 In 2008, the US Congress enacted the Paul Wellstone and Pete Domenici Mental Health Parity and Addiction Equity Act (MHPAEA) requiring insurers to equalize private insurance coverage for mental health and substance use disorder services with coverage for general medical services.

To examine the effects of MHPAEA on substance use disorder treatment. Study Design: We used a difference-in-differences design to compare changes in outcomes among plan enrollees in the years before and after implementation of federal parity (2009-2010) with changes in outcomes among a comparison group of enrollees previously covered by state substance use disorder parity laws.

 Insurance claims data from Aetna Inc health plans in 10 states with state parity laws were used to compare outcomes for plan enrollees in fully insured and self-insured health plans (N = 298,339).

 In the first year of implementation, we found that federal parity did not lead to changes in the proportion of enrollees using substance use disorder treatment. We did find a modest increase in spending on substance use disorder treatment per enrollee ($9.99, 95% confidence interval, 2.54-18.21), but no significant change in identification, treatment initiation, or treatment engagement.

Inclusion of substance use disorder services in the federal parity law did not result in substantial increases in health plan spending. It will be critical to study results for year 2 after regulations affecting the management of care (eg, utilization review, network access) take effect.

Federal parity, as implemented in 2010, is unlikely to lead to large increases in total healthcare spending. 

		In 1 national health plan, federal parity led to a small increase in spending on substance use disorder treatment per enrollee ($9.99, 95% confidence interval, 2.54-18.21) in 2010, the first year after the policy took effect.

		This increase was not due to more individuals initiating treatment.

		Future work should consider the effects of the law in subsequent years after regulations affecting the management of care (eg, utilization review, network access) take effect

In 2008, the US Congress enacted a landmark federal parity law, the Paul Wellstone and Pete Domenici Mental Health Parity and Addiction Equity Act (MHPAEA), requiring health insurers to equalize private insurance coverage for mental health and substance use disorder (SUD) services with coverage for general medical services. Historically, health plans have imposed stringent limits on coverage for mental health and SUD services in the form of high cost sharing, annual inpatient day and outpatient visit limits, and lifetime dollar limits.

 In 2006, prior to implementation of MHPAEA, the majority of workers in private industry faced limits on SUD coverage. For example, 74% of workers had either annual or lifetime limits on the number of hospital days and 71% of workers had either annual or lifetime limits on the number of office visits for SUD.

 In contrast, almost all workers were offered unlimited hospital days and office visits for general medical services. In 2006, the median coinsurance level for SUD outpatient visits was 20% compared with 10% for outpatient medical services, and the median copayment for SUD outpatient visits was $25 compared with $17 for medical services.

Although a majority of states had previously enacted laws requiring parity for coverage of mental health disorders, many fewer states included SUDs in the conditions covered under their parity laws.

 MHPAEA took effect in 2010 and requires that when a private employer with 50 or more employees offers mental health and SUD coverage, all financial requirements (deductibles, copayments, coinsurance) and treatment limits (number of annual inpatient days and outpatient visits covered by insurance) for mental health and SUD must be equal to those for general medical services. The law also extends the Mental Health Parity Act of 1996, which prohibited the use of special annual and lifetime dollar limits for mental health benefits, to include SUD benefits. Health plans that provide out-of-network coverage for general medical services must also provide equal out-of-network coverage for mental health and SUD services.

To date, no published studies have examined the effects of MHPAEA. Most prior studies on the effects of state parity laws and of a comprehensive parity directive implemented in the Federal Employee Health Benefits Program focused on mental health and SUD treatment combined or on mental health treatment only. These studies consistently found that parity did not significantly increase use or spending,

 and that parity did lead to small but significant declines in out-of-pocket spending.

Only 2 prior studies examined the effects of parity on SUD treatment alone. Azzone and colleagues

 used health plan claims data to examine the effect of the Federal Employee Health Benefits Program parity policy on spending outcomes and performance measures based on Healthcare Effectiveness Data and Information Set (HEDIS); they found a significant decline in out-of-pocket SUD spending among treatment users and a small but significant increase in identification of SUD attributable to parity. A second study by Dave and Mukerjee

 examined the effects of state parity laws, including those limited to mental health, on inpatient admissions for SUD treatment and found that parity increased treatment admissions as well as the likelihood that an admission was privately insured.

The expected effects of the MPHAEA are unclear. Because MHPAEA does not require that specific conditions be covered, health plans have some discretion in determining which conditions and services to cover under parity. Relatively low public support for SUD benefits

 has led advocates to express concern that granting insurers discretion will lead to reduced coverage for SUD services. Conversely, SUD service users typically have higher spending compared with mental health service users. When limits are eliminated, a higher share of SUD patients are likely to be affected, which may lead to increases in spending.

Using administrative claims data from Aetna, a large national health insurance company, we study how MHPAEA affected an array of SUD use and spending outcomes, including the proportion of enrollees using any SUD treatment, total annual spending on SUD treatment per enrollee and per user of SUD treatment, total per user out-of-pocket spending on SUD treatment, and proportion of total SUD spending paid out of pocket. We also examine the proportion of health plan enrollees meeting 3 HEDIS-based SUD performance measures—identification, treatment initiation, and treatment engagement&mdash;which were originally developed by the Washington Circle.

In this study, we examine the effects of the law in the first year it took effect: 2010.

We used a difference-in-differences study design to compare changes in outcomes among health plan enrollees 1 year before and 1 year after implementation of MHPAEA (2009-2010) with changes among a comparison group of enrollees covered by state SUD parity laws in place prior to MHPAEA (N = 298,339). Difference-in-differences estimation allows us to distinguish the effects of the federal parity law from other changes affecting medical care utilization. Because MHPAEA extends to enrollees in all privately insured firms with 50 or more employees in the United States, identifying a comparison group presents a challenge. In this study, health plan enrollees already subject to preexisting state SUD parity laws served as the comparison group. As of 2008, 10 states—Colorado, Connecticut, Kentucky, Maine, Maryland, Minnesota, New Hampshire, Oregon, Rhode Island, and Vermont&mdash;had passed parity laws that included SUD services. We therefore limited our study population to Aetna health plan enrollees in these 10 states. Under the Employee Retirement Income Security Act, enrollees in fully insured employer-sponsored health plans were subject to state parity laws, but enrollees in self-insured employer-sponsored health plans operating in these states were exempt from state parity laws. Thus, in these states only enrollees of self-insured plans were newly covered under federal parity. Prior work suggests effects of state parity laws were concentrated on individuals in fully insured plans, and these laws did not affect individuals in self-insured plans.

 Self-insured plan enrollees in these 10 states who were subject to parity for the first time served as our treatment group, whereas fully insured plan enrollees who were already subject to SUD parity at the state level served as our comparison group.

