As health systems seek to incentivize physicians to deliver high-value care, the relationship between physician compensation and health care delivery is an important knowledge gap. To examine physician compensation nationally and its relationship with care delivery, we examined 2012-2015 cross-sectional data on ambulatory primary care physician visits from the National Ambulatory Medical Care Survey. Among 175 762 office visits with 3826 primary care physicians, 15.4% of primary care physicians reported salary-based, 4.5% productivity-based, and 12.9% "mixed" compensation, while 61.4% were practice owners. After adjustment, delivery of out-of-visit/office care was more common for practice owners and "mixed" compensation primary care physicians, while there was little association between compensation type and rates of high- or low-value care delivery. Despite early health reform efforts, the overall landscape of physician compensation has remained strongly tethered to fee-for-service. The lack of consistent association between compensation and care delivery raises questions about the potential impact of payment reform on individual physicians' behavior.

Importance: Time pressure to provide a quick fix is commonly cited as a reason why opioids are frequently prescribed in the United States, but there is little evidence of an association between appointment timing and clinical decision-making. As the workday progresses and appointments run behind schedule, physicians may be more likely to prescribe opioids. Objective: To estimate whether characteristics of appointment timing are associated with clinical decision-making about pain treatment. Design, Setting, and Participants: This cross-sectional study of physician behavior used data from electronic health record systems in primary care offices in the United States to analyze primary care appointments occurring in 2017 for patients with a new painful condition who had not received an opioid prescription within the past year. Main Outcomes and Measures: The association between treatment decisions and 2 dimensions of appointment timing (order of appointment occurrence and delay relative to scheduled start time) were assessed. The rates of opioid prescribing were measured and compared with rates of nonopioid pain medication (ie, nonsteroidal anti-inflammatory drugs) prescribing and referral to physical therapy. All rates were estimated within the same physician using physician fixed effects, adjusting for patient, appointment, and seasonal characteristics. Results: Among 678 319 primary care appointments (642 262 patients; 392 422 [61.1%] women) with 5603 primary care physicians, the likelihood that an appointment resulted in an opioid prescription increased by 33% as the workday progressed (1st to 3rd appointment, 4.0% [95% CI, 3.9%-4.1%] vs 19th to 21st appointment, 5.3% [95% CI. 5.1%-5.6%]; P < .001) and by 17% as appointments ran behind schedule (0-9 minutes late, 4.4% [95% CI, 4.3%-4.6%] vs ≥60 minutes late, 5.2% [95% CI, 5.0%-5.4%]; P < .001). Prescribing of nonsteroidal anti-inflammatory drugs and referral to physical therapy did not display similar patterns. Conclusions and Relevance: These findings suggest that, even within an individual physician's schedule, clinical decision-making for opioid prescribing varies by the timing and lateness of appointments.

Importance: The traditional approach of diagnosis by individual physicians has a high rate of misdiagnosis. Pooling multiple physicians' diagnoses (collective intelligence) is a promising approach to reducing misdiagnoses, but its accuracy in clinical cases is unknown to date. Objective: To assess how the diagnostic accuracy of groups of physicians and trainees compares with the diagnostic accuracy of individual physicians. Design, Setting, and Participants: Cross-sectional study using data from the Human Diagnosis Project (Human Dx), a multicountry data set of ranked differential diagnoses by individual physicians, graduate trainees, and medical students (users) solving user-submitted, structured clinical cases. From May 7, 2014, to October 5, 2016, groups of 2 to 9 randomly selected physicians solved individual cases. Data analysis was performed from March 16, 2017, to July 30, 2018. Main Outcomes and Measures: The primary outcome was diagnostic accuracy, assessed as a correct diagnosis in the top 3 ranked diagnoses for an individual; for groups, the top 3 diagnoses were a collective differential generated using a weighted combination of user diagnoses with a variety of approaches. A version of the McNemar test was used to account for clustering across repeated solvers to compare diagnostic accuracy. Results: Of the 2069 users solving 1572 cases from the Human Dx data set, 1228 (59.4%) were residents or fellows, 431 (20.8%) were attending physicians, and 410 (19.8%) were medical students. Collective intelligence was associated with increasing diagnostic accuracy, from 62.5% (95% CI, 60.1%-64.9%) for individual physicians up to 85.6% (95% CI, 83.9%-87.4%) for groups of 9 (23.0% difference; 95% CI, 14.9%-31.2%; P < .001). The range of improvement varied by the specifications used for combining groups' diagnoses, but groups consistently outperformed individuals regardless of approach. Absolute improvement in accuracy from individuals to groups of 9 varied by presenting symptom from an increase of 17.3% (95% CI, 6.4%-28.2%; P = .002) for abdominal pain to 29.8% (95% CI, 3.7%-55.8%; P = .02) for fever. Groups from 2 users (77.7% accuracy; 95% CI, 70.1%-84.6%) to 9 users (85.5% accuracy; 95% CI, 75.1%-95.9%) outperformed individual specialists in their subspecialty (66.3% accuracy; 95% CI, 59.1%-73.5%; P < .001 vs groups of 2 and 9). Conclusions and Relevance: A collective intelligence approach was associated with higher diagnostic accuracy compared with individuals, including individual specialists whose expertise matched the case diagnosis, across a range of medical cases. Given the few proven strategies to address misdiagnosis, this technique merits further study in clinical settings.

