Showing papers in "Health Services Research in 2013"

Achieving Integration in Mixed Methods Designs—Principles and Practices

Abstract: Mixed methods research offers powerful tools for investigating complex processes and systems in health and health care. This article describes integration principles and practices at three levels in mixed methods research and provides illustrative examples. Integration at the study design level occurs through three basic mixed method designs—exploratory sequential, explanatory sequential, and convergent—and through four advanced frameworks—multistage, intervention, case study, and participatory. Integration at the methods level occurs through four approaches. In connecting, one database links to the other through sampling. With building, one database informs the data collection approach of the other. When merging, the two databases are brought together for analysis. With embedding, data collection and analysis link at multiple points. Integration at the interpretation and reporting level occurs through narrative, data transformation, and joint display. The fit of integration describes the extent the qualitative and quantitative findings cohere. Understanding these principles and practices of integration can help health services researchers leverage the strengths of mixed methods.

Measures of Social Deprivation That Predict Health Care Access and Need within a Rational Area of Primary Care Service Delivery

Abstract: Objective To develop a measure of social deprivation that is associated with health care access and health outcomes at a novel geographic level, primary care service area.

Evaluating survey quality in health services research: a decision framework for assessing nonresponse bias.

Abstract: Objective: To address the issue of nonresponse as problematic and offer appropriate strategies for assessing nonresponse bias. Study Design: A review of current strategies used to assess the quality of survey data and the challenges associated with these strategies is provided along with appropriate post-data collection techniques that researchers should consider. Principal Findings: Response rates are an incomplete assessment of survey data quality and quick reactions to response rate should be avoided. Based on a five-question decision making framework we offer potential ways to assess nonresponse bias along with a description of the advantages and disadvantages to each. Conclusions: It is important that the quality of survey data be considered to assess the relative contribution to the literature of a given study. Authors and funding agencies should consider the potential effects of nonresponse bias both before and after survey administration and report the results of assessments of nonresponse bias in addition to response rates.

Health status and health care experiences among homeless patients in federally supported health centers: findings from the 2009 patient survey.

Abstract: Objective To examine health status and health care experiences of homeless patients in health centers and to compare them with their nonhomeless counterparts.

Nonresponse rates are a problematic indicator of nonresponse bias in survey research

Abstract: Survey researchers are rightly concerned with measuring the level of potential bias in estimates generated from the surveys.2 Bias in estimates can result from measurement error, processing/editing error, coverage error, and nonresponse error (Federal Committee on Statistical Methodology [FCSM] 2001). It is nonresponse bias that is the focus of this editorial and it is also the subject of the paper by Halbesleben and Whitman (2013) that this editorial accompanies. Nonresponse bias is a perennial concern for survey researchers as not everyone we attempt to include in our surveys responds. And to the extent that nonrespondents are different from respondents on the key variables the survey was designed to study these differences could bias the very estimates the survey was designed to make. Because we often have very little information about those who do not respond, survey researchers have long focused on the response rate as a key indicator of survey quality. The assumption is that the more nonresponse there is in a survey, the higher the potential for nonresponse bias. Unfortunately, this assumption has served as a problematic diversion for survey research from the real concern of how survey nonresponse potentially biases survey estimates.

Disparities in Completion of Substance Abuse Treatment between and within Racial and Ethnic Groups