We used de-identified Aetna administrative claims data to conduct this study. These data are from a single insurer, but include employees of many different firms with diverse benefit designs and clearly identifiable information on whether an enrollee is covered by a fully insured or self-insured plan. Importantly, Aetna does not carve out mental health or SUD benefits. We examined enrollee claims 1 year before (2009) and 1 year after (2010) MHPAEA implementation. We included individuals continuously enrolled in Aetna health plans over the full 2-year study period to eliminate variation over time in the underlying population studied. All individuals aged 18 to 62 years in calendar year 2009 were included. We excluded older individuals, who are typically covered by Medicare. We also excluded 135 individuals who switched between fully insured and self-insured plans during the 2-year period. Because most of the state parity laws considered (as well as MHPAEA) do not apply to firms with fewer than 50 employees, individuals employed by these small firms were excluded. Analyses included 162,761 enrollees in self-insured plans and 135,578 enrollees in fully insured plans.

Claims were identified as being SUD related if they met either of 2 criteria. First, we identified enrollees treated for SUD using the first 2 

International Classification of Diseases, 9th Revision, Clinical Modification

 diagnostic codes on each inpatient and outpatient claim using codes for alcohol-or drug-induced mental disorders (291 and 292), alcohol or drug dependence (303 and 304), and nondependent abuse of drugs (excluding tobacco) (305.0 and 305.2-305.9). In addition, any services received in an SUD treatment facility or from a drug or alcohol counselor were considered SUD treatment. We used National Drug Codes to identify prescription medications used specifically to treat SUD (ie, acamprosate, buprenorphine, buprenorphine/naloxone, disulfiram, and naltrexone [oral and sustained release]). Because methadone cannot be prescribed by physicians for the treatment of opioid dependence, we did not include methadone in our analysis. Outcomes in a calendar year were (1) proportion of enrollees using any SUD treatment, (2) total spending on SUD treatment per enrollee, (3) total spending on SUD treatment per user, (4) out-of-pocket spending on SUD treatment per user, and (5) proportion of total SUD spending paid for out of pocket. To calculate an enrollee’s annual total spending, we included all SUD-related inpatient, partial hospitalization, intensive outpatient, and outpatient services, and prescription medications to treat SUD. To calculate an enrollee’s annual out-of-pocket spending, we included the deductible, copayment, and coinsurance for SUD services and prescription medications.

We examined 3 HEDIS-based SUD performance measures. We measured identification as the share of all health plan enrollees who had a new SUD claim within a calendar year. New treatment episodes were those with no SUD treatment during the prior 60 days. We measured treatment initiation as the share of enrollees with a new episode of SUD treatment who initiated treatment within 14 days of their initial diagnosis. Following HEDIS, all patients for whom identification of SUD occurred through a hospital admission were considered to have initiated treatment, but inpatient detoxification services were not considered treatment initiation. We measured treatment engagement as the share of enrollees with a new episode of SUD treatment who received at least 2 SUD services within 30 days of their initial diagnosis. For the treatment engagement measure, multiple services could not occur on the same day. To ensure that we were identifying only new episodes, we did not consider episodes that began during the first 60 days of the calendar year. For both the initiation and engagement measures, we omitted episodes that did not allow for a 30-day follow-up (ie, those that occurred late in the year).

Our explanatory variables were indicators for whether an observation occurred after federal parity implementation (ie, in 2010) and whether the individual was enrolled in a plan newly subject to parity (ie, a self-insured firm). We also controlled for enrollee sex, age (ie, age 18-31, 32-46, 47-62 years in 2009), and state.

We estimated the effect of federal parity using a difference-in-differences model. For binary outcomes we used logistic regression. For spending outcomes we used a 2-part model to estimate the probability of any SUD use and then estimated spending conditional on any use using a generalized linear model with a log link and gamma distribution, as indicated by the results of a modified Park test.

 To estimate the relationship between parity and share of total spending paid out of pocket, we estimated a fractional logit model, which was implemented as a generalized linear model with a logit link and binomial distribution.

 To facilitate interpretation, we transformed relevant coefficients to the original scale of the outcome using the method of recycled predictions. We calculated confidence intervals using a nonparametric block bootstrap method that accounts for repeat observations for individuals.

 This study was exempted from review by Yale University Institutional Review Board.

We compared characteristics of the self-insured treatment group and fully insured comparison group enrollees in 2009 (

). Self-insured enrollees were significantly more likely to be female and younger, although these differences are not large enough to be clinically meaningful. Although differences were small in absolute terms, self-insured enrollees were 57% more likely than fully insured enrollees to have an SUD diagnosis (1.1 % vs 0.7 %).

 reports difference-in-differences estimates for the probability of use of SUD treatment and total spending on SUD treatment per enrollee. After accounting for secular trends in the use of SUD treatment, we found no significant difference in the probability of using SUD treatment attributable to MHPAEA. We did find a significant increase of $9.99 (95% confidence interval [CI], $2.54-$18.21) in total spending on SUD treatment per enrollee attributable to MHPAEA, compared with a base rate of $36.51 in the self-insured group. We found no significant difference in total spending on SUD treatment per user, although the point estimate was relatively large ($608). 

 indicates that we detected no effect of MHPAEA on out-of-pocket spending on SUD treatment or proportion of spending paid out of pocket among users.

 summarizes the effects of MHPAEA on the HEDIS- based identification, treatment initiation, and treatment engagement performance measures. We found no significant effect on identification of SUD, treatment initiation, or treatment engagement associated with implementation of MHPAEA.

This study is the first to examine the effects of MHPAEA on SUD treatment use and spending. The new law led to a significant increase in SUD spending of $9.99 per health plan enrollee. The average cost of an employer-sponsored individual health insurance policy in 2010 was $5049,

 suggesting this is a negligible increase. During the congressional debate over passage of MHPAEA, employers and health plans raised the concern that the law would greatly increase healthcare spending.

 This study suggests that, at least related to SUD treatment, this concern was unfounded. Once a controversial and much debated issue, the inclusion of SUD services in parity appears unlikely to affect either health plan profitability or overall rates of insurance coverage.

That we found no change in out-of-pocket SUD spending suggests federal parity did not lead to increases in financial protection for SUD treatments. This result is surprising given data indicating that higher cost sharing for SUD services was common prior to MHPAEA.

Under parity, 2 potentially offsetting changes to out-of-pocket spending might be expected. First, equalization of cost sharing between general and medical care services might lead to lower cost sharing for SUD services and thus reduce out-of-pocket spending. Second, increases in the number or intensity of services due to reduced out-of-pocket price or elimination of treatment limits might lead to increases in total costs and total out-of-pocket spending. To shed light on the relative importance of these competing effects, we examined the proportion of total costs paid out of pocket. If cost sharing were declining, we would expect the proportion of total costs paid out of pocket to decline as well. That we found no change in the proportion of total costs paid out of pocket suggests that the elimination of treatment limits may be an important cause of increased spending. This would be the effect expected if increases in treatment expenditures result from the elimination of day or visit limits rather than reductions in cost sharing.

