within 21 primary care practices of 2 large integrated health systems in New England, we measured the proportion of women screened overall and by age using 2 screening definition categories: any mammogram and screening mammogram.

 Of the 81,352 women in our cohort, 54,903 (67.5%) had at least 1 mammogram during the time period, 48,314 (59.4%) had a screening mammogram. Women aged 50 to 69 years were the highest proportion screened (82.4% any mammogram, 75% screening indication); 72.6% of women at age 40 had a screening mammogram with a median of 70% (range = 54.3%-84.8%) among the practices. Of women aged at least 75 years, 63.3% had a screening mammogram, with the median of 63.9% (range = 37.2%-78.3%) among the practices. Of women who had 2 or more mammograms, 79.5% were screened annually.

Primary care practice patterns for breast cancer screening are not well aligned with some evidence-based guidelines and accountability metrics. Metrics and incentives should be designed with more uniformity and should also include shared decision making when the evidence does not clearly support one single conclusion.

This study explores the inherent challenges of shifting primary care breast cancer screening practice in response to evidence-based guidelines while also attending to accountability and performance measures. 

		Studies examining practice-level patterns for breast cancer screening are limited. 

		Breast cancer screening guidelines and accountability metrics are inconsistent with each other. 

		Primary care practice patterns are not shifting with screening guidelines.

		The heterogeneity of accountability measurement may be a barrier for the uptake of new guidelines in response to recent evidence. 

		Given that practices seek to provide guideline-based care and achieve quality and accountability metrics, alignment of these measures is important for patient care.

Breast cancer screening practices are often debated in clinical practice, public health, and national dialogues. A multitude of medical professional organizations endorse specific sets of breast cancer screening guidelines, such as that of the United States Preventive Services Task Force (USPSTF),

and the American College of Radiology (ACR),

 among others. These guidelines vary to some extent, and represent different interpretations of largely the same evidence base. At the same time, in an effort to improve quality of care delivery and hold organizations accountable for the dollars they spend, a number of quality measures are in use by healthcare organizations and practices, such as Healthcare Effectiveness Data and Information Set (HEDIS) measures from the National Quality Forum (NQF)

These measures are often tied to fiscal and other organizational incentives through contractual payment mechanisms.

Currently, tested models of care delivery, such as accountable care organizations (ACOs)

 and patient-centered medical homes, are employing measures for “best practices” and for qualification/recognition, as well to establish whether bonuses or savings will be paid out. Specifically, in many of these risk-based contracts, if the provider entity shows worse quality on a number of established quality measures—often including cancer screening&mdash;provider organizations may not be able to participate in any shared savings generated by new payment models.

 In other pay-for-performance models, set thresholds for improvement in certain quality measures result in higher payment bonuses.

 Further, practices often have contracts with private insurers that specify guidelines of care for breast cancer screening, such as annual screening in women aged 40 to 49 years. Breast cancer screening is one of the measures that healthcare systems and practices are typically required to report. Given that physicians and practices seek to provide guideline-based care and achieve quality/accountability metrics, understanding the alignment of these measures is important for patient care.

 presents a summary of common evidence-based guidelines and quality measures for breast cancer screening.

 Major evidence-based guidelines that are endorsed or disseminated by professional organizations include the USPSTF, which, since 2009, has recommended that biennial screening typically begin at age 50 through age 74 for women of average breast cancer risk—although for women aged 40 to 49, a risk- and preference-based decision should be made.

 The most recently released USPSTF guidelines (2016) continue to support the previous 2009 screening guidelines, but include a statement that there is little evidence at this time on the effectiveness of digital tomosynthesis (3D mammography) or additional screening for women with dense breasts.

 In contrast, the ACS and ACR recommend that average-risk women begin screening at age 40, and continue annually, without a pre-specified ending age.

Measures for quality, payment, or other forms of accountability are usually derived from the same evidence as the professional organizations, but are not always aligned with guidelines. For example, prior to 2014, the HEDIS and ACO breast cancer screening metric was based on women initiating screening at age 40 and continuing until age 69 every 2 years, and also included any mammogram, not necessarily with a screening indication.

