Zirui Song, MD, PhD

The COVID-19 pandemic led to a significant disruption, then recovery, of health care services use. Prior research has not examined the relative rates of resumption of high-value and low-value care. We examined the use of 6 common low-value services that received a D grade from the US Preventive Services Task Force compared with clinically comparable high-value services in a large commercially insured population nationwide from before the pandemic to April 1, 2021. We found that, overall, low-value services and high-value services were disrupted similarly. In aggregate, low-value care declined to 56.2% and high-value care to 53.2% in the initial month of the pandemic (April 2020) relative to baseline (number of visits in 2019 normalized by relevant enrolled population), then rebounded to 83.1% of baseline for low-value services and 95.0% of baseline for high-value services by January 2021. Substantial heterogeneity appeared across clinical contexts, such as prostate cancer screening for men 70 years and older rebounding to 111.8% of baseline and asymptomatic chronic obstructive pulmonary disease screening remaining at 38.5% of baseline in January 2021. This suggests that although, on average, resuming lower-value services may have been perceived to be a lesser priority by providers and patients, the pandemic may have had heterogeneous effects on consumer and provider decision-making along the dimension of clinical value. This enhances our understanding of how disruptions affect the relationship between clinical value and usage of different services and suggests the need for more targeted interventions to reduce low-value care.

We found that, overall, low-value services and high-value services were disrupted similarly during COVID-19, but low-value services rebounded, on average, more slowly than their high-value counterparts.

In aggregate, low-value care declined to 56.2% and high-value care to 53.2% in the initial month of the pandemic (April 2020) relative to their 2019 baseline.

In aggregate, low-value care rebounded to 83.1% of baseline, whereas high-value care rebounded to 95.0% of baseline by January 2021.

Substantial heterogeneity exists across services, suggesting that the pandemic may have had heterogeneous effects on consumer and provider decision-making along the dimension of clinical value.

The COVID-19 pandemic disrupted many aspects of health care delivery. In both outpatient and emergency settings, large declines in utilization were documented in the early months of the pandemic,

 with the volume of services rebounding toward prepandemic levels in the third quarter of 2020. To date, much of the focus has been on delayed or forgone high-value services, such as cancer screenings,

 care for stroke and myocardial infarction,

 which could potentially result in long-term negative health consequences.

However, the ways in which the COVID-19 pandemic may have affected the use of low-value services, relative to high-value services, remain poorly understood. Low-value care in the United States accounted for approximately 10% to 20% of total health care spending before the pandemic and has proven difficult to reduce.

 Some have hypothesized that the pandemic-related care disruptions and recovery could reduce low-value spending and substitute such spending with, on average, higher-value spending,

 but there has been no empirical work evaluating those predictions. To address this gap in knowledge, we compared changes in the use of sentinel low-value services with changes in use of comparable high-value services over the course of the pandemic. Our results shed light on how clinical value is inculcated into physician and consumer decision-making and on the role of 1-time disruptions in changing or failing to change low-value care usage.

We analyzed a subset of services with a D grade from the US Preventive Services Task Force (USPSTF). A grade of D is assigned to a service when “there is moderate or high certainty that the service has no net benefit or that the harms outweigh the benefits”

 and the USPSTF discourages use of the service. Grade D services, therefore, are among the most rigorously developed low-value services to target for reduction. Of the 19 services with a D rating from the USPSTF, 6 were commonly used by the commercially insured population and reliably identifiable in the claims data (

]). The remaining D-graded services were excluded for the following reasons: (1) lack of coding specificity (eg, pancreatic cancer screening), (2) the patient population addressed by the D grade was difficult to identify from claims (eg, women who are not at increased risk of breast cancer), or (3) the service was rarely identified in the data set (eg, cervical cancer screening for female patients younger than 21 years or testicular cancer screening for male patients).

To understand how changes in volume attributable to the pandemic may have differed by clinical value, utilization of each of the 6 low-value services was compared with use of services of relatively higher clinical value that required similar resources and patient time to complete (

). For example, in 2 service pairs, blood tests that do not require a separate clinic visit were compared (eg, prostate specific-antigen [PSA] screening after age 70 years [low value] was compared with hepatitis C virus screening after age 70 years [high value]). One pair compared services that are routinely performed during a clinic visit (eg, asymptomatic bacteriuria testing [low value] was compared with serum cholesterol testing and glycated hemoglobin A

] testing [high value]). Another pair compared prescription medications (eg, hormone replacement therapy [low value] and statins [high value]). An additional pair compared services that necessitated a specific clinic visit (eg, chronic obstructive pulmonary disease [COPD] screening for asymptomatic individuals [low value] was compared with lung cancer screening [high value]). Diagnosis codes for low-value and high-value services can be found in the eAppendix.

We used deidentified claims data spanning January 2019 through March 2021 from the OptumLabs Data Warehouse, which includes medical and pharmacy claims for commercial and Medicare Advantage enrollees representing a mixture of geographical regions across the United States. We measured monthly utilization in terms of unique patient encounters for each service from January 1, 2019, to April 1, 2021. In our main specification, we normalized the number of observed encounters by the relevant enrolled population (eg, for prostate cancer screening for patients 70 years and older, we divide the number of encounters by number of men 70 years and older who were enrolled during the month) to ensure that our results were not driven by changes in the age or gender composition of enrollees. Changes in utilization were measured relative to baseline utilization during the first quarter of 2019 to allow for comparisons between services to be standardized to a prepandemic month and not driven by endogenous differences in levels during the pandemic.

We also compared volumes of our low-value services with overall trends in the volume of medical activity in the delivery system, represented by unique patient visits for all services. We excluded telemedicine from this count of visits because our set of low-value services, with the exception of prescription drugs, require an in-person visit to a health facility, and thus all in-person visits were a more suitable comparison group. The Harvard Medical School institutional review board deemed this study exempt from review, owing to the use of deidentified data.

Weighting all low-value services equally, our results show that enrollment-adjusted volume of low-value services decreased to 56.2% of 2019 baseline volume in April 2020. By January 2021, the volume of low-value services had rebounded to 83.1% of its 2019 baseline. This is in comparison with all enrollment-adjusted nontelemedicine visits that fell to 63.1% of 2019 volumes in April 2020 and rebounded to 104.1% in January 2021, or equally weighted high-value services that fell to 53.2% of 2019 volume in April 2020 and rebounded to 95.0% in January 2021.