We also note that our data included claims for SUD treatments paid for by private insurance. If MHPAEA resulted in a switch from SUD services self-financed by families to services financed by insurance, out-of-pocket spending reported in claims might have increased, but actual out-of-pocket spending by SUD treatment users might have declined. We found no significant increase in SUD identification, treatment initiation, or treatment engagement after parity implementation. We note that these measures only captured whether individuals were more likely to begin treatment and did not measure the duration or the intensity of treatment. Although not definitive, that we found increases in SUD total expenditures of 27% (an increase of $10 with an initial expenditure of $37 per enrollee for the self-insured group, as indicated in Table 2) without similar increases in identification suggests there were increases in treatment duration or treatment intensity after MHPAEA implementation. The point estimate of total spending per SUD treatment user suggests an increase in yearly total spending of approximately $600 per year, although this finding is not statistically significantly different from zero due to imprecision in the estimate caused by the relatively small number of SUD treatment users in our sample. In these data the payment for a physician visit for SUD treatment of low to moderate intensity was approximately $70, suggesting this increased spending may yield 8 additional visits for SUD treatment. Treatment increases may be an important benefit of parity given the level of unmet need associated with SUD and evidence that treatment provides important benefits and reduces social costs.

Understanding the effects of MHPAEA on SUD treatment is critical because provisions of the Patient Protection and Affordable Care Act will extend the law to the individual and small group insurance markets under state and federal health insurance exchanges beginning in 2014. Substance use disorder services are required as part of the essential health benefit under the Affordable Care Act, although states may differ on the scope of SUD benefits covered.

Strengths of this study include the use of a comparison group to control for secular trends in SUD treatment use, geographic heterogeneity in the study population, and detailed SUD treatment use and spending data. Like all studies, our research is limited by the type of data considered. These data did not include patient interviews or medical chart reviews to confirm information regarding diagnosis or services received, or information on treatments not financed by insurance. There is little reason to expect that these data limitations differentially affected the intervention and comparison groups, suggesting our estimates of changes in use and spending were unbiased. Yet if providers systematically changed recorded diagnoses in response to parity (due to increased coverage for SUD treatments), our results would be biased. In addition, we found little difference in reimbursement levels between the self-insured and fully insured groups.

Another limitation is that MHPAEA may lead to multiple insurance market changes, including declines in the out-ofpocket price of services, increases in supply-side constraints imposed by insurers (ie, prior authorization, referral restrictions), and reduced stigma associated with SUD treatment, which may all affect use and spending. Although our study design allowed us to determine the net effect of parity, we were not able to disentangle these competing mechanisms.

A third consideration is that preexisting state parity laws were not identical to MHPAEA; therefore, fully insured enrollees in our comparison group might have experienced some change in benefits as they moved from being subject to the less comprehensive state parity laws to the more comprehensive MHPAEA in 2010. A fourth limitation is that we did not consider changes in costs for treatment of substance abuse—related medical conditions (eg, alcoholic cirrhosis, hepatitis). A fifth limitation relates to the generalizability of our findings. We evaluated the effects of parity on individuals insured by a single health insurer in 10 states with preexisting state SUD parity laws. Thus our results may not be generalizable to other insurance or population contexts.

Finally, this study examined only the first year after MHPAEA took effect. The interim final regulations of MHPAEA, which were released in February 2010 and took effect for most plans in 2011, prohibited plans from using so-called nonquantitative treatment limits for mental health and SUD benefits unless these limits were comparable to those used for general medical services.

 Nonquantitative treatment limits include medical management standards, prior authorization, utilization review, prescription drug formulary design, standards for provider admission to participate in a network, and provider reimbursement. It is possible that these regulations could lead to different effects of the law; therefore, it is critical for future research to examine use and spending in response to MHPAEA in subsequent years.

From Perelman School of Medicine, University of Pennsylvania (AJE, MOH), Philadelphia, PA; Yale School of Public Health (SHB, DAF), New Haven, CT; Yale School of Medicine (DAF), New Haven, CT; Aetna (HU, DL), Hartford, CT; John Hopkins Bloomberg School of Public Health (CLB), Baltimore, MD.

This study was funded by the National Institute on Drug Abuse (grant DA 026414).

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

Concept and design (SHB, AJE, DAF, CLB); acquisition of data (SHB, DL, CLB); analysis and interpretation of data (SHB, AJE, MOH, DAF, HU, DL, CLB); drafting of the manuscript (SHB, CLB); critical revision of the manuscript for important intellectual content (SHB, AJE, MOH, DAF, HU, DL, CLB); statistical analysis (SHB, AJE, MOH, CLB); obtaining funding (SHB, DAF, CLB), administrative, technical, or logistic support (SHB, AJE, MOH, DL, CLB); and supervision (SHB, CLB).

 Susan H. Busch, PhD, Yale School of Public Health, 60 College St, New Haven CT 06520-8034.1. Barry C, Sindelar JL. Equity in private insurance coverage for substance abuse: a perspective on parity. Health Aff (Millwood). 2007; 26(6):w706-w716.

2. Gabel JR, Whitmore H, Pickreign JD, Levitt KR, Coffey RM, Vandivort- Warren R. Substance abuse benefits: still limited after all these years. Health Aff (Millwood). 2007;26(4):w474-w482.

3. National Alliance on Mental Illness. State mental health parity laws. http://www.nami.org/Content/ContentGroups/Policy/Issues_Spotlights/ Parity1/State_Parity_Chart_0709.pdf. Revised July 2009. Accessed September 17, 2012.

4. Goldman HH, Frank RG, Burnam MA, et al. Behavioral health insurance parity for federal employees. N Engl J Med. 2006;354(13): 1378-1386.

5. Barry CL, Busch SH. Do state parity laws reduce the financial burden on families of children with mental health care needs? Health Serv Res. 2007;42(3, pt 1):1061-1084.

6. Pacula RL, Sturm R. Mental health parity legislation: much ado about nothing? Health Serv Res. 2000;35(1, pt 2): 263-275.

Articles

POLICY

Federal parity led to an increase in spending on substance use disorder treatment.

The Effects of Federal Parity on Substance Use Disorder Treatment

To investigate the source of the weak correlation across geographic areas between Medicare and private insurance spending.

