How Should Medicare Pay for Artificial Intelligence?

This Viewpoint examines artificial intelligence­enabled clinical services, existing payment structures, and the economics of artificial intelligence pricing.

Artificial intelligence in Medicare: utilization, spending, and access to AI-enabled clinical software.

OBJECTIVES: In 2018, CMS established reimbursement for the first Medicare-covered artificial intelligence (AI)-enabled clinical software: CT fractional flow reserve (FFRCT) to assist in the diagnosis of coronary artery disease. This study quantified Medicare utilization of and spending on FFRCT from 2018 through 2022 and characterized adopting hospitals, clinicians, and patients. STUDY DESIGN: Analysis, using 100% Medicare fee-for-service claims data, of the hospitals, clinicians, and patients who performed or received coronary CT angiography with or without FFRCT. METHODS: We measured annual trends in utilization of and spending on FFRCT among hospitals and clinicians from 2018 through 2022. Characteristics of FFRCT-adopting and nonadopting hospitals and clinicians were compared, as well as the characteristics of patients who received FFRCT vs those who did not. RESULTS: From 2018 to 2022, FFRCT billing volume in Medicare increased more than 11-fold (from 1083 to 12,363 claims). Compared with nonbilling hospitals, FFRCT-billing hospitals were more likely to be larger, part of a health system, nonprofit, and financially profitable. FFRCT-billing clinicians worked in larger group practices and were more likely to be cardiac specialists. FFRCT-receiving patients were more likely to be male and White and less likely to be dually enrolled in Medicaid or receiving disability benefits. CONCLUSIONS: In the initial 5 years of Medicare reimbursement for FFRCT, growth was concentrated among well-resourced hospitals and clinicians. As Medicare begins to reimburse clinicians for the use of AI-enabled clinical software such as FFRCT, it is crucial to monitor the diffusion of these services to ensure equal access.

Mitigating Racial And Ethnic Bias And Advancing Health Equity In Clinical Algorithms: A Scoping Review.

In August 2022 the Department of Health and Human Services (HHS) issued a notice of proposed rulemaking prohibiting covered entities, which include health care providers and health plans, from discriminating against individuals when using clinical algorithms in decision making. However, HHS did not provide specific guidelines on how covered entities should prevent discrimination. We conducted a scoping review of literature published during the period 2011-22 to identify health care applications, frameworks, reviews and perspectives, and assessment tools that identify and mitigate bias in clinical algorithms, with a specific focus on racial and ethnic bias. Our scoping review encompassed 109 articles comprising 45 empirical health care applications that included tools tested in health care settings, 16 frameworks, and 48 reviews and perspectives. We identified a wide range of technical, operational, and systemwide bias mitigation strategies for clinical algorithms, but there was no consensus in the literature on a single best practice that covered entities could employ to meet the HHS requirements. Future research should identify optimal bias mitigation methods for various scenarios, depending on factors such as patient population, clinical setting, algorithm design, and types of bias to be addressed.

Improving the Performance of Risk Adjustment Systems: Constrained Regressions, Reinsurance, and Variable Selection.

Modifications of risk-adjustment systems used to pay health plans in individual health insurance markets typically seek to reduce selection incentives at the individual and group levels by adding variables to the payment formula. Adding variables can be costly and lead to unintended incentives for upcoding or service utilization. While these drawbacks are recognized, they are hard to quantify and difficult to balance against the concrete, measurable improvements in fit that may be achieved by adding variables to the formula. This paper takes a different approach to improving the performance of health plan payment systems. Using the HHS-HHC V0519 model from the Marketplaces as a starting point, we constrain fit at the individual and group level to be as good or better than the current payment model while reducing the number of variables in the model. We introduce three elements in the design of plan payment: reinsurance, constrained regressions, and machine learning methods for variable selection. The fit performance of our alternative formulas with many fewer variables is as good or better than the current HHS-HHC V0519 formula.

The effect of Medicaid expansion on Medicaid participation, payer mix, and labor supply in primary care.