We decomposed the aggregate results into 4 categories: (1) services that were mostly unaffected by COVID-19, (2) services that experienced a decline followed by a robust rebound comparable with high-value services, (3) services that experienced a decline followed by a slower rebound than high-value services, and (4) services that declined but have not experienced a rebound. The first category included services for which we observed limited or no decline in utilization attributable to COVID-19. This set of services included bacterial vaginosis screening and hormone replacement therapy, which both remained above 84% of monthly 2019 first-quarter volumes in April 2020 and in January 2021. The second included services whose volume declined because of COVID-19 but showed a robust rebound similar to their high-value counterparts. This category included prostate cancer screenings, which rebounded to 111.8% of 2019 volume by January 2021, a stronger rebound than that of comparable higher-value services of HbA

 testing, hepatitis B and C screening, and cholesterol testing for individuals 70 years and older, which had rebound volumes of 102.2%, 99.5%, and 73.6% of 2019 volume, respectively. The third category contained services for which a pandemic-associated decline was observed and the rebound was slower than that of comparable higher-value services. Asymptomatic bacteriuria testing and herpes screening are in this category. These low-value services have rebounded in volume over time, although the rates of rebound for asymptomatic bacteriuria testing remained below those of comparable services, such as cholesterol testing, HIV testing, and HbA

 testing, and herpes screening lagged behind testing for hepatitis B and C. Finally, the fourth category contained services whose utilization declined during the pandemic and has not rebounded substantially. COPD screening declined by nearly 80% compared with the baseline volume in January 2020, which is twice the magnitude of decline for lung cancer screening or all nontelemedicine visits in our data. The decline was sustained with levels 60% below baseline 2019 levels, even in the first quarter of 2021.

Although our results may not be causal, they document a sharp decline in most low-value services at the onset of the COVID-19 pandemic, with magnitudes generally similar to changes in utilization of comparison higher-value services. The rebound of these low-value services, however, was not uniform. Of the 4 low-value services that experienced a sharp decline, most exhibited a slower rebound than comparable higher-value services and slower rebound relative to all health care visits. As the 

 shows, when weighting each of the low-value and high-value services equally to create aggregate indices, the index of lower-value services rebounded slower than the index of high-value services. This suggests that resuming lower-value services may have been perceived to be a lesser priority by providers and patients as health care encounters rebounded during the later months of the pandemic. At the same time, robust rebound in some low-value health care services suggests that other factors may also play an important role in this type of decision-making. Our results show that, although low-value care did, on average, rebound to a lower degree, there is significant heterogeneity across services. This would suggest that health systems would need interventions that are more targeted than a 1-time disruption to reduce usage of low-value services. Furthermore, because the heterogeneity in services may be driven by idiosyncratic factors unrelated to perceived value, our results should be coupled with robust qualitative analyses to further elucidate how providers and patients made decisions about service provision and usage, respectively, and consequently what policies need to be designed to better align this behavior with value.

Our study has several limitations. First, data were limited to 12 months from the start of the pandemic, and some of the patterns observed may not persist in the future. However, most claims-based studies of health care use during the pandemic have yet to contain 12 months of postpandemic data. Second, as this is a “Trends from the Field” paper, we did not attempt to isolate specific clinical mechanisms or explanations for the observed changes in utilization. Although the initial drop-off in utilization was likely because of patients forgoing nonurgent care and outpatient providers reducing access,

 why some low-value services rebounded more than others is likely explained by clinical factors specific to the service. For example, the drop-off in COPD screening may have persisted because COPD screening requires patients to perform lung maneuvers without a mask that are particularly high risk for providers during COVID-19. In contrast, blood tests such as the PSA screening can be done with the patient masked. Thus, this paper does not establish a causal link between value and patterns of care during the pandemic. Other factors that may be associated with changing utilization patterns for these services may be unrelated to clinical value. In fact, the heterogeneity of results with respect to value suggests that this is the case and, as a result, stakeholders interested in improving the value of care must better understand the diverse drivers of utilization. Third, the analyses focused on commercially insured populations who are younger and likely of higher socioeconomic status than the general US population and may differ from other populations, such as those covered under Medicaid or traditional Medicare.

To our knowledge, this is the first large-scale claims-based analysis comparing changes in the volume of high-value and low-value services during the COVID-19 pandemic and its recovery. The finding that multiple low-value services rebounded slower than comparable high-value services suggests that the pandemic may have had heterogeneous effects on consumer and provider decision-making along the dimension of clinical value. Whether such clinically nuanced effects are largely transient or whether they persist to help nudge the delivery system toward higher value remains to be seen. Given that some low-value services had a robust rebound, there is a need for interventions that more clearly target low-value care as opposed to blunt disruptions such as the pandemic to reduce these types of low-value care. 


 Harvard University (MS), Cambridge, MA; OptumLabs (MS), Eden Prairie, MN; Harvard Medical School (ZS), Boston, MA; Massachusetts General Hospital (ZS), Boston, MA; Center for Primary Care, Harvard Medical School (ZS), Boston, MA; Department of Health Care Policy, Harvard Medical School (MEC), Boston, MA; Department of Internal Medicine and Center for Value-Based Insurance Design, University of Michigan (AMF), Ann Arbor, MI.

 Drs Song and Chernew gratefully acknowledge funding from Arnold Ventures. Dr Song gratefully acknowledges support from the Office of the Director, National Institutes of Health (NIH) (DP5-OD024564).

 Dr Song is a practicing physician at Massachusetts General Hospital, including caring for patients with COVID-19; has submitted a grant to the NIH related to the value of care during the COVID-19 pandemic; and has received personal fees from the Research Triangle Institute for work on Medicare risk adjustment, from Google Ventures for academic lectures on health policy outside of this work, from V-BID Health for presentation of ongoing research, and for providing consultation in legal cases. Dr Chernew is the Chair of the Medicare Payment Advisory Commission (MedPAC); received personal fees from MITRE and sits on an advisory board for National Institute of Health Care Management; equity in Archway Health, Waymark Inc, V-BID Health, and Virta Health; grants from Ballad Health, Arnold Ventures, Commonwealth Fund, Signify Health, BCBSA, and HCSC; and is co–editor in chief of 

®). Dr Fendrick reports consulting fees from AbbVie, Bayer, Centivo, Covered California, Emblem Health, Exact Sciences, GRAIL, Harvard University, Health & Wellness Innovations, Health at Scale Technologies, HealthCorum, Hygieia, MedZed, Merck, Mother Goose Health, Phathom Pharmaceuticals, Sempre Health, Silverfern Health, State of Minnesota, Teledoc Health, US Department of Defense, Virginia Center for Health Innovation, Wellth, Wildflower Health, Yale–New Haven Health System, and Zansors; research support from the Agency for Healthcare Research and Quality, Arnold Ventures, Boehringer Ingelheim, Gary and Mary West Health Policy Center, National Pharmaceutical Council, Patient-Centered Outcomes Research Institute, PhRMA, Robert Wood Johnson Foundation, and State of Michigan/CMS; and serving as co-editor-in-chief of 

, member of the Medicare Evidence Development & Coverage Advisory Committee, and partner of V-BID Health, LLC. Ms Shahzad reports no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article. All opinions expressed are those of the authors and not any organization which they are affiliated with.

 Concept and design (MS, ZS, MEC, AMF); acquisition of data (MS, MEC); analysis and interpretation of data (MS, ZS, MEC, AMF); drafting of the manuscript (MS); critical revision of the manuscript for important intellectual content (MS, ZS, MEC, AMF); statistical analysis (MS, ZS); obtaining funding (ZS); and supervision (ZS, MEC, AMF).