 We obtained Medicare spending data at the hospital referral region (HRR) level for 2007 from the Dartmouth Atlas, and commercial claims from large employers for 2007 from the Truven MarketScan Database. We constructed county-level data on hospital market structure from Medicare patient flows and obtained county-level data on the Medicare wage index from the Centers for Medicare & Medicaid Services website. We aggregated these sources to the HRR level. We decomposed Medicare and private spending into 2 components: price and volume. We also decomposed Medicare and private prices into 2 components: a common measure of cost and a sector-specific markup. We computed correlations between Medicare and private prices and volumes, and the correlation of each sector’s price and volume with cost and markup.

 We found that Medicare and private prices are strongly positively correlated, largely because both are keyed off of common costs. Consistent with previous work, we found that Medicare and private volumes are strongly positively correlated as well.

The weak correlation between Medicare and private spending is consistent with these 2 empirical regularities. It is mathematically due to negative correlations between each sector’s price and the other sector’s volume. In particular, we found that private prices have important spillover effects on Medicare volume. Future research on the effects of competition should take account of this phenomenon.

This retrospective, descriptive analysis investigated the source of the weak correlation across geographic areas between Medicare and private insurance spending.

		Medicare and private prices are strongly positively correlated, largely because both are keyed off of common costs. Medicare and private volumes also are strongly positively correlated.

		The weak correlation between Medicare and private spending is consistent with these 2 empirical regularities. It is mathematically due to negative correlations between each sector’s price and the other sector’s volume.

		Private prices have important spillover effects on Medicare volume. Future research on the effects of competition should consider this phenomenon.

Area variation in Medicare fee-for-service spending has been extensively documented. However, the extent to which area variation in spending by private insurers mirrors the variation in Medicare is an open empirical issue. Numerous studies have found a significant positive correlation between Medicare and the privately insured in terms of service use, such as hospital admissions, days per admission, or days in the intensive care unit at the end of life.

 In contrast, the correlation between Medicare and private spending is weakly positive or even negative, depending on the type of service and the specifics of the populations in question.

Understanding why Medicare and private spending are so weakly correlated is important. One hypothesis for the weak correlation in pending is that the prices paid by Medicare are weakly (or even inversely) correlated with the prices paid by private insurers. If true, this could be evidence that Medicare’s prices fail to reflect costs, perhaps due to political manipulation of reimbursement rates, or that private prices fail to reflect costs, perhaps due to provider market power. These explanations have very different implications for Medicare policy.

Yet no study has tested this hypothesis directly. More generally, no study has offered a unified framework in which to understand the relationships between the components of Medicare and private insurer spending. This study sought to fill this gap. We decomposed Medicare and private insurance spending into 2 components: a price index and price-adjusted spending, which we used as a measure of quantity. We calculated the simple correlations among the components of spending and the correlations between the components, the Medicare wage index, and price-cost markups.

These correlations rejected the hypothesis that Medicare’s prices are weakly or inversely correlated with those of private insurance. Rather, they are strongly positively correlated, largely reflecting differences in costs (as measured by the wage index) across areas. We provide evidence for an alternative explanation for the strong positive correlation between Medicare and private utilization and a weak correlation between Medicare and private spending.

We examined how the components of inpatient spending among fee-for-service Medicare beneficiaries are related to the components of spending among a nonelderly, privately insured population. We wrote inpatient spending per enrollee in either Medicare or private coverage from hospital referral region (HRR) i as the product of a price index for HRR i and quantity (price-adjusted spending) per enrollee in i:

	Spending per enrollee in HRR = price index

Strictly speaking, geographic price indices for enrollees in fee-for-service Medicare and private insurance plans measure different things. Medicare reimbursement rates are determined by a complex nationwide administered price system, so geographic variation in Medicare “prices” necessarily arises out of geographic variation in whatever factors the Medicare formula considers. In contrast, commercial rates depend on numerous unrelated negotiations between insurers and providers, so geographic variation in commercial prices depends on a much less well-specified, broader range of influences.

We adopted the formula used in the Dartmouth Atlas for our Medicare price index.

 According to the formula, the Medicare price level in an HRR is equal to the ratio of total spending in the HRR divided by base diagnosis-related group (DRG) payments plus adjusted outlier payments. Thus, the formula captures variation across areas in the wage index and in medical education and disproportionate-share (graduate medical education, indirect medical education, and disproportionate share hospital) payments, given each HRR’s DRG mix and outlier payments. Our private insurance price index is equal to total spending in an HRR divided by what spending would have been if all hospitals in the HRR charged the national average set of DRG prices. This index captures variation across areas in the prices that insurers pay, given each HRR’s DRG mix.

We also decomposed the price indices into 2 factors: the part that is due to differences in costs that are common to Medicare and the private sector, and a markup, which we allowed to vary across sectors. We used the Medicare wage index to represent common costs. We defined the markup as the residual from a regression of the log of the price index on the log of the Medicare wage index. Thus, for either the Medicare or private insurer price index in HRR i:

	Price index in HRR = f(Medicare wage index in HRR)

where f(.) is defined as exp[a + b*ln(wage index)], and a and b are coefficients from the regression described above. We provide a formal presentation of our decomposition in the 

We used data from several sources. First, we used HRR-level data from the Dartmouth Atlas website for 2007. These data included total Medicare inpatient spending per admission on fee-for-service beneficiaries, total price-adjusted inpatient spending per admission, the number of elderly beneficiaries, and the number of admissions per beneficiary. All of these variables are adjusted by Dartmouth for differences in the age, sex, and racial composition of areas. From this information, we constructed the price index described above. These data cover a total of 27,978,661 individuals in all 306 HRRs.

Second, we used Truven MarketScan Commercial Claims and Encounters data on the nonelderly privately insured from 2007. We limited our sample to individuals with preferred provider organization insurance and imposed the additional selection criteria suggested by Baker and colleagues.

 These data include information on the enrollees’ county of residence, which we mapped to HRRs according to the method proposed by Chernew and coauthors.

 We adjusted the variables for differences in age and sex across HRRs (we could not adjust for differences in race because the MarketScan data do not include race). These data cover a total of 14,902,153 beneficiaries in all 306 HRRs.

Third, we used data on the Medicare wage index by county from the Centers for Medicare & Medicaid Services website, which we mapped to HRRs using the method described above.

Fourth, we used data on hospital market structure that we constructed from individual-level 2007 data on Medicare patient flows according to the method proposed by Kessler and McClellan

; we provide details in the Appendix. We counted hospitals in a common system as commonly owned. 

 presents the enrollee-weighted and unweighted means and standard deviations of the variables that we analyzed. These descriptive statistics reflect the well-known properties of Medicare and private insurance area variation. There is considerable variation across areas in Medicare spending, even after adjusting for differences in prices.

 Spending per enrollee in the MarketScan data is much lower, reflecting the much lower hospital admissions rate in the nonelderly population. The enrollee-weighted means of the Medicare and MarketScan price indices are both 1 by construction.