Evidence of increased health care utilization associated with the Medicaid expansion suggests that clinicians increased capacity to meet demand. However, little is known about the mechanism underlying this response. Using a novel source of all-payer data, we quantified clinicians' response to the Medicaid expansion - examining whether and how they changed their Medicaid participation decisions, payer mix, and overall labor supply. Primary care clinicians in expansion states provided an average of 49 additional appointments per year (a 21% relative increase) for patients insured by Medicaid, compared to clinicians in non-expansion states - with new-patient visits representing half (25 appointments) of this overall increase. Clinicians did not increase their labor supply to accommodate these additional appointments. They instead offset the 1.7 percentage point average increase in Medicaid payer mix with an equivalent reduction in commercial payer mix. However, this reduction in commercial patient share represented only a 2.8% relative decrease, with commercially insured patients still comprising the majority of the average clinician's patient panel. Subsample analyses revealed a larger increase in care for Medicaid patients among clinicians with high Medicaid participation preceding the eligibility expansion.

Identifying undercompensated groups defined by multiple attributes in risk adjustment.

OBJECTIVE: To identify undercompensated groups in plan payment risk adjustment that are defined by multiple attributes with a systematic new approach, improving on the arbitrary and inconsistent nature of existing evaluations. METHODS: Extending the concept of variable importance for single attributes, we construct a measure of 'group importance' in the random forests algorithm to identify groups with multiple attributes that are undercompensated by current risk adjustment formulas. Using 2016-2018 IBM MarketScan and 2015-2018 Medicare claims and enrolment data, we evaluate two risk adjustment scenarios: the risk adjustment formula used in the individual health insurance Marketplaces and the risk adjustment formula used in Medicare. RESULTS: A number of previously unidentified groups with multiple chronic conditions are undercompensated in the Marketplaces risk adjustment formula, while groups without chronic conditions tend to be overcompensated in the Marketplaces. The magnitude of undercompensation when defining groups with multiple attributes is many times larger than with single attributes. No complex groups were found to be consistently undercompensated or overcompensated in the Medicare risk adjustment formula. CONCLUSIONS: Our method is effective at identifying complex undercompensated groups in health plan payment risk adjustment where undercompensation creates incentives for insurers to discriminate against these groups. This work provides policy-makers with new information on potential targets of discrimination in the healthcare system and a path towards more equitable health coverage.

Impact of Changes in Chain Restaurant Calories over Time on Obesity Risk.

BACKGROUND: Prior research on the restaurant environment and obesity risk is limited by cross-sectional data and a focus on specific geographic areas. OBJECTIVE: To measure the impact of changes in chain restaurant calories over time on body mass index (BMI). DESIGN: We used a first-difference model to examine whether changes from 2012 to 2015 in chain restaurant calories per capita were associated with percent changes in BMI. We also examined differences by race and county income, restaurant type, and initial body weight categories. SETTING: USA (207 counties across 39 states). PARTICIPANTS: 447,873 adult patients who visited an athenahealth medical provider in 2012 and 2015 where BMI was measured. MAIN OUTCOMES MEASURED: Percent change in objectively measured BMI from 2012 to 2015. RESULTS: Across all patients, changes in chain restaurant calories per capita were not associated with percent changes in BMI. For Black or Hispanic adults, a 10% increase in exposure to chain restaurant calories per capita was associated with a 0.16 percentage-point increase in BMI (95% CI 0.03, 0.30). This translates into a predicted weight increase of 0.89 pounds (or a 0.53% BMI increase) for an average weight woman at the 90th percentile of increases in the restaurant environment from 2012 to 2015 versus an increase 0.39 pounds (or 0.23% BMI increase) at the 10th percentile. Greater increases in exposure to chain restaurant calories also significantly increased BMI for Black or Hispanic adults receiving healthcare services in lower-income counties (0.26, 95% CI 0.04, 0.49) and with overweight/obesity (0.16, 95% CI 0.04, 0.29). LIMITATIONS: Generalizability to non-chain restaurants is unknown and the sample of athenahealth patients is relatively homogenous. CONCLUSIONS: Increased exposure to chain restaurant calories per capita was associated with increased weight gain among Black or Hispanic adults.