Recent reductions in hospital readmission rates have been attributed to the Hospital Readmissions Reduction Program. However, admission rates also declined during the same period. We found that because the probability of an admission occurring soon after another is lower when there are fewer admissions per patient, the reduction in admission rates may explain much of the reduction in readmission rates.

BACKGROUND: Treatment by high-opioid prescribing physicians in the emergency department (ED) is associated with higher rates of long-term opioid use among Medicare beneficiaries. However, it is unclear if this result is true in other high-risk populations such as Veterans. OBJECTIVE: To estimate the effect of exposure to high-opioid prescribing physicians on long-term opioid use for opioid-naïve Veterans. DESIGN: Observational study using Veterans Health Administration (VA) encounter and prescription data. SETTING AND PARTICIPANTS: Veterans with an index ED visit at any VA facility in 2012 and without opioid prescriptions in the prior 6 months in the VA system ("opioid naïve"). MEASUREMENTS: We assigned patients to emergency physicians and categorized physicians into within-hospital quartiles based on their opioid prescribing rates. Our primary outcome was long-term opioid use, defined as 6 months of days supplied in the 12 months subsequent to the ED visit. We compared rates of long-term opioid use among patients treated by high versus low quartile prescribers, adjusting for patient demographic, clinical characteristics, and ED diagnoses. RESULTS: We identified 57,738 and 86,393 opioid-naïve Veterans managed by 362 and 440 low and high quartile prescribers, respectively. Patient characteristics were similar across groups. ED opioid prescribing rates varied more than threefold between the low and high quartile prescribers within hospitals (6.4% vs. 20.8%, p < 0.001). The frequency of long-term opioid use was higher among Veterans treated by high versus low quartile prescribers, though above the threshold for statistical significance (1.39% vs. 1.26%; adjusted OR 1.11, 95% CI 0.997-1.24, p = 0.056). In subgroup analyses, there were significant associations for patients with back pain (adjusted OR 1.25, 95% CI 1.01-1.55, p = 0.04) and for those with a history of depression (adjusted OR 1.28, 95% CI 1.08-1.51, p = 0.004). CONCLUSIONS: ED physician opioid prescribing varied by over 300% within facility, with a statistically non-significant increased rate of long-term use among opioid-naïve Veterans exposed to the highest intensity prescribers.

Few patients with opioid use disorder receive medication for addiction treatment. In 2017 the Comprehensive Addiction and Recovery Act enabled nurse practitioners (NPs) and physician assistants (PAs) to obtain federal waivers allowing them to prescribe buprenorphine, a key medication for opioid use disorder. The waiver expansion was intended to increase patients' access to opioid use treatment, which was particularly important for rural areas with few physicians. However, little is known about the adoption of these waivers by NPs or PAs in rural areas. Using federal data, we examined waiver adoption in rural areas and its association with scope-of-practice regulations, which set the extent to which NPs or PAs can prescribe medication. From 2016 to 2019 the number of waivered clinicians per 100,000 population in rural areas increased by 111 percent. NPs and PAs accounted for more than half of this increase and were the first waivered clinicians in 285 rural counties with 5.7 million residents. In rural areas, broad scope-of-practice regulations were associated with twice as many waivered NPs per 100,000 population as restricted scopes of practice were. The rapid growth in the numbers of NPs and PAs with buprenorphine waivers is a promising development in improving access to addiction treatment in rural areas.