 These practice quality measures were instituted at many practices, and may have provided an impetus or financial incentive to achieve the best measures possible by adhering to their screening parameters. At the same time, women receive recommendations on breast cancer screening not only from primary care, but also from radiology and obstetrics and gynecology practices, which typically recommend an annual interval.

From 2009 to 2013, providers wanting to adopt USPSTF breast cancer screening guidelines would not be able to fully do so and perform well using HEDIS, ACO, and NCQA measures. This is due to the number of women aged 40 to 49 years who may choose not to be screened per USPSTF guidelines/recommendations, and thus, they would appear “not current,” which could reduce the rates on which practice performance was measured. On the other hand, providers using ACS or ACR breast cancer screening guidelines would be simultaneously concordant with the accountability metrics of HEDIS and others. In 2014, HEDIS and ACO measures were changed for breast cancer screening such that they were concordant with that of USPSTF for a starting age of 50 at 2-year screening intervals until age 74.

 This environment of heterogeneous guidelines, differential uptake of guidelines over time, and concurrent goals of providing patient-centered care, as well as meeting practice-based metrics, can potentially create unavoidable discordance of breast cancer screening practices within the range of recommendations.

The objective of this study was to examine breast cancer screening practice patterns to assess concordance with evidence-based guidelines and accountability metrics for primary care within a sample of practices from 2 large regional healthcare systems.

This study was conducted within one of the consortium member networks of the National Cancer Institute (NCI)—funded Population-based Research Optimizing Screening for Personalized Regimens (PROSPR),

 which is focused on breast cancer screening. The PROSPR Research Center (PRC) includes data on breast cancer screening within the primary care populations of the Dartmouth-Hitchcock regional network in New Hampshire and the Brigham and Women’s Hospital system in greater Boston and surrounding areas in Massachusetts. Our PRC comprises 37 primary care facilities and 10 radiology facilities in the bi-state region, and includes data from January 2011 through September 2014 on a primary care population (the PRC cohort) of women between the ages of 30 and 89 years, who had at least 1 primary care visit in the past 24 months within our respective healthcare systems.

To assess a biennial screening interval, the analysis was restricted to those women who became part of our PRC cohort anytime between January 1, 2011, and June 30, 2012, which allowed at least 24 months (plus an additional 3 months to account for the scheduling and completion of a mammography exam) from the time a woman became a part of the PRC cohort until September 2014 (n = 83,725). We excluded from the study, 16 primary care facilities with fewer than 100 women visiting during the study period. This resulted in a final cohort of 81,352 women among the 21 primary care facilities. The study was approved prior to data collection by the Institutional Review Boards of Dartmouth College and Brigham and Women’s Hospital.

Actual screening patterns were measured using data from our PRC database. Data are routinely collected for our PRC cohort on breast imaging (including breast screening exams within 2 years prior to becoming part of the PRC cohort), follow-up, breast pathology, breast cancer diagnosis, and vital status. Specifically, for this study, we used the following data elements: entry into the PRC cohort, the primary care facility, age at PRC cohort entry, date and exam indication for 2D and 3D breast images (mammograms), age at mammography, and vital status. Data sources used included the electronic health record (EHR), radiology information system databases, and institutional cancer registries. All data from our PRC were systematically mapped into a single database with common data elements.

Using the PRC database, we categorized 2 measures for determining the receipt of breast cancer screening based on a woman’s breast images that occurred during the study period. The first screening measure category included receipt of any mammogram (screening or diagnostic), which corresponds to HEDIS and ACO metrics. The second category was receipt of a mammogram that specifically indicated that the purpose of the exam was screening, which corresponds to USPSTF guidelines. Additionally, the woman’s receipt of a screening mammogram within 2 years of her PRC cohort entry date was included in the screening measure definitions. The age at PRC cohort entry was categorized as follows (in years): under 40, 40 to 44, 45 to 49, 50 to 69, 70 to 74, 75 to 79, 80 to 84, and 85 or older. We reported the overall population age distribution by age groups and provided the number and percent of each population age group’s screening measures.