 Mahnum Shahzad, BA, Harvard University, 14 Story St, Cambridge, MA 02138. Email: mahnum_shahzad@g.harvard.edu.

1. Mehrotra A, Chernew ME, Linetsky D, Hatch H, Cutler DA. The impact of the COVID-19 pandemic on outpatient visits: practices are adapting to the new normal. The Commonwealth Fund. June 25, 2020. Accessed September 2, 2020. https://www.commonwealthfund.org/publications/2020/jun/impact-covid-19-pandemic-outpatient-visits-practices-adapting-new-normal

2. Mehrotra A, Chernew ME, Linetsky D, Hatch H, Cutler DA. The impact of the COVID-19 pandemic on outpatient visits: changing patterns of care in the newest COVID-19 hot spots. The Commonwealth Fund. August 13, 2020. Accessed September 2, 2020. https://www.commonwealthfund.org/publications/2020/aug/impact-covid-19-pandemic-outpatient-visits-changing-patterns-care-newest

3. Chatterji P, Li Y. Effects of the COVID-19 pandemic on outpatient providers in the US. National Bureau of Economic Research working paper No. 27173. May 2020. Accessed September 2, 2020. https://www.nber.org/papers/w27173.pdf

4. Ziedan E, Simon KI, Wing C. Effects of state COVID-19 closure policy on non-COVID-19 health care utilization. National Bureau of Economic Research working paper No. 27621. July 2020. Accessed January 18, 2021. https://www.nber.org/papers/w27621

5. Abelson R. Doctors without patients: ‘our waiting rooms are like ghost towns.’ 

. May 5, 2020. Accessed September 9, 2020. https://www.nytimes.com/2020/05/05/health/coronavirus-primary-care-doctor.html

6. Jeffery MM, D’Onofrio G, Paek H, et al. Trends in emergency department visits and hospital admissions in health care systems in 5 states in the first months of the COVID-19 pandemic in the US. 

. 2020;180(10):1328-1333. doi:10.1001/jamainternmed.2020.3288

7. Rosenbaum L. The untold toll — the pandemic’s effects on patients without COVID-19. 

. 2020;382(24):2368-2371. doi:10.1056/NEJMms2009984

8. Hartnett KP, Kite-Powell A, DeVies J, et al; National Syndromic Surveillance Program Community of Practice. Impact of the COVID-19 pandemic on emergency department visits — United States, January 1, 2019–May 30, 2020. 

. 2020;69(23):699-704. doi:10.15585/mmwr.mm6923e1

9. McWilliams JM, Russo A, Mehrotra A. Implications of early health care spending reductions for expected spending as the COVID-19 pandemic evolves. 

. 2021;181(1):118-120. doi:10.1001/jamainternmed.2020.5333

10. Chen RC, Haynes K, Du S, Barron J, Katz AJ. Association of cancer screening deficit in the United States with the COVID-19 pandemic. 

. 2021;7(6):878-884. doi:10.1001/jamaoncol.2021.0884

11. Birkmeyer JD, Barnato A, Birkmeyer N, Bessler R, Skinner J. The impact of the COVID-19 pandemic on hospital admissions in the United States. 

. 2020;39(11):2010-2017. doi:10.1377/hlthaff.2020.00980

12. Clement J, Jacobi M, Greenwood BN. Patient access to chronic medications during the COVID-19 pandemic: evidence from a comprehensive dataset of US insurance claims. 

. 2021;16(4):e0249453. doi:10.1371/journal.pone.0249453

13. Patel SY, McCoy RG, Barnett ML, Shah ND, Mehrotra A. Diabetes care and glycemic control during the COVID-19 pandemic in the United States. 

. 2021;181(10):1412-1414. doi:10.1001/jamainternmed.2021.3047

14. Shrank WH, Rogstad TL, Parekh N. Waste in the US health care system: estimated costs and potential for savings. 

. 2019;322(15):1501-1509. doi:10.1001/jama.2019.13978

15. Oakes AH, Segal JB. The COVID-19 pandemic can help us understand low-value health care. 

. October 27, 2020. Accessed July 23, 2021. https://www.healthaffairs.org/do/10.1377/hblog20201023.522078/full/

16. Sorenson C, Japinga M, Crook H, McClellan M. Building a better health care system post-COVID-19: steps for reducing low-value and wasteful care. 

. August 21, 2020. Accessed June 7, 2021. https://catalyst.nejm.org/doi/full/10.1056/CAT.20.0368

17. Bhatia RS, Shojania KG, Levinson W. Cost of contact: redesigning healthcare in the age of COVID. 

. 2021;30(3):236-239. doi:10.1136/bmjqs-2020-011624

18. Gupta R, Simpson LA, Morgan DJ. Prioritizing high-value, equitable care after the COVID-19 shutdown: an opportunity for a healthcare renaissance. 

. 2021;16(2):114-116. doi:10.12788/jhm.3526

19. Kerlin MP, Costa DK, Davis BS, Admon AJ, Vranas KC, Kahn JM. Actions taken by US hospitals to prepare for increased demand for intensive care during the first wave of COVID-19: a national survey. 

. 2021;160(2):519-528. doi:10.1016/j.chest.2021.03.005

20. Grade definitions. US Preventive Services Task Force. Accessed June 7, 2021. https://www.uspreventiveservicestaskforce.org/uspstf/about-uspstf/methods-and-processes/grade-definitions

21. Das AK, Mishra DK, Gopalan SS. Reduced access to care among older American adults during CoVID-19 pandemic: results from a prospective cohort study. 

. 2021;2(4):1240. doi:10.52768/2766-7820/1240

Articles

Use of low-value care services during COVID-19 exhibits substantial heterogeneity but, on average, shows declines similar to the use of high-value services; low-value care use lags behind high-value care use in the rebound phase.

Changes in Use of Low-Value Services During the COVID-19 Pandemic

 In the move toward value-based payment, new payment models have largely been designed by payers and focused on the role of primary care providers. We examine a new phase of payment reform wherein providers, mostly specialists, are designing alternative payment models (APMs) for their own practices through a task force, called the Physician-Focused Payment Model Technical Advisory Committee, created by the Medicare Access and CHIP Reauthorization Act of 2015. Although it is a potentially notable shift in payment reform, little is known about the content of these proposals to date.

 Qualitative systematic review of physician-focused payment model proposals submitted to CMS.

 We analyzed the first wave of new payment models proposed. For each of the 24 proposals submitted by physicians and physician groups, we assessed the models on their 10 key dimensions and evaluated underlying themes across all or many of the models to gain insights into what providers are looking for in APMs within the constraints of the rules established by the HHS secretary.

 Key features of the models and our analysis include bearing financial risk, a reliance on case management, embrace of new technologies, and consideration of legal barriers.

 We discuss how specialists may help lead in the evolving payment landscape and recommend how these models might be improved. Payers and policy makers could benefit from our findings, which reflect how providers view financial risk in APMs and provide guidance on the types of payment reforms that they may embrace in the journey toward value.