Fair regression for health care spending.

The distribution of health care payments to insurance plans has substantial consequences for social policy. Risk adjustment formulas predict spending in health insurance markets in order to provide fair benefits and health care coverage for all enrollees, regardless of their health status. Unfortunately, current risk adjustment formulas are known to underpredict spending for specific groups of enrollees leading to undercompensated payments to health insurers. This incentivizes insurers to design their plans such that individuals in undercompensated groups will be less likely to enroll, impacting access to health care for these groups. To improve risk adjustment formulas for undercompensated groups, we expand on concepts from the statistics, computer science, and health economics literature to develop new fair regression methods for continuous outcomes by building fairness considerations directly into the objective function. We additionally propose a novel measure of fairness while asserting that a suite of metrics is necessary in order to evaluate risk adjustment formulas more fully. Our data application using the IBM MarketScan Research Databases and simulation studies demonstrates that these new fair regression methods may lead to massive improvements in group fairness (eg, 98%) with only small reductions in overall fit (eg, 4%).

Comparison of crowd-sourced, electronic health records based, and traditional health-care based influenza-tracking systems at multiple spatial resolutions in the United States of America.

BACKGROUND: Influenza causes an estimated 3000 to 50,000 deaths per year in the United States of America (US). Timely and representative data can help local, state, and national public health officials monitor and respond to outbreaks of seasonal influenza. Data from cloud-based electronic health records (EHR) and crowd-sourced influenza surveillance systems have the potential to provide complementary, near real-time estimates of influenza activity. The objectives of this paper are to compare two novel influenza-tracking systems with three traditional healthcare-based influenza surveillance systems at four spatial resolutions: national, regional, state, and city, and to determine the minimum number of participants in these systems required to produce influenza activity estimates that resemble the historical trends recorded by traditional surveillance systems. METHODS: We compared influenza activity estimates from five influenza surveillance systems: 1) patient visits for influenza-like illness (ILI) from the US Outpatient ILI Surveillance Network (ILINet), 2) virologic data from World Health Organization (WHO) Collaborating and National Respiratory and Enteric Virus Surveillance System (NREVSS) Laboratories, 3) Emergency Department (ED) syndromic surveillance from Boston, Massachusetts, 4) patient visits for ILI from EHR, and 5) reports of ILI from the crowd-sourced system, Flu Near You (FNY), by calculating correlations between these systems across four influenza seasons, 2012-16, at four different spatial resolutions in the US. For the crowd-sourced system, we also used a bootstrapping statistical approach to estimate the minimum number of reports necessary to produce a meaningful signal at a given spatial resolution. RESULTS: In general, as the spatial resolution increased, correlation values between all influenza surveillance systems decreased. Influenza-like Illness rates in geographic areas with more than 250 crowd-sourced participants or with more than 20,000 visit counts for EHR tracked government-lead estimates of influenza activity. CONCLUSIONS: With a sufficient number of reports, data from novel influenza surveillance systems can complement traditional healthcare-based systems at multiple spatial resolutions.

Physicians' Participation In Medicaid Increased Only Slightly Following Expansion.

While most primary care physicians treated at least one Medicaid patient in 2013, Medicaid represented a small share of their payer mix. Following Medicaid eligibility expansion in 2014, most physicians maintained or slightly increased their Medicaid participation, with greater increases observed in states that expanded eligibility. Nevertheless, Medicaid patients remained concentrated among relatively few physicians after expansion.

Accurate Influenza Monitoring and Forecasting Using Novel Internet Data Streams: A Case Study in the Boston Metropolis.