Total spending per enrollee in MarketScan is slightly less variable across areas than total spending per enrollee in Medicare; the ratio of the standard deviation to the mean is 0.159, compared with 0.181 in Medicare. But the portion of this variation due to variation in prices is much greater in MarketScan than Medicare; the standard deviation of the MarketScan price index is 0.179 compared with 0.129 in Medicare, which is consistent with previous work.

 This is reflected in the fact that the standard deviation across areas of the price markup over the wage index is 3 times greater in MarketScan.

presents the correlation between Medicare and MarketScan total spending, quantities, and prices (

 values for the null hypothesis of zero correlation are in parentheses). Total spending is weakly positively correlated (r = 0.1135, 

 = .0473), but quantity, as measured by price-adjusted spending, is very strongly positively correlated (r = 0.6467, 

 <.0001). This finding is consistent with the previous literature, which finds strong positive correlations between the rates of use of particular services in the  elderly and nonelderly populations. It is also consistent with the hypothesis that physicians develop a single practice style that they use for patients of different ages and insurance statuses. Medicare and MarketScan prices are also strongly positively  correlated (r = 0.4476, 

 <.0001). This rejection of the hypothesis that the prices paid by Medicare are weakly (or even inversely) correlated with the prices paid by private insurers suggests that another factor is responsible for the weak correlation between  Medicare and MarketScan total spending.

Instead, according to Table 2, the weak correlation between sectors’ total spending is due to the negative correlation between each sector’s price and the other’s quantity. In other words, Medicare volume is low where private prices are high (r = —0.4324), and MarketScan volume is low where Medicare prices are high (r = –0.3748). Mathematically, this is what is causing the weak  correlation between Medicare and MarketScan total spending. (If Medicare spending is the product of prices a and quantities b, and  MarketScan spending is the product of prices c and quantities d, then Cov(ab, cd) ≈ w

) This is consistent with a wide volume of literature that finds that price reductions in Medicare induce increases in private volume.

 investigates this result further. It presents the correlations of the sectors’ prices and quantities with the Medicare wage index and each sector’s markup. According to the first column, the wage index is strongly positively correlated with both sectors’ prices, although more strongly with Medicare prices (r = 0.9001) than with private prices (r = 0.5262). It is also negatively correlated with both Medicare(r = —0.2548) and MarketScan (r = –0.4123) quantities. For example, the 3 HRRs with the lowest price-adjusted Medicare spending (Bend, Oregon, at $2644; San Luis Obispo, California, at $2677; and Santa Barbara, California, at $2763) all have relatively high wage indices (1.04, 1.12, and 1.12, respectively); by comparison, the 3 HRRs with the highest adjusted   Medicare spending (Alexandria, Louisiana, at $5666; McAllen, Texas, at $5836; and Monroe, Louisiana, at $5925) all have relatively low wage indices (0.78, 0.87, and 0.79, respectively). If we assume the wage index measures costs common to both sectors, when the wage index falls, both Medicare and private patients become absolutely more profitable, but Medicare patients become relatively less profitable than private patients, because the correlation between the wage index and the Medicare price is higher. This explains why a unit decrease in the wage index leads to increases in both Medicare and private quantities, with a greater increase in the private quantity.

According to the second column, the MarketScan markup is strongly positively correlated with the MarketScan price (r = 0.8739), and negatively correlated with both the Medicare (r = —0.3576) and the MarketScan (r = –0.2958) quantities. When the MarketScan markup rises, private quantity falls, and Medicare quantity falls even more. The 

 translates the negative correlation between MarketScan markup and Medicare quantity into dollar terms. The HRRs with the lowest MarketScan markups have average price-adjusted Medicare spending of $4426, whereas those with the highest markups have Medicare spending of $3808—a difference of almost 1 standard deviation. This is consistent with a model in which private demand slopes downward and providers respond to the declining profitability of private patients by increasing the services they deliver to Medicare patients.

According to the third column of Table 3, the Medicare markup is strongly positively correlated with the Medicare price (r = 0.4431, 

 <.0001), but roughly uncorrelated with the MarketScan price and the Medicare or MarketScan quantity. This is likely due to the minimal amount of variation in the Medicare markup; we discuss alternative explanations for this finding in the Discussion section.

Area variation in Medicare spending has long been viewed as evidence of inefficiency in the program.

 At the same time, health services researchers have observed that area variation in private insurance spending is only weakly positively (or even negatively) correlated with Medicare variations. Given that physicians likely have a single practice style that they use for both elderly and nonelderly patients, this presents a puzzle: if utilization is correlated across sectors, why isn’t spending?

In this study, we explain the source of this apparent anomaly. We decomposed Medicare and private insurance spending into 2 components: a price index and price-adjusted spending, which we use as a measure of quantity or volume. We showed that the weak correlation in overall spending is not due to weak or inverse correlation in Medicare and private prices. Not surprisingly, Medicare and private prices are strongly positively correlated across areas, largely because both are keyed off common costs (as measured by the Medicare wage index). Instead, the weak correlation is due to the negative correlation between each sector’s price index and the other’s volume.

We documented 2 channels through which this negative cross-correlation might occur. First, increases in common costs are associated with increases in both Medicare and private insurance prices, and increases in private prices are associated with decreases in private volumes. Second, increases in private insurance markups are associated with decreases in Medicare volumes. Previous research on the sources of Medicare variation has focused on beneficiary health, socioeconomic status, and preferences. Our results point to another possibility: prices in the private, under-age-65 insurance market. In analyses not reported in the tables, we showed that hospital market concentration is a likely source of this second effect: the correlation between private markups and the hospital HHI is 0.3797 (

To our knowledge, ours is the first study to identify this effect. The literature on the spillover effects of prices on utilization focuses on the impact of changes in public prices on private quantities rather than the impact of private prices on public quantities.

 We provide empirical evidence for the hypothesis proposed by Chernew and colleagues: that providers respond to the declining profitability of private patients by reducing the share of time and resources they devote to them compared with public patients.

Our analysis has significant limitations. It does not identify the causal relationships that are at the root of the negative association between one sector’s prices and the other’s volume. For example, we cannot say whether high private prices themselves cause low Medicare volumes, or whether some underlying third factor (like common costs or hospital market concentration) causes both. For purposes of policy, distinguishing between these alternatives is important. To do this would require specification of a formal model of the process through which prices and volumes are determined, which was beyond the scope of this study. In addition, our analysis was limited to inpatient acute care hospital spending. Results for outpatient services, prescription drugs, and postacute care may differ, thus complicating the explanations that we offer for the weak correlation in overall spending. Finally, our study design had only minimal independent variation in public prices. Therefore, the fact that we did not find significant spillovers from public prices to private volumes cannot be interpreted as evidence of the absence of such an effect; understanding the relative importance of private-to-public and public-to-private spillovers is an important topic for future work.