In proposals of new payment models to the Physician-Focused Payment Model Technical Advisory Committee, specialists and other providers have indicated that they would be amenable to assuming financial risk.

Collectively, the payment models that the providers have designed suggest that many providers see investment in new technologies and increased time spent on case management as strategies to succeed in a value-based framework.

Policy makers should critically consider which proposals may slow healthcare spending and should consider how antikickback provisions might hinder the formation of value-based arrangements, as well as how the newly proposed payment models might interact with existing alternative payment models.

Historically, policy efforts to slow healthcare spending centered on changing the way providers are paid—especially those imposed by payers&mdash;have not always been met with enthusiasm by the provider community. Medicare’s mandatory episode-based payment models for cardiac and orthopedic care, for example, met substantial resistance from providers and were ultimately scaled back.

 Physician groups, including the American Medical Association (AMA), have in turn called for voluntary, physician-led payment innovations as an alternative to top-down payment reform. CMS, in a “New Directions” vision statement for its Center for Medicare and Medicaid Innovation, also highlighted voluntary payment models—including for specialty physicians&mdash;as a new focus.

To date, little is known about the appetite of physicians, particularly specialists, to design their own payment models aimed at achieving higher quality at lower cost. Payment innovations by public and private payers have largely focused on the role of primary care physicians and the quality of primary care decisions or outcomes. Specialists thus far have had fewer opportunities to design or implement alternative payment models (APMs) directly pertaining to their scope of practice.

In this paper, we examine the first broad, large-scale effort by providers to design new payment models for their own specialties. We evaluated the payment models submitted to the Physician-Focused Payment Model Technical Advisory Committee (PTAC)—a task force created by the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA). The PTAC reviews payment models submitted by providers and submits recommendations to the secretary of HHS, who may reject, revise, or implement each proposal. The recommendations are structured around 10 PTAC criteria (

As of November 2018, physicians and provider groups submitted 24 of 25 total PTAC proposals,

 which vary in scope, clinical focus, and economic dimensions. These proposals reflect how providers wish to be paid in a manner that generally departs from pure fee-for-service (FFS). Collectively, the submitters are composed of 6 professional societies, 13 provider groups, a state government agency, 2 academic medical centers, a nonprofit association, and an individual physician. In most of these proposals, specialists came forward with new reimbursement models involving financial risk for core services provided by their specialties, which is a stark contrast from the largely primary care—based APMs thus far. Although CMS has not yet launched demonstrations for PTAC proposals, this novel trend and the requests of enterprising providers and their organizations offer insights into possible future directions of payment reform and the challenges that such models may face.

We conducted a systematic review of the 24 physician-focused payment model proposals along 10 dimensions: patient population, population size, physician specialty, specialty size, services covered, payment type, duration of payment coverage, upside financial incentives, downside risk, and technological innovations in care delivery. These were selected to provide a combination of both broad and specialty-specific insights into the types of payment models that physicians are seeking. In addition to proposing novel APMs, the proposals also include requests to change the Medicare fee schedule, along with minor technical changes to payment rules. We excluded 1 proposal submitted by a non—healthcare provider.

Of the 24 proposals, 19 focused on new payment models for specific specialties, including surgery, gastroenterology, pulmonology, nephrology, neurology, oncology, urology, palliative care, interventional cardiology, and emergency medicine. Several key attributes of the proposals are depicted in the 

 A detailed table describing the 24 proposals across all 10 dimensions is included in the 

). Although the proposals differed in the patients, providers, and clinical scenarios that they concerned, we observed important common features across the models.

Given the HHS secretary’s criteria to improve value, most of the proposals chose to demonstrate greater value through an arrangement that required the provider to accept some financial risk for costs of care of their patients. The ways that the proposals structured that risk provide insight into the terms that were considered acceptable. In 21 of 24 proposals, provider entities asked for a shared savings and shared losses (ie, “2-sided”) payment model wherein physicians are rewarded if spending is below a prospective risk-adjusted payment benchmark and penalized if spending exceeds the benchmark. In addition, providers asked for payment adjustments, including varying degrees of bonuses, based on performance on quality measures. The 21 proposals generally capped financial gains and losses with stop-gain and stop-loss limits that varied from 4% to 20% and were frequently about 8% and 9%.

 Most accountable care organizations (ACOs) in Medicare have thus far operated without downside risk. The fact that these specialists are proposing to assume risk suggests a willingness to coordinate care, manage population health, and engage in other activities typically performed by primary care or other allied providers (eg, longitudinal follow-up). Although the submitters are a self-selected group, it is likely that recent experiments with downside risk, through Next Generation ACOs and other advancements in private payer contracts, have provided some level of familiarity with or confidence in such an approach. Another motivator may have been the potential for the providers to achieve “Advanced APM” status under MACRA to earn the 5% bonus while avoiding reporting requirements and payment adjustments under the Merit-based Incentive Payment System (MIPS). Although these models have some degree of risk-bearing, many left important payment details unresolved (eg, how payment adjustments will be calculated or how they will be distributed among providers within the APM), making it difficult to determine how much risk individual providers would bear. A recurring challenge is the trade-off between the technical difficulties of limiting financial risk to disease-specific spending (ie, attributable to the specialist) and the risk of accountability for total costs of care, which is more straightforward to calculate but more difficult to expect a single specialty to manage effectively.

The most frequently proposed mechanism for risk-bearing was bundled payments for episodes of specialty care, ranging from surgeries and dialysis to home hospitalizations and hepatitis C treatment. A common challenge among these bundled payment approaches is identifying the triggering event for an episode. For instance, the end-stage renal disease (ESRD) proposal grappled with whether starting dialysis or declining biomarkers of kidney function is the appropriate trigger for the bundle. This is a key issue, as evidence from Medicare’s voluntary bundled payment program for lower extremity joint replacements highlighted the possibility that savings may be offset by an increase in the number of bundles triggered, especially among healthier patients. (However, it should be noted that this concern has not materialized in evaluations to date of the CMS mandatory bundled payment program for lower extremity joint replacements.

) CMS should monitor how the design of the triggering event might lead to inappropriate initiation of bundles. Mitigating inappropriate triggering is likely important for its intended purpose of slowing spending. Further, although bundled payments have been shown to reduce spending for surgical episodes like joint replacements, a recent evaluation of Medicare’s bundled payment initiatives for medical conditions (including congestive heart failure, chronic obstructive pulmonary disease [COPD], sepsis, and acute myocardial infarction) found no significant changes in Medicare payments, length of stay, emergency department use, hospital readmission, or mortality.

 Policy makers should thus critically assess the cost-saving potential of proposals that bundle services for chronic medical conditions, such as ESRD.