BACKGROUND: Influenza outbreaks pose major challenges to public health around the world, leading to thousands of deaths a year in the United States alone. Accurate systems that track influenza activity at the city level are necessary to provide actionable information that can be used for clinical, hospital, and community outbreak preparation. OBJECTIVE: Although Internet-based real-time data sources such as Google searches and tweets have been successfully used to produce influenza activity estimates ahead of traditional health care-based systems at national and state levels, influenza tracking and forecasting at finer spatial resolutions, such as the city level, remain an open question. Our study aimed to present a precise, near real-time methodology capable of producing influenza estimates ahead of those collected and published by the Boston Public Health Commission (BPHC) for the Boston metropolitan area. This approach has great potential to be extended to other cities with access to similar data sources. METHODS: We first tested the ability of Google searches, Twitter posts, electronic health records, and a crowd-sourced influenza reporting system to detect influenza activity in the Boston metropolis separately. We then adapted a multivariate dynamic regression method named ARGO (autoregression with general online information), designed for tracking influenza at the national level, and showed that it effectively uses the above data sources to monitor and forecast influenza at the city level 1 week ahead of the current date. Finally, we presented an ensemble-based approach capable of combining information from models based on multiple data sources to more robustly nowcast as well as forecast influenza activity in the Boston metropolitan area. The performances of our models were evaluated in an out-of-sample fashion over 4 influenza seasons within 2012-2016, as well as a holdout validation period from 2016 to 2017. RESULTS: Our ensemble-based methods incorporating information from diverse models based on multiple data sources, including ARGO, produced the most robust and accurate results. The observed Pearson correlations between our out-of-sample flu activity estimates and those historically reported by the BPHC were 0.98 in nowcasting influenza and 0.94 in forecasting influenza 1 week ahead of the current date. CONCLUSIONS: We show that information from Internet-based data sources, when combined using an informed, robust methodology, can be effectively used as early indicators of influenza activity at fine geographic resolutions.

Reframing the unaffordability debate: patient responsibility for physician care.

OBJECTIVES: Public discussion suggests that rising out-of-pocket costs have dramatically reduced the value of insurance and led to patients doing without needed care. Our aim was to ascertain trends in patient responsibility for cost sharing. STUDY DESIGN: We used data from an organization that serves over 78,000 healthcare providers and has access to visit-level data, including the amounts paid by patients. These practices are broadly representative of physicians and patients nationally. METHODS: We analyzed trends in patient obligations among a cohort of about 21,000 providers in 1078 practices who had used athenahealth software since 2011, including primary care physicians, obstetricians and gynecologists, surgeons, and some other specialists. Our analysis focused on what commercially insured patients pay out of pocket when seeking ambulatory care. RESULTS: The average patient obligation for approximately 2.5 million primary care visits each year rose from $23.52 per visit in 2011 to $26.40 per visit in 2015, for an overall increase of $2.88, or about 3% annually. This rate of increase is moderate and below growth in overall healthcare spending during the same time period. CONCLUSIONS: Average increases in patient obligations for outpatient visits in recent years have been fairly moderate, and multiple sources of survey data suggest that consumers' concerns about overall affordability are decreasing. The high cost of healthcare continues to pose challenges, both at the individual level and for society as a whole. Nevertheless, it is important that potential strategies to improve affordability are informed by trends in patient obligations.

A comparison of conservative interventions and their effectiveness for coccydynia: a systematic review.

STUDY DESIGN: Systematic literature review. OBJECTIVE: To investigate the various conservative interventions for coccydynia and determine their effectiveness. BACKGROUND: Coccydynia is the term used to describe pain in the coccygeal region. Pain in this region is typically caused by local trauma. Sitting is typically the most painful position for patients having coccydynia. Various methods of treating coccydynia are found in the literature but to our knowledge no systematic review has been performed that compared the effectiveness of these interventions. METHODS: Searches were performed for research studies using electronic databases (Cochrane Library, CINAHL, Medline, PEDro, Scopus, and Sports Discus) from January 2002 through July 2012. The quality of the papers was assessed using the GRADE approach. RESULTS: Seven papers were located that satisfied the inclusion and exclusion criteria (2 RCTs, 5 observational studies). The level of evidence ranged from moderate to very low quality and recommendations for use ranged from weak recommendations for use to weak recommendations against use. CONCLUSIONS: Due to the dearth of research available and the low levels of evidence in the published studies that were located we are unable to recommend the most effective conservative intervention for the treatment of coccydynia. Additional research is needed regarding the treatment for this painful condition.