BACKGROUND: In 2016, Medicare implemented Comprehensive Care for Joint Replacement (CJR), a national mandatory bundled-payment model for hip or knee replacement in randomly selected metropolitan statistical areas. Hospitals in such areas receive bonuses or pay penalties based on Medicare spending per hip- or knee-replacement episode (defined as the hospitalization plus 90 days after discharge). METHODS: We conducted difference-in-differences analyses using Medicare claims from 2015 through 2017, encompassing the first 2 years of bundled payments in the CJR program. We evaluated hip- or knee-replacement episodes in 75 metropolitan statistical areas randomly assigned to mandatory participation in the CJR program (bundled-payment metropolitan statistical areas, hereafter referred to as "treatment" areas) as compared with those in 121 control areas, before and after implementation of the CJR model. The primary outcomes were institutional spending per hip- or knee-replacement episode (i.e., Medicare payments to institutions, primarily to hospitals and post-acute care facilities), rates of postsurgical complications, and the percentage of "high-risk" patients (i.e., patients for whom there was an elevated risk of spending - a measure of patient selection). Analyses were adjusted for the hospital and characteristics of the patients and procedures. RESULTS: From 2015 through 2017, there were 280,161 hip- or knee-replacement procedures in 803 hospitals in treatment areas and 377,278 procedures in 962 hospitals in control areas. After the initiation of the CJR model, there were greater decreases in institutional spending per joint-replacement episode in treatment areas than in control areas (differential change [i.e., the between-group difference in the change from the period before the CJR model], -$812, or a -3.1% differential decrease relative to the treatment-group baseline; P<0.001). The differential reduction was driven largely by a 5.9% relative decrease in the percentage of episodes in which patients were discharged to post-acute care facilities. The CJR program did not have a significant differential effect on the composite rate of complications (P=0.67) or on the percentage of joint-replacement procedures performed in high-risk patients (P=0.81). CONCLUSIONS: In the first 2 years of the CJR program, there was a modest reduction in spending per hip- or knee-replacement episode, without an increase in rates of complications. (Funded by the Commonwealth Fund and the National Institute on Aging of the National Institutes of Health.).

Importance: In several pay-for-performance programs, Medicare ties payments to readmission rates but accounts only for a limited set of patient characteristics-and no measures of social risk-when assessing performance of health care providers (clinicians, practices, hospitals, or other organizations). Debate continues over whether accounting for social risk would mitigate inappropriate penalties or would establish lower standards of care for disadvantaged patients if they are served by lower-quality providers. Objectives: To assess changes in hospital performance on readmission rates after adjusting for additional clinical and social patient characteristics by using methods that distinguish the association between patient characteristics and readmission from between-hospital differences in quality. Design, Setting, and Participants: Using Medicare claims for admissions in 2013 through 2014 and linked US Census data, we assessed several clinical and social characteristics of patients that are not currently used for risk adjustment in the Hospital Readmission Reduction Program. We compared hospital readmission rates with and without adjustment for these additional characteristics, using only the average within-hospital associations between patient characteristics and readmission as the basis for adjustment, thereby appropriately excluding hospitals' distinct contributions to readmission from the adjustment. Main Outcomes and Measures: All-cause readmission within 30 days of discharge. Results: The study sample consisted of 1 169 014 index admissions among 1 003 664 unique Medicare beneficiaries (41.5% men; mean [SD] age, 79.9 [8.3] years) in 2215 hospitals. Compared with adjustment for patient characteristics currently implemented by Medicare, adjustment for the additional characteristics reduced overall variation in hospital readmission rates by 9.6%, changed rates upward or downward by 0.37 to 0.72 percentage points for the 10% of hospitals most affected by the additional adjustments (±30.3% to ±58.9% of the hospital-level standard deviation), and would be expected to reduce penalties (in relative terms) by 52%, 46%, and 41% for hospitals with the largest 1%, 5%, and 10% of penalty reductions, respectively. The additional adjustments reduced the mean difference in readmission rates between hospitals in the top and bottom quintiles of high-risk patients by 0.53 percentage points (95% CI, 0.50-0.55; P < .001), or 54% of the difference estimated with CMS adjustments alone. Both clinical and social characteristics contributed to these reductions, and these reductions were considerably greater for conditions targeted by the Hospital Readmission Reduction Program. Adjustment for social characteristics resulted in greater changes in rates of readmission or death than in rates of readmission alone. Conclusions and Relevance: Hospitals serving higher-risk patients may be penalized substantially because of the patients they serve rather than their quality of care. Adjusting solely for within-hospital associations may allow adjustment for additional patient characteristics to mitigate unintended consequences of pay for performance without holding hospitals to different standards because of the patients they serve.