Given the dearth of specialty-based quality measures used by payers today, it is notable that these specialists embraced the HHS secretary’s quality measurement criterion and are willing to subject their payments to quality measures specific to their domain of practice. It is worth noting, however, that many of these specialty-specific quality measures have not been previously implemented on a broad scale. Although expanding quality measurement into specialty care is a meaningful innovation, to the extent that untested measures may be poor markers of improved quality, payment adjustments may not accurately reward specialists for their efforts. Further, given that quality measures play the important role of protecting against underprovision of care (ie, skimping) in the setting of prospective payment contracts, untested quality measures may pose a patient safety risk if they fail to ensure proper care. This concern of skimping is particularly relevant in the case of vulnerable populations, who may be harmed by models that, in tying reimbursement to quality, inadvertently encourage physicians to avoid caring for those patients with the greatest medical and social needs.

 Using evidence from other countries is one approach taken by some proposals to demonstrate the validity of their measures. Experience and learning gained by testing novel quality measures designed by specialists will likely provide valuable evidence for future payment models. Finally, the process of incorporating novel measures might also allow payers to assess the effectiveness of such measures in changing provider behavior.

The proposals also frequently featured case management, often as a key contributor to savings. Twelve of the 24 proposals requested a payment for tasks such as patient monitoring and surveillance. For example, one model

 proposed payments for population health management of elderly patients with chronic conditions, and another

 proposed bundled payments for services that would include home visits and follow-up for patients recovering from acute incidents. Today, patient engagement and monitoring—which are often considered valuable and cost-effective&mdash;are not typically reimbursed by CMS. For instance, the Government Accountability Office showed in a report

 that provider groups frequently cite inadequate payment as a significant barrier to the use of telehealth in Medicare. At least half of the proposals advocated for lowering that barrier.

Most of the new models proposed per-member per-month (PMPM) payments to cover the cost of these services. The COPD and Asthma Monitoring Project (CAMP) proposal,

 for example, requested a $175 PMPM fee for remote monitoring of high-risk patients with COPD and asthma, citing studies whose findings suggest that Medicare will ultimately save money.

 Proposals that featured case management similarly argued for monitoring and surveillance of patients with ESRD,

 as well as in other clinical scenarios, such as palliative care.

includes selected case management features and proposed funding mechanisms.

 These proposals requested that CMS reward them for preventing unnecessary interventions through monitoring and incremental adjustments in care. For the providers, the proposed PMPM payments offset the costs of the time and resources required to engage in these functions. For CMS, these requests present an interesting alternative path toward potential cost control and quality improvement outside the centralized development of APMs. However, PMPM payments involve no risk sharing and are subject to the same FFS-based incentives that policy makers are looking to mitigate.

To enhance the effectiveness of case management, many proposals sought to engage allied professionals, including nurses, social workers, nutritionists, pharmacists, midwives, physical therapists, and others, in care delivery. For instance, the Annual Wellness Visit proposal

 requested that nurses be allowed to perform Medicare-billable annual wellness visits under physician supervision to increase the capability of understaffed rural health clinics. These providers already play critical roles today, but formalizing their contributions within specialty-based APMs at the request of physicians themselves is a notable development. For example, the Patient-Centered Headache Care Payment proposal,

 submitted by the American Academy of Neurology, includes a patient care coordinator as part of the Headache Care Team who is responsible for periodic patient follow-up with questionnaires, headache diaries, and referral tracking and management—important tasks that are not currently paid for by Medicare.

An issue that surfaces in the proposals is whether models based on case management services can be effective in smaller group practices with fewer patients and allied providers. These new payment models may encourage provider consolidation, as larger organizations are likely in a better position to manage financial risk and comply with quality reporting and health information technology (IT) requirements. Nonetheless, most of the proposals attempt to address how small practices could participate in the new APMs. Some, such as Project Sonar,

 propose that small practices form larger purchasing coalitions to contract with shared service providers.

 The contracted services may include quality and outcomes reporting, IT requirements, and care management, allowing small group and solo providers to engage in these important functions. The proposals indicate that this approach is being piloted in various settings across the country. The proposed models could further encourage participation among small practices by facilitating integration with MIPS—a feature that is often overlooked by models that focus on the Advanced APM track of MACRA&mdash;which is otherwise less accessible for small practices. Lastly, proposed models that rely on a specific quality management or accounting system may struggle to engage small practices, as they may face high switching costs for infrastructure and need more flexibility.

Many proposals emphasized care coordination, suggesting that additional revenues through chronic care management codes may not be adequately offsetting practice costs of coordination. However, specialty-based models inherently face challenges in coordinating across a range of specialists caring for a given patient, particularly compared with a primary care—based APM in which an internist or family physician naturally coordinates the care team. This challenge may be exacerbated by the lack of well-accepted standards for what roles specialists should play in cases where specialists and primary care physicians both coordinate care; the potential for disrupting team dynamics and confusing patients is not trivial. Therefore, the proposed models could be improved by more specific delineations of provider roles and responsibilities, particularly with respect to how the patient’s primary care provider fits into the model. Moreover, the models would benefit from clearer guidance on how savings could be shared among members of the care team. Given the lack of precedent, establishing standards for the distribution of responsibilities and incentives across specialties could be a meaningful contribution of PTAC. That said, policy makers should be cautious about the potential of coordination to generate savings, given sparse evidence that care coordination actually lowers spending (although it may improve patient care).

Several groups proposed the implementation of innovative technologies to support care delivery. For example, Project CAMP

 proposed a smartphone application to track patient-reported symptoms, computerized decision support and “smart alarms” for pulmonologists, and Bluetooth-enabled digital peak flow meters that transmit clinical data to providers in real time. The Home Hospitalization proposal

 relies on Bluetooth-enabled medical devices (eg, blood-pressure cuff, pulse oximeter, scale, stethoscope) for biometric data tracking as part of its tablet-based telehealth platform that also features video communication. Selected technological innovations in the proposed models are described in 

Providers view new technologies as a way to improve patient engagement and increase the value of care. However, a central question remains as to who will pay for such technologies. Five of the proposals requested payments for the use of new technologies. However, CMS does not usually fund start-up costs behind new technologies, especially those not approved by the FDA. HHS, in a recent comment on this matter, indicated that payment models testing proprietary technology, such as Project CAMP’s peak flow meter, will not be pursued.

 Thus, whether paying for such new technologies will induce investment in telehealth and whether it can save money remain open questions. The risk of underinvestment in technologies that benefit patients through improved monitoring or decision making also continues to exist. To this end, CMS might consider increasing chronic care management fees to account for the technologies needed to manage patients in the digital age. That aside, the proposed models would benefit by considering more creative ways to support new technologies than simply requesting new billing codes. Innovative strategies may include folding the start-up costs into an enhanced bundled payment or sharing risk with the companies that manufacture the technologies, which are approaches taken by a few proposals.

Five of the proposals requested safe harbor designations from Stark laws, suggesting that these laws—which exist at both the federal and state levels and prohibit physicians from making referrals to those with whom a financial relationship exists&mdash;may act as unintended barriers to improving value in care delivery. The American Hospital Association advocated in Congressional testimony for exceptions under the antikickback statute to advance new APMs, arguing that the statute’s strict liability penalizes providers seeking to collaborate and create efficiencies.

 The AMA currently supports a US Senate bill that would allow such exceptions.