Nonetheless, our work shows that the weak correlation between spending in the 2 sectors does not, by itself, imply anything about  the processes by which prices are determined, or about the relative efficiency in one sector versus the other. However, our finding that hospital market concentration is strongly positively correlated with private payer markups supports the concern voiced by other investigators that private sector purchasers are more vulnerable to provider market power.

Future research on the policy  implications of area variations should take more careful account of this issue.  

 Frome Stanford University (LCB, KB, DPK), Stanford, CA.

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

 Concept and design (LCB, KB, DPK); acquisition of data (LCB, KB, DPK); analysis and interpretation of data (LCB, KB, DPK); drafting of the manuscript (LCB, KB, DPK); critical revision of the manuscript for important intellectual content (LCB, KB, DPK); statistical analysis (LCB, KB, DPK); obtaining funding (LCB, KB), administrative, technical, or logistic support (LCB, KB, DPK); and supervision (LCB, KB, DPK).

 Dr Daniel P. Kessler, Stanford University Law, GSB, and Hoover Institution, 434 Galvez Mall, Stanford, CA 94305. E-mail: fkessler@stanford.edu.1. Chernew ME, Sabik LM, Chandra A, Gibson TB, Newhouse JP. Geographic correlation between large-firm commercial spending and Medicare spending. Am J Manag Care. 2010;16(2):131-138.

2. Baker LC, Fisher ES, Wennberg JE. Variations in hospital resource use for Medicare and privately insured populations in California. Health Aff (Millwood). 2008;27(2):w123-w134.

3. Stiefel M, Feigenbaum P, Fisher ES. The Dartmouth Atlas applied to Kaiser Permanente: analysis of variation in care at the end of life. Perm J. 2008;12(1):4-9.

WEB EXCLUSIVE

The weak correlation between Medicare and commercial insurance spending is due to negative correlations between each sector's price and the other sector's volume.

Why Are Medicare and Commercial Insurance Spending Weakly Correlated?

To explore potential weight-related disparities in the quality of care for adults with diabetes in a large managed care health plan according to recommended quality indicators.

 A total of 164,721 Kaiser Permanente Southern California members aged 18 to 75 years with diabetes who had 1 or more health encounters with a weight and height measurement between July 1, 2007, and June 30, 2008, were identified. The 9 Healthcare Effectiveness Data and Information Set (HEDIS) quality-of-care measures were examined: 4 screening measures (glycated hemoglobin [A1C], retinal examination, lipids, and nephropathy) and 5 control measures (2 for A1C, 2 for blood  pressure, and 1 for low-density lipoprotein cholesterol [LDLC]), using data extracted from electronic health records (EHRs). Weight and height from the EHR were used to calculate body mass index (BMI). Adjusted odds ratios and 95% confidence intervals were calculated to examine the association between BMI categories and HEDIS quality-of-care measures.

Among individuals with diabetes, 12% were in the healthy-weight category, 30% were overweight, and 28%, 17%, and 14% were in obese categories I, II, and III, respectively. Overweight and obese individuals were more likely than their healthy-weight counterparts to have screening measures performed. However, among those tested or screened, overweight and obese individuals were less likely to have their A1C and blood pressure controlled. LDL-C control increased as BMI increased.

 These findings highlight the need for interventions to improve glycemic and blood pressure control among overweight and obese patients with diabetes.

This study highlights the need for increased attention by patients, providers, and healthcare systems to improving blood pressure and glycemic control among overweight and obese patients with diabetes.

		Overweight and obese individuals were more likely than their healthy-weight counterparts to have glycated hemoglobin (A1C) testing, retinal examinations, and lipid and nephropathy screenings performed.

		Among those tested or screened, overweight and obese individuals were less likely to have their A1C and blood pressure controlled; low-density lipoprotein cholesterol control increased as body mass index increased.

The prevalence of obesity in the United States has risen dramatically in the past 50 years and continues to rise. Currently, more than 30% of the American population is considered obese.

 Obese individuals have more comorbidities and a higher rate of healthcare utilization than non-obese individuals.

 There are well-documented associations between obesity and several chronic  conditions, including diabetes.

Diabetes is associated with high rates of cardiovascular disease, retinopathy, peripheral vascular disease, neuropathy, and nephropathy.

Many of these complications can be delayed or prevented with effective management; however, many people with diabetes are not meeting the American Diabetes Association (ADA) goals for risk factor control.

An analysis of the National Health and Nutrition Examination Survey 2001-2002 found that the percentage of adults in the United States with diabetes who were not at goal was 50% for glycated hemoglobin (A1C), 53% for blood pressure (BP), 65% for low-density lipoprotein cholesterol (LDL-C), and 49% for triglycerides.

Obese patients have reported feeling stigmatized by healthcare professionals

; those who feel stigmatized may be more likely to  delay or avoid seeking healthcare services.

 In addition, a number of studies have documented negative healthcare provider attitudes toward obese patients,

perhaps making it less likely that healthcare professionals will work with their obese patients on behavioral changes. The extent of the impact of stigmatization and bias on quality of care in obese patients is not well   understood, especially for those patients with chronic health conditions such as diabetes. Most of the available research relies on self-report rather than using medical record data to assess actual utilization.

The aim of this study was to examine whether there were weight-related disparities in risk factor screening and control rates based  upon the National Committee for Quality Assurance (NCQA) Healthcare Effectiveness Data and Information Set (HEDIS) performance measures for diabetes among members of a large, integrated managed healthcare organization. The HEDIS performance measures are widely used by health plans in the United States to measure performance on quality of care and services across a range of health conditions including diabetes.

This study was conducted within Kaiser Permanente Southern California (KPSC), an integrated healthcare system that provides comprehensive health services for approximately 3.5 million residents of southern California. Members of KPSC are socioeconomically diverse and broadly representative of the general population of southern California.

We identified all KPSC members aged 18 to 75 years continuously enrolled in the health plan between July 1, 2007, and June 30, 2008, with no more than one 45-day gap in enrollment and with a diagnosis of diabetes mellitus (N = 169,077). A diabetes diagnosis was based on the HEDIS specifications of at least 2 outpatient or 1 inpatient

 International Classification of Diseases, Ninth Revision, Clinical Modification

) diagnosis codes of 250, 357.2, 362.0, 366.41, or 648.0, or having been dispensed insulin or an oral anti-hyperglycemic medication (excluding metformin alone) during the measurement period. We then excluded 3901 members (2.3%) who did not have at least 1 health encounter with a recorded weight and height measurement during the study period and 455 with a body mass index (BMI) of less than 18.5 kg/m2 (0.3%), resulting in a final study sample of 164,721 individuals. Information from KPSC electronic health records (EHRs) including encounter, pharmacy, and laboratory data were linked with administrative membership information for these analyses. The study protocol was reviewed and approved by the KPSC Institutional Review Board.