For analyzing the proportion of women who initiated and continued screening, including the time interval between screens (

), we used the USPSTF criteria category (receipt of a mammogram indicated as screening). We assessed the percent of women initiating screening at the age of 40 and at the age of 50, and continued screening at age 75 or older for each primary care facility. We defined the age-40 cutoff with a 27-month window to account for time to schedule and complete an exam. For example, if a woman received a screening mammogram up to 27 months following her 40th birthday, she would be considered as initiating screening at age 40. We summarized the frequency and proportion of women within 5-year age categories for each of the screening intervals: annual, biennial, and over 2 years, in addition to the median and interquartile range (IQR) of screening interval in days. The screening intervals were defined as 9 to 18 months for annual, 18 to 27 months for biennial, and over 27 months for more than 2 years.

Our study cohort included 81,352 total women, with 38,897 (47.8%) aged 50 to 74 years (

). Overall, 54,903 of 81,352 (67.5%) had a mammography exam of any type (screening or diagnostic), and 48,314 of 81,352 (59.4%) had a screening mammogram during our study period, with the majority of mammography exams occurring within the 50-to-69 age group (82.4% and 75%, respectively) (Table 3). Seventy percent of women aged 40 to 44 had a screening mammogram and 77.3% had a mammogram of any type (screening or diagnostic). Among the older age groups, the proportion with a mammogram steadily decreased; however, over a quarter (27.5%) of women 85 years or older had received a screening mammogram (Table 3).

Examining the proportion of women by age categories, those with a screening mammogram at age 40 was 72.6% overall, and across the primary care facilities, the median was 70% with a range of 54.3% to 84.8% (Figure). For women who initiated with a screening mammogram at age 50, the overall proportion with mammography was 75.5%, with a median of 77.2% and a range of 51.4% to 91.8% across the primary care facilities. Sixty-three percent of women 75 years or older continued screening, and among the primary care facilities, the median was 63.9% with a range of 37.2% to 78.3% (Figure).

Of the women undergoing screening mammography (n = 32,275), the vast majority had a mammogram on an annual basis (79.5%) (Table 2). As the age category increased, the proportion of screening women in the annual screening group also increased (Table 2). Median screening intervals overall and by age group reflect the high proportion of annual screeners (Table 2). The median (IQR) screening interval among those who screened annually was 379 days (IQR, 369-419; data not shown).

This study explores the inherent challenges of shifting breast cancer screening practice in response to evidence-based guidelines, while also attending to accountability and performance measures. We found that screening mammography—in other words, the purpose of the exam was for screening&mdash;appears to be largely initiated in the 40s age group, and stays fairly consistent until later life. For older women (ie, 75 years or older), there is a notable drop-off in screening, but well over half the women in this age group were screened, and well over a third of women aged 80 to 84 were screened. Among women who have been screened, the vast majority do so at an annual interval across all ages, with median screening interval bearing that out.

Using the definition that receipt of any mammogram (screening or diagnostic) can be considered having been screened—such as in HEDIS and ACO measures&mdash;for a 2-year period, the proportion of women screened appears much higher. This work suggests that practice patterns are not shifting with the USPSTF guidelines, and that the heterogeneity of accountability and performance measurement may be a barrier to change given the lack of uniformity of payment incentives, performance measures, and evidence-based guidelines.

By examining which breast cancer screening guidelines primary care practices follow within the complex landscape of evidence-based guidelines, and the increasing need for accountability metrics, we found that ACS and ACR guidelines were followed, as was the age of initiation for the 2012 HEDIS measure. Notably, the USPSTF and HEDIS (2014 edition) starting age and interval were not well followed in practice, but would be met by default because the measure of biennial mammography would be satisfied by women screening annually; further, screening at age 50 would be met by women who had started in their 40s, and there is no penalty for overscreening. The USPSTF and HEDIS stopping ages also were not reflected in the screening practice patterns, because a large proportion of women continued on with screening after age 75, which is in alignment with ACR and ACS guidelines.