 Most recently, 2 former HHS secretaries called for safe harbor exceptions to Stark laws, which they argue have become roadblocks in the move toward a value-based system by making hospitals reluctant to reward physicians for best practices and causing confusion from inconsistent interpretation.

 Given that APMs rely on more teamwork among providers across specialties, exemptions from antikickback statutes may facilitate contractual agreements among providers. Whether such relief may be generalized in APMs through federal regulation or whether decisions will be made on a case-by-case basis is unknown. A recent Request for Information from CMS calling for ideas on Stark law reform signals momentum on this issue.

 However, although safe harbor designations and other waivers of antikickback provisions may facilitate the formation of APMs that span the care continuum, policy makers should continue to critically assess financial relationships that have the potential to create perverse incentives for overutilization.

The Affordable Care Act commenced a move toward value-based payments. In the last decade, new payment models have largely been designed by payers and policy makers and mostly focus on the role of primary care providers. The payment models submitted to the PTAC, in contrast, were designed by providers themselves with an explicit goal of improving the value of care in specialty care settings. Among these are some potentially innovative ideas, such as longitudinal accountability centered on specialty care, more granular measurement of quality using clinical data, care coordination for chronic illness, and novel technology that aids in patient monitoring and communication. These models propose to increase value through financial risk on the costs of care, albeit with limited downside risk. As of November 2018, 13 of these proposals had been recommended for testing or implementation, 5 had not been recommended, and others were still being considered. More than a dozen additional letters of intent have been submitted to the PTAC, indicating continued momentum in this effort. Most of these future proposals will also focus on specialties, such as radiation oncology, orthopedic surgery, and allergy, asthma, and immunology.

Of note, it is unclear how these proposed models would interact with existing APMs. Many of the proposals overlap clinically with existing models, such as Bundled Payments for Care Improvement models and the Medicare Shared Savings Program (MSSP). This overlap raises 2 issues. First, it may be unclear how to reconcile costs of care when the same patient is included in different models. Second, when participating in parallel programs, evaluation of the impact of each model becomes more challenging methodologically. Some have called for more guidance from CMS on potential interactions and if certain models may take precedence in financial reconciliation or distributions of shared savings or losses.

 Several models considered this overlap in their APM design. For example, the Comprehensive Care Physician Payment Model indicated how its proposals could be added as supplemental payments or penalties on top of MSSP ACOs.

 However, most proposals did not address possible model interactions.

These payment model proposals by healthcare providers offer initial insight into the characteristics of value-based payment arrangements that specialists in the United States may be willing to accept. In addition, they provide guidance on where physicians may see opportunities to improve the delivery of care. They may help inform the future direction of payment reform—toward more risk for specialties, a focus on case management, and new technologies to improve patient engagement. Many of these models face obstacles, from the reorganization of provider organizations to legal barriers, and pose certain risks that need to be mitigated, including safety concerns around untested quality measures and preservation of some FFS incentives. Regardless, these models suggest that some providers, especially specialists, might support payment reform if it is done in a way that incorporates their input. Ultimately, the road from payment reform to delivery system improvement is lengthy&mdash;requiring ingenuity, diligence, and commitment from providers willing to challenge the status quo. This new work by forward-thinking providers to shape payment and accountability for themselves is an important step in this journey.

 Harvard Medical School (SG, TGF, ZS), Boston, MA; Massachusetts General Hospital (TGF, ZS), Boston, MA; Johns Hopkins Medicine (KKP), Baltimore, MD; The Brookings Institution (KKP), Washington, DC.

 Supported by a grant from the Office of the Director, National Institutes of Health (NIH Director’s Early Independence Award, DP5 OD024564, to Dr Song).

 Dr Ferris is a member of the Physician-Focused Payment Model Technical Advisory Committee (PTAC), to which the proposals analyzed in this study were submitted. Dr Patel is a vice president of Johns Hopkins Health System, a member of PTAC, and a board member of Dignity Health, SSM Health, and Paladina Health. Dr Song reports speaking fees from the International Foundation of Employee Benefit Plans, outside the submitted work. Mr Gondi reports 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 (SG, TGF, KKP, ZS); acquisition of data (SG, ZS); analysis and interpretation of data (SG, TGF, KKP, ZS); drafting of the manuscript (SG, KKP); critical revision of the manuscript for important intellectual content (SG, TGF, KKP, ZS); obtaining funding (ZS); administrative, technical, or logistic support (SG); and supervision (TGF, ZS).

 Zirui Song, MD, PhD, Department of Health Care Policy, Harvard Medical School, 180A Longwood Ave, Boston, MA 02115. Email: song@hcp.med.harvard.edu.

1. CMS, HHS. Medicare program; cancellation of Advancing Care Coordination Through Episode Payment and Cardiac Rehabilitation Incentive payment models; changes to Comprehensive Care for Joint Replacement payment model (CMS-5524-P). 

2. Centers for Medicare & Medicaid Services: Innovation Center new direction. CMS website. innovation.cms.gov/Files/x/newdirection-rfi.pdf. Accessed November 25, 2017.

3. Bowen A, Burton N, Muhlestein D. Medicare alternative payment models: not every provider has a path forward. Leavitt Partners website. leavittpartners.com/whitepaper/medicare-alternative-payment-models-not-every-provider-path-forward. Published September 11, 2017. Accessed December 5, 2017.

4. Proposal submissions: Physician-Focused Payment Model Technical Advisory Committee. Office of the Assistant Secretary for Planning and Evaluation website. aspe.hhs.gov/proposal-submissions-physician-focused-payment-model-technical-advisory-committee. Accessed November 30, 2017.

5. CMS, HHS. Medicare program; Merit-Based Incentive Payment System (MIPS) and Alternative Payment Model (APM) incentive under the physician fee schedule, and criteria for physician-focused payment models. 

6. 2018 physician specialty data report: executive summary. Association of American Medical Colleges website. aamc.org/download/492910/data/2018executivesummary.pdf. Published 2018. Accessed July 29, 2019.

7. Fisher ES. Medicare’s bundled payment program for joint replacement: promise and peril? 

. 2016;316(12):1262-1264. doi: 10.1001/jama.2016.12525.

8. Joynt Maddox KE, Orav EJ, Zheng J, Epstein AM. Evaluation of Medicare’s bundled payments initiative for medical conditions. 

. 2018;379(3):260-269. doi: 10.1056/NEJMsa1801569.

9. Joynt Maddox KE. Financial incentives and vulnerable populations—will alternative payment models help or hurt? 

. 2018;378(11):977-979. doi: 10.1056/NEJMp1715455.

10. Coalition to Transform Advanced Care. Advanced Care Model (ACM) service delivery and advanced alternative payment model. Office of the Assistant Secretary for Planning and Evaluation website. aspe.hhs.gov/system/files/pdf/255906/ProposalACMCTAC.pdf. Published February 7, 2017. Accessed November 30, 2017.