Abstract: Concern about disparities in substance abuse treatment (SAT) outcomes among racial/ethnic groups in the United States has led to efforts to identify differences in substance abuse patterns and response to treatment. Yet this research has focused on between-group differences, mainly of African Americans and Whites (see Grella and Joshi 1999; Smith and Weisner 2000; Green et al. 2002; Hser et al. 2003, 2004; Satre, Mertens, and Weisner 2004), with little attention to differences between these two groups and Latinos, the fastest growing population in SAT (Morgenstern and Bux 2003; Marsh et al. 2009; Guerrero et al. 2012). The few comparative studies that include Latinos (Marsh et al. 2009) also obfuscate service and psychosocial factors within these groups that may significantly impact their ability to complete treatment. Using multicross-sectional annual data (2006–2009) from adult participants who received treatment for the first time in Los Angeles County, this study tests the extent to which between- and within-group differences in individual and program characteristics exist for African American, Latino, and White clients, and how these differences interact with treatment completion in outpatient settings. Successful completion of SAT is a well-established process outcome measure associated with long-term outcomes, such as less future criminal involvement and fewer readmissions (Evans, Li, and Hser 2009; Garnick et al. 2009). As such, this measure is particularly relevant for clients during their first exposure to treatment because successful completion reflects achievement of treatment goals at the client level and, under health care reform legislation, it may become a prevalent measure of program performance at the system level (Arndt 2010; Borys 2011). By examining individual- and service-level measures for an adequate sample of first-time clients, mainly those referred by the criminal justice system, this study involved a comparative analysis of the impact these factors have on treatment completion across racial/ethnic groups. Conceptual Framework Most research on SAT disparities across the United States has focused on individual factors to explain group differences in treatment completion between minorities and Whites. This research has highlighted the factors such as differences in client demographics, primary substance used, and addiction severity (Jacobson, Robinson, and Bluthenthal 2007a,b; Arndt 2010). In particular, findings from the national Treatment Episode Data Set pointed to seven client characteristics associated with a higher likelihood of successfully completing SAT: (1) non-Latino White, (2) female, (3) older than 40, (4) more than 12 years of education, (5) employed, (6) use of alcohol as primary substance, and (7) less than daily substance use at admission (Substance Abuse and Mental Health Services Administration [SAMHSA] 2009). Specifically, primary drug used and severity of drug use at intake, as well as psychosocial stressors, are individual factors associated with higher risk to drop out of treatment. Studies show that use of heroin, methamphetamine, and cocaine compared with alcohol is associated with reduced likelihood of treatment completion (Bluthenthal, Jacobson, and Robinson 2007; SAMHSA 2009). In Los Angeles County in 2006, African Americans were most likely to report using cocaine/crack than other drugs, Latinos were most likely to report using heroin, and Whites were most likely to report using amphetamine (Bluthenthal, Jacobson, and Robinson 2007). In addition, it is well documented that African Americans and Latinos enter treatment with more health, mental health, and social problems than Whites, which can contribute to reduced treatment completion (Marsh et al. 2009). Overall, the aggregate effect of primary drug used and severity of drug use, as well as the prevalence of mental health problems and homelessness, place minorities at a disadvantage in terms of successfully meeting the demands of a structured treatment program (Grella and Stein 2006; Ngo et al. 2009; Niv, Pham, and Hser 2009; Van Dorn, Swanson, and Swartz 2009). Thus, Hypothesis 1 posited that after accounting for primary drug used and days of drug use before admission, as well as history of mental disorder and homelessness status, African American and Latino clients would report lower odds of completing treatment compared with White clients. The literature on treatment completion has also suggested that minority status and primary drug used, such as cocaine, methamphetamine, and heroin compared with alcohol, are interacting factors associated with lower odds of completing treatment (Bluthenthal, Jacobson, and Robinson 2007; SAMHSA 2009). Thus, Hypothesis 2 posited that African Americans and Latinos using cocaine, methamphetamine, and heroin as their primary drug problem would be less likely to complete treatment than Whites and individuals using alcohol. It is well established that African American and Latinos face significant challenges to accessing and remaining in treatment long enough to complete treatment successfully (McKay et al. 2003; Tonigan 2003). However, emerging evidence has highlighted significant heterogeneity within African Americans and Latinos in terms of primary drug of choice, severity of drug use before admission, and prevalence of psychosocial stressors that may inhibit efforts to meet treatment goals (Arndt 2010; Guerrero et al. 2012). Thus, Hypothesis 3 posited that use of illegal drugs as primary drug problem, days of drug use before admission, history of mental disorder, and homelessness status would be associated with lower odds of completing treatment within members of each racial and ethnic group. There is growing recognition of the significant role of specific substance abuse treatment services and system factors in helping individuals achieve treatment completion (Marsh et al. 2009; Marsh, Shin, and Cao 2010). Among those seeking help for substance abuse issues, wait time to treatment entry is the most commonly cited service barrier (Claus and Kindleberger 2002; Appel et al. 2004), whereas unmet service needs are generally factors associated with a reduced likelihood for African Americans and Latinos to complete treatment (Jacobson, Robinson, and Bluthenthal 2007a; Marsh et al. 2009; Niv, Pham, and Hser 2009; Shim et al. 2009). Increasing evidence also suggests that referral source is related to treatment completion (SAMHSA 2009). In particular, drug and probation court referrals to SAT have aimed to facilitate rapid access to social services to achieve timely completion of treatment as a condition of probationary status (Evans, Li, and Hser 2008, 2009). Although African Americans and Latinos are disproportionally represented in the criminal justice system, it is not clear whether individuals benefit from rapid access to treatment and the additional supervision that court referrals offer to ensure successful completion of substance abuse treatment. Thus, Hypothesis 4 posited that fewer days of wait time to treatment entry and referral by the criminal justice system would be associated with higher odds of completing treatment for all members of racial and ethnic groups.