BACKGROUND: Younger children in a school grade cohort may be more likely to receive a diagnosis of attention deficit-hyperactivity disorder (ADHD) than their older peers because of age-based variation in behavior that may be attributed to ADHD rather than to the younger age of the children. Most U.S. states have arbitrary age cutoffs for entry into public school. Therefore, within the same grade, children with birthdays close to the cutoff date can differ in age by nearly 1 year. METHODS: We used data from 2007 through 2015 from a large insurance database to compare the rate of ADHD diagnosis among children born in August with that among children born in September in states with and states without the requirement that children be 5 years old by September 1 for enrollment in kindergarten. ADHD diagnosis was determined on the basis of diagnosis codes from the International Classification of Diseases, 9th Revision. We also used prescription records to compare ADHD treatment between children born in August and children born in September in states with and states without the cutoff date of September 1. RESULTS: The study population included 407,846 children in all U.S. states who were born in the period from 2007 through 2009 and were followed through December 2015. The rate of claims-based ADHD diagnosis among children in states with a September 1 cutoff was 85.1 per 10,000 children (309 cases among 36,319 children; 95% confidence interval [CI], 75.6 to 94.2) among those born in August and 63.6 per 10,000 children (225 cases among 35,353 children; 95% CI, 55.4 to 71.9) among those born in September, an absolute difference of 21.5 per 10,000 children (95% CI, 8.8 to 34.0); the corresponding difference in states without the September 1 cutoff was 8.9 per 10,000 children (95% CI, -14.9 to 20.8). The rate of ADHD treatment was 52.9 per 10,000 children (192 of 36,319 children; 95% CI, 45.4 to 60.3) among those born in August and 40.4 per 10,000 children (143 of 35,353 children; 95% CI, 33.8 to 47.1) among those born in September, an absolute difference of 12.5 per 10,000 children (95% CI, 2.43 to 22.4). These differences were not observed for other month-to-month comparisons, nor were they observed in states with non-September cutoff dates for starting kindergarten. In addition, in states with a September 1 cutoff, no significant differences between August-born and September-born children were observed in rates of asthma, diabetes, or obesity. CONCLUSIONS: Rates of diagnosis and treatment of ADHD are higher among children born in August than among children born in September in states with a September 1 cutoff for kindergarten entry. (Funded by the National Institutes of Health.).

Antibiotic use is a primary driver of antibiotic resistance. However, antibiotic use can be distributed in different ways in a population, and the association between the distribution of use and antibiotic resistance has not been explored. Here, we tested the hypothesis that repeated use of antibiotics has a stronger association with population-wide antibiotic resistance than broadly-distributed, low-intensity use. First, we characterized the distribution of outpatient antibiotic use across US states, finding that antibiotic use is uneven and that repeated use of antibiotics makes up a minority of antibiotic use. Second, we compared antibiotic use with resistance for 72 pathogen-antibiotic combinations across states. Finally, having partitioned total use into extensive and intensive margins, we found that intense use had a weaker association with resistance than extensive use. If the use-resistance relationship is causal, these results suggest that reducing total use and selection intensity will require reducing broadly distributed, low-intensity use.

OBJECTIVES: 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. STUDY DESIGN: Observational study of Massachusetts residents aged 21 to 64 years with any HMO or PPO insurance coverage from 2010 to 2013. METHODS: 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. RESULTS: 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 (P <.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; P <.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; P <.001). CONCLUSIONS: 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.

Medicaid provides health insurance to more than seventy million Americans, yet there has been little systematic analysis of what factors influence enrollees' satisfaction with and access to care. Using a nationally representative survey of more than 270,000 Medicaid enrollees in 2014-15, we examined the consumer perspective on care in Medicaid. Average satisfaction ratings were 7.9 out of 10.0, but there were significant disparities across racial/ethnic groups. Satisfaction and access measures were generally similar among enrollees in managed care versus fee-for-service Medicaid. Access was significantly better in states where more physicians per capita accepted Medicaid patients: A one-standard-deviation increase in participating physicians per 100,000 population was associated with a 4.6-percentage-point increase in having a personal doctor. This was particularly true in fee-for-service Medicaid, and measures of physicians per capita were stronger predictors of patient experience than the simple percentage of doctors who accept Medicaid. Among those in Medicaid managed care, greater spending per enrollee was a significant predictor of satisfaction and access. Our findings emphasize that physician availability makes a difference in patients' experiences in Medicaid, and they indicate that racial/ethnic disparities in those experiences persist even among a uniformly insured population.