Body mass index was calculated by dividing weight in kilograms by height in meters squared and categorized according to World Health Organization guidelines: healthy weight was a BMI of 18.5 to 24.9 kg/m

, overweight was a BMI of 25.0 to 29.9 kg/m

, obese class I was a BMI of 30.0 to 34.9 kg/ m

, obese class II was a BMI of 35.0 to 39.9 kg/m

, and obese class III was a BMI of 40 kg/m

 or more.20 If more than 1 BMI measurement was recorded during the study period, the first recorded measurement was used.

Quality-of-care measures for comprehensive diabetes care specified by NCQA included 4 screening measures: A1C testing, eye (retinal) examination, LDL-C screening, and medical attention for nephropathy (microalbuminuria screening or evidence of  nephropathy), and 5 control measures: A1C control (≤9.0% [≤75 mmol/mol] and <7.0% [<53 mmol/mol]), LDL-C control (<2.6  mmol/L), and BP control (<130/80 mm Hg and <140/90 mm Hg) (

 The ADA recommendations for patients with diabetes  are BP less than 130/80 mm Hg and A1C less than 7.0%.

 All recorded quality-of-care measures were performed in the HEDIS  measurement year between July 1, 2007, and June 30, 2008, with the exception of lipid screening and retinal examination, which  could be performed in the measurement year or the year prior to the measurement year. 

Age was calculated from date of birth to July 1, 2007; income was based on 2000 census data geocoded at the block level; and race/ethnicity and number of healthcare encounters during the study period were ascertained from administrative data. All persons of Hispanic ethnicity were coded as Hispanic, regardless of race, and then non-Hispanic persons were categorized based on their race. Individuals with no information on race or Hispanic ethnicity were categorized as unknown. As an overall measure of disease burden, the Deyo adaption of the Charlson Comorbidity Index was calculated using 

 diagnosis codes from inpatient and outpatient encounters within 12 months prior to the measurement period.

Descriptive statistics were used to describe demographic characteristics (age, sex, race/ethnicity, and income) by the BMI categories. Means and standard deviations (SDs) were calculated for continuous variables, and percentages were calculated for categorical variables. Demographic characteristics were compared across BMI categories using the Cochran-Armitage trend test for categorical variables and analysis of variance for continuous variables.

Performance rates for each NCQA HEDIS measure were calculated as the percentage of individuals who achieved the measure overall and by BMI category. According to HEDIS specifications, patients without a BP measure or laboratory test result during the study period were counted as not being at goal for that measure.

The Cochran-Armitage trend test was used to assess trends in performance rates with increasing BMI.

Multivariate logistic regression was used to examine the association between BMI and the individual HEDIS measures after adjusting for age, sex, race/ethnicity, income, number of comorbidities, and number of healthcare encounters during the study period. The presence of a linear trend across BMI categories was tested by including the median of each BMI category as a continuous independent variable in the model. All analyses were performed using SAS software version 9.2 (SAS Institute Inc, Cary, North Carolina).

The study sample was composed of 164,721 individuals with diabetes and a mean age of 56.9 (11.2) years. Overall, 12% of individuals were in the healthy-weight category, 30% were overweight, and 28%, 17%, and 14% were in obese categories I, II, and III, respectively (

). About half of the individuals were male, approximately one-third were Hispanic, and slightly more than half resided in a census block with a household income of less than $50,000. The mean (SD) BMI was 32.4 kg/m

), and BMI was inversely related to mean age. The proportion of males was lower at higher BMI. The proportions of whites and blacks and individuals in the lowest 2 income categories increased with increasing BMI (

Nearly all (94%) individuals had an A1C test, 58% had a retinal examination, 93% had a lipid screening profile performed, and 95% had a documented screening for or evidence of nephropathy. In unadjusted analyses, there was a statistically significant trend toward an increase in performance rates for all screening measures with increasing BMI. Conversely, levels of control for A1C, LDL-C, and BP all decreased as BMI increased. Only 41% of all individuals achieved the optimal control target of <7% (<53 mmol/mol) for A1C. More than half (58%) achieved the LDL-C goal of less than 2.6 mmol/L, 82% achieved the BP goal of less than 140/90 mm Hg, and 54% achieved the BP goal of less than 130/80 mm Hg. The largest disparity across BMI categories was seen  for BP less than 130/80 mm Hg, where control decreased across all BMI categories from 63% in healthy-weight individuals to 46%  of individuals in obese class III (data not shown).

After adjusting for age, sex, race/ethnicity, income, number of comorbidities, and number of healthcare encounters, overweight and obese individuals were more likely to have all 4 screening measures (A1C testing, retinal examination, nephropathy screening, and  lipid screening) compared with their healthy-weight counterparts (

). The odds of having a retinal examination, however, did not consistently increase with increasing BMI (

for trend = .4410). Of note, the adjusted odds ratios (ORs) (95% confidence interval  [CI]) of nephropathy screening in overweight and obese class I, II, and III individuals compared with healthy-weight individuals were 1.17 (1.08-1.27), 1.44 (1.32-1.57), 1.62 (1.48-1.78), and 1.84 (1.66-2.04), respectively (

 for trend <.0001). After adjustment, overweight and obese individuals were less likely to have optimal glycemic control (A1C <7% [<53 mmol/mol]) and BP control (<130/80 mm Hg) than their healthy-weight counterparts. In contrast, LDL-C control consistently increased across BMI categories, with class III obese individuals being more likely (OR = 1.23; 95% CI, 1.17-1.29) to achieve goal compared with their healthy-weight counterparts.

Among a diverse population of individuals with diabetes, our findings suggest that BMI was not a barrier to care for recommended screenings and examinations. We found that those individuals who were overweight and obese were more likely than their healthy-weight counterparts to have the recommended screening tests and examinations performed but were less likely to have their  glucose and BP controlled. Interestingly, we found the opposite for LDL-C control, which increased with increasing BMI.

Our findings are similar to those of several other studies; however, those studies did not exclusively include individuals with diabetes. Data from the California Men’s Health Study demonstrated that overweight and obese men were significantly more likely to have their glucose, cholesterol, and triglycerides tested regardless of whether they had diabetes.

 In the National Ambulatory  Medical Care Survey, obese participants were more likely to receive glucose and A1C testing, although it is not known what  proportion of these patients had diabetes.

 In a quality-of-care study among Medicare beneficiaries and recipients of Veterans  Health  Administration healthcare services, a higher BMI was associated with  increased odds of receiving A1C testing and lipid screening among patients with diabetes.