Few studies have examined concordance of practice-level breast cancer screening guidelines and patterns in relation to prevailing evidence-based recommendations and accountability metrics. Several studies, however, have examined patient

adoption of breast cancer screening guidelines—particularly those related to the 2009 USPSTF recommendations, which were changed from the grade B 2002 recommendation of screening mammography every 1 to 2 years, starting at age 40. In a 2009 western Washington study of 18 providers, one study reported 3 main reasons for low intention to change breast cancer screening practices in response to the USPSTF recommendations: lack of confidence in the evidence, limited availability of low-cost mammography, and desire to offer more services.

 Although the representativeness of this study is unknown for US primary care providers, it highlights the possibility that factors driving clinical decision making by providers may or may not be congruent with prevailing national recommendations and delivery models.

In addition to recommendations from medical organizations, new care delivery models, such as the patient-centered medical home, seek to employ standardized measures to track cost, use, and clinical quality in order to promote benefits to patient care,

 For the PCMH model, breast cancer screening follows the HEDIS and NCQA 2012 measure set, which are fundamentally derived from the same evidence base as other breast cancer screening recommendations. Notably, care delivery models and incentives are typically based in primary care and not directly in radiology practices, which play a key role in breast cancer screening practices and often use ACR and/or ACS guidelines. Differing guidelines for breast cancer screening allow for variation in adoption based on clinical specialty, practice and provider preferences, institutional financial incentives, and regional norms. For breast cancer screening, lack of uniformity to the guidelines and measures almost ensures a heterogeneous approach to breast cancer screening in clinical practice. Heterogeneity in guidelines may create conundrums for practices and their patients and providers, especially if providers within practices and/or patients differ in which they choose to follow.

This issue is not limited to breast cancer screening, but is seen in other areas of clinical practice, with hypertension being a prime example. There is currently controversy over the appropriate cut point for the target systolic blood pressure (SBP) in individuals over 60 years of age.

 Specifically in question, is whether the target SBP should be 150 or 140 mm Hg for this age group,

 with some guidelines set at 150 mm Hg; however, the HEDIS and CMS measures still have the goal of 140 mm Hg.

 Given the confusion and the lagging change in high-stakes measures for payment and shared results, it may be no surprise that providers are reticent to change practice, as has been demonstrated.

Organizations and practices increasingly have the ability to track actual practice patterns for breast cancer screening through both claims and EHR-based measures, and are, in fact, incentivized to do so for accountability measures through use of EHRs and claims data. Assessing, at the practice and provider levels, how well breast cancer screening patterns match organizational and/or provider-specific goals based on the guidelines and measures of choice, is an important component of streamlining how breast cancer screening is delivered. However, incorporating women’s screening preferences in quality of care metrics is essential for ensuring high-quality patient-centered care.

Several population-based studies, based on national survey data, have examined breast cancer screening in relation to the USPSTF recommendations. In a Medicare sample from 2005 to 2010, an abrupt decline in screening mammography use (—4.3%) was noted in 2010 relative to 2009.

 In contrast, among 5.5 million privately insured women aged 40 to 64 from 2006 to 2011, a small decrease was seen in screening among women aged 40 to 49 years 2 months after the USPSTF guidelines; but by 2 years following the recommendations, a small increase was seen in screening mammography rates across all ages.

 No changes in screening patterns were observed after the 2009 USPSTF guidelines in a 2006 to 2010 Medical Expenditure Panel Survey of women 40 years or older.

 Our studies are in accord with those reporting no effect of the USPSTF guidelines on practice patterns, and extend the generalizability of that finding to women within primary care and for the added years of observation beyond 2010.

Recent studies of the 2009 USPSTF recommendation changes also show lack of awareness of both the guidelines and risks and benefits of screening by women, leading to resistance in adopting new guidelines.

 Women’s preferences and expectations may be shaped by media, providers, and social networks, and may not match evolving adoption of new breast cancer screening guidelines by their providers and practices. If practices hope to deliver the best patient-centered care, then cultivating the closest alignment of guidelines, accountability measures, and preferences in breast cancer screening should be a key objective.