11. Icahn School of Medicine at Mount Sinai. “HaH Plus” (Hospital at Home Plus): provider-focused payment model. Office of the Assistant Secretary for Planning and Evaluation website. aspe.hhs.gov/system/files/pdf/255906/HaHPlusProviderFocusedPaymentModel.pdf. Published May 2, 2017. Accessed November 30, 2017.

Physician groups have begun designing alternative payment models for their own specialties, proposing that CMS include financial risk, funding for new technologies, and legal waivers.

Physician-Initiated Payment Reform: A New Path Toward Value

 As US healthcare spending increases, insurers are focusing attention on decreasing potentially avoidable specialist care. Little recent research has assessed whether the design of modern health maintenance organization (HMO) insurance is associated with lower utilization of outpatient specialty care versus less restrictive preferred provider organization (PPO) plans.

 Observational study of Massachusetts residents aged 21 to 64 years with any HMO or PPO insurance coverage from 2010 to 2013.

 We examined rates and patterns of primary care visits, new specialist visits, and specialist spending among HMO versus PPO enrollees. We estimated multivariable regression models for each outcome, adjusting for patient and insurance characteristics.

 From 2010 to 2013, 546,397 and 295,427 individuals had continuous HMO or PPO coverage, respectively. HMO patients had fewer annual new specialist visits per member versus PPO patients (unadjusted, 0.37 vs 0.43), a difference after adjustment of 0.05 annual visits, or a 12% relative decrease among HMO members (

 <.001). These visits were more likely to be with a specialist in the same health system as the patient’s primary care physician (44.9% vs 40.7%; adjusted difference, 2.8 percentage points; 

 <.001). Mean annual spending on new specialist visits and subsequent follow-up per member was lower in HMO versus PPO patients (unadjusted, $104.10 vs $128.10), translating to 12% lower annual spending (adjusted difference, —$16.26; 

 Having HMO insurance was associated with lower rates of new specialist visits and lower spending on specialist visits, and these visits were less likely to occur across multiple health systems. The impact of this change on overall spending and clinical outcomes remains unknown.

Is specialist “gatekeeping” in modern health maintenance organization (HMO) insurance associated with differences in outpatient patterns of care?

Compared with preferred provider organization insurance, HMO insurance was associated with lower rates of new specialist visits.

These visits were less likely to occur across multiple health systems.

HMO insurance was also associated with 12% lower specialist visit spending, which was largely driven by lower use, not lower price.

HMO gatekeeping may meaningfully reduce specialist utilization, although the impact of this change on overall spending and clinical outcomes remains unknown.The majority of private health insurance options in the United States fall into 1 of 2 broad categories: preferred provider organization (PPO) plans and health maintenance organization (HMO) plans. PPO plans generally have broadly inclusive physician networks with little barrier to self-referral, whereas lower-premium HMO plans generally have less inclusive networks and rely on provider-facing managed care strategies, such as requiring primary care physicians (PCPs) to approve referrals to access specialists (ie, “gatekeeping”).

 As of 2014, across the United States, HMO insurance products accounted for 24% of enrollment in the employer-sponsored insurance market, whereas PPO products accounted for nearly 50% of the market.

 In Massachusetts, the setting for this analysis, HMO insurance was 39% of the commercial market, with an additional 19% with point-of-service (POS) coverage and 37% with PPO coverage in 2016.

 One concept underpinning the provider-facing strategies used by HMOs is that PCPs can reduce the use of low-value specialty care and ensuing downstream utilization by either treating a condition in the primary care setting or directing specialty referrals to higher-value providers within their health system. This is in contrast to the consumer-facing cost-control strategies seen in PPO plans that may impose higher cost sharing but retain open provider choice and access at the discretion of the enrollee.

In general, HMO plans require enrollees to identify a PCP to help direct their downstream utilization. Conceptually and empirically, a single PCP with overall responsibility for a patient’s care could lead to better-coordinated care while reducing potentially avoidable outpatient specialist referrals.

 Such gatekeeping arrangements, however, can also create problems for both patients and PCPs when disagreements arise over whether a specialist is needed.

The extent to which such gatekeeping affects specialist utilization or costs has mixed results in studies that are largely more than 15 years old.

 Little recent research has assessed whether the design of modern HMO insurance is associated with lower utilization of outpatient specialty care compared with modern PPO plans. Moreover, broadly understanding the effect of insurance design on specialty utilization across a large heterogeneous payer market is an important policy-relevant question recently made feasible through the availability of statewide all-payer claims databases (APCDs).

We used the Massachusetts APCD to examine the association between insurance design and outpatient care utilization patterns in multiple HMO and PPO arrangements. In Massachusetts, the primary distinction between HMO and PPO plans relates to gatekeeping requirements (ie, a referral requires PCP approval), in contrast to the existence of “closed-model” HMOs like Kaiser Permanente (ie, regardless of PCP approval, referrals outside the internal network are largely unavailable), which do not currently have market share in the state. We hypothesized that HMO enrollment would be associated with more PCP visits and fewer new specialist visits, with a larger share of specialty care received within the PCP’s health system.

The Massachusetts APCD contains detailed data on healthcare utilization, insurance eligibility, and provider credentialing across all commercial payers and public health insurance programs in Massachusetts, representing approximately 90% of the nonelderly population in Massachusetts.

Our study cohort included all Massachusetts residents aged 21 to 64 years who had 4 calendar years of continuous enrollment with any commercial HMO or PPO product from an APCD-participating commercial insurance provider from 2010 to 2013; public insurance was not included. We focused on members with this length of continuous enrollment because we were interested in the long-term association between HMO or PPO membership and specialist use among patients with stable benefit design. Enrollees may change from an HMO to a PPO plan or vice versa because of their plans for specialist use, which we wanted to reduce as a source of bias. We did not include children given the unique characteristics of the pediatric healthcare market in the state. The key exposure of interest was whether an individual was a continuous member of an HMO versus a PPO product, as defined by the insurer submitting data in the APCD member eligibility file. These designations are used by the insurers themselves to distinguish HMO from PPO plans for administering health plan benefits and requirements. We defined an HMO product as any product designated as HMO or POS (9.6% of all members, or 15.2% of HMO members in the final study sample), which have very similar benefit designs among the large insurers in Massachusetts.

 PPO products were defined as those with a designation of PPO or exclusive provider organization (EPO; 1.5% of all members, or 4.3% of PPO members in the final study sample), which have similar flexibility to PPO plans.

 We used diagnoses from all outpatient and inpatient claims in 2010 to measure baseline comorbidities for risk adjustment. We then performed cross-sectional analyses that focused on the period from 2011 to 2013. We defined continuous enrollment as 11 or more months of insurance coverage in a year, such that those with a short period of discontinuous enrollment, often due to an administrative error, were not excluded from the analyses. This project was approved by the Committee on Human Subjects at Harvard Medical School.

To attribute patients to PCPs, we assigned enrollees to the PCP with the plurality of office visits in a calendar year (“empirical” PCP assignment).