A study among primary care patients examining diabetes and lipid screening rates found that higher BMI predicted higher screening rates for triglycerides, high-density lipoprotein cholesterol, LDL-C, and A1C; however, only 10% of this study population had diabetes.

None of these aforementioned studies assessed the HEDIS-recommended control measures for diabetes care. To our knowledge, only 1 study to date has examined control rates. Rose and colleagues

 found that among nearly 50,000 individuals over an 18-month period, obese individuals were less likely to have good control of BP, LDL-C, and fasting glucose compared with healthy-weight individuals.

Despite evidence that overweight and obese individuals may avoid seeking healthcare services due to stigmatization, those with the added burden of chronic disease such as diabetes seek medical care more frequently,

thus giving healthcare providers additional opportunities to perform screening examinations and tests. However, since retinal examinations are performed by ophthalmologists outside the primary care setting, increased use of healthcare services among individuals with a chronic disease may not extend to retinal examinations, thus potentially explaining the low screening rates we found for this measure. The high rates of nephropathy screening among obese patients in our study may be related to the high prevalence of nephropathy among people with diabetes and the severity of treatment for end-stage renal disease. Additionally, obesity is a possible risk factor for nephropathy.

 we found weight-related differences for BP control in our study. We also found weight-related differences in A1C control. These findings are important because controlling BP and A1C can reduce the risk of microvascular complications such as nephropathy and cardiovascular disease among persons with diabetes.

 The unexpected finding of increased lipid control with increasing BMI may be due to the ADA recommendation of statin therapy regardless of lipid levels for patients with  diabetes who have cardiovascular disease or 1 or more risk factors for cardiovascular disease.

 Additionally, obese individuals  with elevated LDL-C from a subset of the Framingham Offspring and Third Generation cohorts were found to be more likely to be treated with lipid-lowering agents.

Although we found higher screening rates among overweight and obese individuals, the lower rates of glycemic control and BP control suggest that these patients may need additional encouragement and tools to achieve good control. Provider-led interventions to achieve BP control, however, have shown mixed results.

 Promising results were seen with nurse-led interventions using structured algorithms to improve BP control in people with diabetes.

 Similar nurse-led interventions could be implemented to improve glycemic control and BP control among patients with diabetes. Home BP monitoring and automated telephone interventions have proved to be effective at reducing BP and blood glucose levels and may be cost-effective strategies at the population level.

Moreover, the pharmacist-led home BP monitoring program achieved greater impact among participants with diabetes.

Several limitations of these analyses should be considered. The main limitation is that we only included individuals with a recorded BMI. It is possible that individuals without a documented BMI differ from the study population. However, the number of individuals missing height and weight measurements to calculate BMI was only 2.3% of the sample (n = 3901). Due to the fact that this study was conducted using data from 1 integrated healthcare system in 1 region of the country, our results may not be generalizable to  populations in other geographic areas or in other types of healthcare delivery systems. The large sample size allowed for the detection of small, statistically significant differences. While these differences may not be clinically relevant on an individual level,  they are indicative of a larger trend on a population level. Further, KPSC has laboratory facilities in many of its medical centers, and  results are available online for review by members after the tests are completed. This convenience and availability of results may increase the overall likelihood of patients completing recommended blood tests. Finally, this study adjusted for but did not assess the influence of race/ethnicity on screening performance or control rates. However, 1 study found that race/ethnicity was not consistently associated with lower screening rates in black and Hispanic populations compared with whites.

The strengths of this study include its large and racially/ethnically diverse composition that is reflective of the general population in southern California, a broad age range, and the use of information in the EHR rather than self-reported BMI data. Misclassification errors can result with the use of self-reported BMI data because individuals are likely to underreport weight and to over-report height.

Among patients with diabetes enrolled in a large integrated managed care organization, we found weight-related disparities in A1C and BP control. These findings support the need for increased attention by patients, providers, and healthcare systems to improving BP and glycemic control among overweight and obese patients with diabetes. This study also highlighted the need for better retinal examination performance across all BMI categories, including for healthy-weight individuals. Future studies should focus on identifying modifiable healthcare system, provider, and patient factors associated with health outcomes in overweight and obese individuals with diabetes.

From Kaiser Permanente Southern California (DSR, KJC, JML, TNH, KR), Pasedena, CA.

The research was funded by Southern California Permanente Medical Group.

 The authors (DSR, KJC, JML, TNH, KR) report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Concept and design (KJC, KR); acquisition of data (DSR, KJC, JML, TNH, KR); analysis and interpretation of data (DSR, KJC, JML, TNH, KR); drafting of the manuscript (DSR, KJC, JML, TNH, KR); critical revision of the manuscript for important intellectual content (JML, TNH, KR); statistical analysis (DSR, THN); technical, or logistic support (THN, KR).

 Kristi Reynolds, PHD, MPH, Department of Research and Development, Kaiser Permanente Southern California, 100 S Los Robles, 2nd Fl, Pasadena, CA 91101. E-mail: Kristi.Reynolds@KP.org.1. Bertakis KD, Azari R. Obesity and the use of health care services. Obes Res. 2005;13(2):372-379.

2. Flegal KM, Graubard BI, Williamson DF, Gail MH. Cause-specific excess deaths associated with underweight, overweight, and obesity. JAMA. 2007;298(17):2028-2037.

3. Ma J, Xiao L, Stafford RS. Adult obesity and office-based quality of care in the US. Obesity (Silver Spring). 2009;17(5):1077-1085.

4. Malnick SD, Knobler H. The medical complications of obesity. QJM. 2006;99(9):565-579.

5. Molenaar EA, Hwang S-J, Vasan RS, et al. Burden and rates of treatment and control of cardiovascular disease risk factors in obesity: the Framingham Heart Study. Diabetes Care. 2008;31(7):1367-1372.

This study explores potential weight-related disparities in the quality of care for adults with diabetes in a large health plan according to recommended quality indicators.

January 2014

The Effects of Federal Parity on Substance Use Disorder Treatment

Why Are Medicare and Commercial Insurance Spending Weakly Correlated?

Optimal Management of Diabetes Among Overweight and Obese Adults

Evaluation of Electronic Medical Record Administrative Data Linked Database (EMRALD)

Specialist Participation in Healthcare Delivery Transformation: Influence of Patient Self-Referral

Comorbidities and Cardiovascular Disease Risk in Older Breast Cancer Survivors

Electronic Health Risk Assessment Adoption in an Integrated Healthcare System

Evidence-Based Guidelines to Determine Follow-up Intervals: A Call for Action

Process of Care Compliance Is Associated With Fewer Diabetes Complications

Patient-Centered Medical Home Transformation With Payment Reform: Patient Experience Outcomes