A strength of this study was the focus on practice-level patterns, of which there is little reported in the literature. However, a limitation is that we did not capture patient preferences, and thus, we cannot account for their potential role in the patterns observed. A multi-level study is needed to simultaneously examine patient-, provider-, and practice-level breast cancer screening patterns and preferences—ideally, in a variety of healthcare settings and locales, which may capture important warranted variation in processes of care.

There is a complex landscape of breast cancer screening with its host of recommendations, guidelines, and measures. Adding to that complexity is the multilevel nature of healthcare systems—with patients, providers, practices, and networks&mdash;all with their own objectives, preferences, and accountabilities. Increasing the complexity further, is the diversity of clinical settings and professional lenses through which breast cancer screening may be viewed. For example, general internal medicine and family practice may differ from radiology in the application of evidence, choice of guidelines, or need for specific accountability measures. The situation is similar with the hypertension recommendations, in which general internal medicine, family practice, and cardiology may not all endorse or use the same SBP measure. With heterogeneity in clinical guidelines across specialties&mdash;despite their grounding in the same evidence base and differing priorities within the multi-level healthcare environment (patient-centered care vs fiscally motivated metrics)&mdash;it is not surprising that there is room for improvement in the provision of patient-centered care related to breast cancer screening. Additional efforts are needed to reconcile conflicting guidelines and to harmonize accountability and performance measures with evidence-based guidelines.

 Department of Biomedical Data Science (TO, JW, MG, SP), and Norris Cotton Cancer Center (TO, ANAT), and The Dartmouth Institute for Health Policy and Clinical Practice (TO, CB, ANAT), Geisel School of Medicine at Dartmouth, Lebanon, NH; Division of General Internal Medicine, Brigham and Women’s Hospital (JSH, AB, KH), Boston, MA; Department of Medicine, Dartmouth Hitchcock Medical Center (CB), Lebanon, NH.

The design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript was supported as part of the National Cancer Institute—funded consortium, Population-based Research Optimizing Screening through Personalized Regimens (PROSPR) (U54 CA163307). 

Drs Tosteson and Haas received a grant from the National Cancer Institute. Dr Bitton serves as a part-time advisor to the Center for Medicare and Medicaid Innovation (CMMI) on their comprehensive primary care initiative (no involvement on setting quality metrics for CMS or CMMI). The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article. 

Concept and design (AB, MG, JSH, KH, TO, SP, ANAT); acquisition of data (MG, JSH, KH, TO, SP); analysis and interpretation of data (AB, CB, MG, JSH, TO, SP, ANAT, JW); drafting of the manuscript (AB, CB, MG, JSH, TO, JW); critical revision of the manuscript for important intellectual content (AB, CB, MG, JSH, TO, ANAT); statistical analysis (JW); provision of patients or study materials (MG, JSH); obtaining funding (JSH, TO, ANAT); administrative, technical, or logistic support (MG, JSH, KH, TO); and supervision (TO). 

 Martha Goodrich, MS, Geisel School of Medicine at Dartmouth, 1 Medical Center Dr, Lebanon, NH 03756. E-mail: Martha.e.goodrich@dartmouth.edu. 

This study measured breast cancer screening practice patterns in relation to evidence-based guidelines and accountability metrics, and found closer alignment is needed for providing patient-centered care.

January 2017

Value-Based Payment in Implementing Evidence-Based Care: The Mental Health Integration Program in Washington State

The Challenge of Paying for Cost-Effective Cures

Alignment of Breast Cancer Screening Guidelines, Accountability Metrics, and Practice Patterns

An Assessment of the CHIP/Medicaid Quality Measure for ADHD

An Expanded Portfolio of Survival Metrics for Assessing Anticancer Agents

Electronic Health Records and the Frequency of Diagnostic Test Orders

The Value of Survival Gains in Myelodysplastic Syndromes

The Effect of Massachusetts Health Reform on Access to Care for Medicaid Beneficiaries

Patient-Centered Care: Turning the Rhetoric Into Reality

The Social Value of Childhood Vaccination in the United States