 Nonphysician primary care providers, such as nurse practitioners or physician assistants, were not included because they were not consistently identified in provider classifications provided by insurers. We identified visits to primary care based on claim lines with evaluation and management (E&M) Current Procedural Terminology (CPT) codes to physicians classified as general internal medicine, general practice, or family medicine. If there were no primary care visits, we assigned enrollees to the non—primary care office-based specialist (eg, excluding radiology, pathology) with the plurality of visits. Any ties between physicians were broken randomly, and enrollees with no visits were not assigned to a PCP or another physician. We also examined patterns of PCP assignment as recorded by insurers, which we labeled “insurance” PCP assignment.

We also assigned physicians to larger physician organizations or health systems in Massachusetts using a publicly available database of Massachusetts hospital and physician group national provider identifiers,

 grouped into physician organizations using the affiliations defined by the Massachusetts Health Care Delivery System Map, a manually curated listing of hospital and clinic affiliations within larger provider organizations (eg, which satellite clinics are part of Partners HealthCare) published by the Blue Cross Blue Shield of Massachusetts Foundation.

 We assigned 76% of physicians to organizations using this approach, and the remainder were largely unaffiliated independent physicians.

We assessed several dimensions of outpatient care delivery potentially influenced by type of insurance coverage. For primary care utilization, we calculated the average annual number of PCP visits of any kind per member from 2011 to 2013 and the proportion of patients with no PCP visits in a year. For primary care assignment, we assessed the proportion of patients ever assigned a PCP by their insurer and the percentage of all PCP visits occurring with those providers, as well as the consistency of empirically attributed and insurer-defined PCP assignment measured by the number of assigned PCPs per member over the 3-year study period.

We defined a new specialist visit as any claim containing a new visit E&M CPT code with a non-PCP, which can only be billed if a patient has not been seen by any physician within that practice in the previous 3 years.

 We then measured characteristics of new specialist visits that could be associated with higher likelihood of PCP-initiated referral, as opposed to patient self-referral. We did not attempt to identify specialist-to-specialist referrals because HMO referral policy focuses on PCPs as the key provider for approving referrals. We examined whether a new specialist visit’s claim designated a referring physician and whether that referring physician was the patient’s assigned PCP. We also measured the time in days between a new specialist visit and the most recent PCP visit for the patient in the past 365 days. When we had information on organizational affiliation for both the PCP and the specialist (“health system match” was available for 59% of new specialist visits), we assessed whether the PCP for the most recent visit or the assigned PCP was affiliated with the same provider organization as the specialist for the new visit.

We measured the percentage of patients with any new specialist visit from 2011 to 2013 and the average number of new specialist visits per patient in that period. Because HMO gatekeeping may have its most pronounced effect on the highest utilizers, we also examined the proportion of patients seeing 3 or more new specialists or seeing new specialists in 2 or more provider organizations, which we identified as approximately the top decile of utilization for the entire study population.

Lastly, we measured the average cost per new specialist visit and all subsequent visits with that specialist over the subsequent year by summing all costs (including insurer reimbursement and any out-of-pocket spending) from all E&M claims for new specialist visits for members, as well as E&M costs from any downstream established office visits with the same specialist seen in the following year. To test whether differences in spending by insurance type were driven by differences in price versus quantity of specialist visits, we standardized specialist visit prices using the average HMO cost per evaluation and management CPT code as a standard price for all encounters.

To measure comorbidities, we calculated Hierarchical Condition Category (HCC) scores from all diagnosis codes in the baseline year using software available from CMS. The HCC score represents the predicted risk of healthcare spending beyond average in the next year; a score of 1.0 corresponds to a risk profile for a Medicare enrollee with average expected annual spending.

 To derive a proxy measure of socioeconomic status, we linked individuals’ 5-digit zip code, the most granular geographic unit available, to the publicly available area deprivation index (ADI), a measure derived using a weighted combination of 17 Census-level indicators of socioeconomic disadvantage.

 We also included the median co-pays for primary care and specialist physician visits for each separate insurance product in 2010 as a measure of plan generosity. Other variables included member age as of December 31, 2010; sex; fixed effects for insurance carrier (eg, Blue Cross Blue Shield vs UnitedHealthcare); insurance plan risk type (fully insured vs self-insured); and employer size (ranging from individual to ≥500 employees; see 

We compared the characteristics of HMO and PPO patients and the unadjusted rates of the outcomes using χ

 tests. To estimate adjusted comparisons between HMO and PPO patients, we fitted a series of multivariable regression models for each outcome, with the key explanatory variable being an indicator for whether a patient was a member of an HMO versus a PPO insurance product. We fitted models with different link functions depending on the outcome: linear models for visit counts, costs, and visit intervals; Poisson models for counts of unique physicians or health systems; and logistic models for all other outcomes, which were binary or fractional in nature. All models adjusted for patient age, sex, HCC score, ADI, median primary care and specialist co-pays in that plan (with the exception of spending outcomes, which excluded co-pay covariates), and employer size. We also included a fixed effect for insurance carrier to control for unmeasured characteristics of carriers’ networks or policies that might also influence the outcomes of interest. As a sensitivity analysis, we also estimated logistic regression models predicting the propensity to have HMO versus PPO insurance and then used propensity score weighting to more robustly address imbalance on observable variables between HMO and PPO patients and replicated all of our model results (see 

Because of the mix of functional forms (eg, linear, Poisson, logistic), we used bootstrap methods within each model to present the effects of HMO membership as the adjusted difference in each individual outcome between the average HMO member and the average PPO member. For each effect, we drew 1000 simulations of the set of regression coefficients using the asymptotic multivariate normal approximation to the log-likelihood function, and we present the mean and 95% CI from these simulations.

 Analyses were performed in R version 3.1.3 and SAS version 9.4 (SAS Institute; Cary, North Carolina). The 95% CI around reported estimates reflects 0.025 in each tail, or 

The study sample included 546,397 and 295,427 individuals with 4 years of continuous coverage in an HMO or PPO insurance product, respectively, from 2010 to 2013. HMO enrollees were younger than PPO patients (45.7 vs 46.1 years, respectively; 

 <.001), had slightly lower HCC scores (0.34 vs 0.36; 

 = .01), had higher ADI scores (ie, were more socioeconomically disadvantaged; 83.8 vs 80.4; 

 <.001), and were significantly less likely to be employed by an employer with 500 or more employees (51.7% vs 81.1%; 

 <.001; Table 1). Sample characteristics after propensity score weighting were identically balanced (

Is specialist “gatekeeping” in modern health maintenance organization (HMO) insurance associated with differences in outpatient care? The study finds that HMO gatekeeping may meaningfully reduce specialist utilization.

Zirui Song, MD, PhD

Articles

Changes in Use of Low-Value Services During the COVID-19 Pandemic

Physician-Initiated Payment Reform: A New Path Toward Value

Gatekeeping and Patterns of Outpatient Care Post Healthcare Reform

Incorporating Value Into Physician Payment and Patient Cost Sharing

The Emerging Business Models and Value Proposition of Mobile Health Clinics

Addressing the Primary Care Workforce Crisis