Squeezing the Balloon: Propensity Scores and Unmeasured Covariate Balance

Abstract: The strength of randomized controlled trials (RCTs) is the assumption that randomized treatment assignment yields a balanced distribution of covariates thought to be related to outcome between the treatment and control groups (Rubin 2001). Published studies of RCT results traditionally report a table displaying the balance in measured covariates (e.g., patient age, gender, baseline clinical conditions, etc.) between the treatment and control groups. Demonstrated balance of measured covariates across treatment groups is intended to lend credence that such balance extends to unmeasured covariates (Berk 2004). In the context of observational (nonrandomized) data, researchers have espoused designing treatment effect studies that mimic the measured covariate balancing properties of RCTs (Rosenbaum and Rubin 1983a,b; Rubin 1997, 2001, 2007; Joffe and Rosenbaum 1999; Shah et al. 2005). The use of a propensity score (PS)—the probability a patient received treatment given the patient's measured covariate values—has become a mainstay in efforts to find measured covariate balance in observational data studies to estimate treatment effects. It has been said that PS-based methods “can be used to design observational studies in a way analogous to the way randomized experiments are designed” (Rubin 2001) with a design attempting to “assemble groups of treated and control units such that within each group the distributions of covariates is balanced” (Rubin 2001). While methodologists are quick to qualify that achieving balance in measured covariates between groups of treated and untreated patients does not “guarantee” balance in unmeasured covariates across groups, measured covariate balance often creates an “expectation” of unmeasured covariate balance as in RCTs (Ward and Johnson 2008). Indeed, a review of the PS literature noted that “many of the articles in our review” imply that “propensity scores might also balance the unknown confounders between exposure groups” (Shah et al. 2005). Several PS-based algorithms have been suggested to create patient samples that are balanced in measured covariates between treated and untreated patients. These algorithms range from stratification (D'Agostino 1998) and matching based on propensity scores (Hall, Summers, and Oberchain 2003; Frolich 2007; Stuart 2010) to using patient-specific propensity scores to weight observations (Rosenbaum 1987; Robins, Hernan, and Brunback 2000). Treatment effect inferences are then made by contrasting average outcomes between treated and untreated patients with similar propensity scores (and correspondingly similar distributions of measured covariates). These algorithms yield unbiased treatment effect estimates only if after balancing measured covariates, unmeasured covariates are “ignorable” or that the remaining unmeasured covariates that affected treatment choice are independent of outcome (Rosenbaum and Rubin 1983a,b; Joffe and Rosenbaum 1999). Unmeasured covariates affecting treatment choice are ignorable if either (1) they have no relationship (either directly or indirectly) with outcome, or (2) they are balanced between treatment and control groups after balancing measured covariates. Neither of these conditions can be verified directly with data available to researchers. The condition that the unmeasured covariates affecting treatment choice have no relationship with outcome is identical to the assumption required to yield unbiased estimates in standard multivariate regression-based treatment effect estimators—treatment is orthogonal to the error term in the outcome relationship after adjusting for the measured covariates included in the regression model (Angrist and Pischke 2009). Stated differently, this condition assumes that none of the unmeasured covariates affecting treatment choice confound the relationship between treatment and outcome. This orthogonal assumption requires theory-based persuasion by researchers for acceptance. Therefore, the conceptual advantage of PS-based methods relative to standard regression appears to hinge on the assumption that balancing measured covariates between treated and nontreated patients leads to unmeasured covariate balance between treated and nontreated patients. If this assumption holds, unbiased treatment effect estimates can be obtained without relying on theory to support the orthogonal assumption. However, PS-based analyses of treatment effects using observational data largely ignore what seems to be a fundamental question—why did patients with the same or similar propensity scores receive different treatments? Intuitively, it would seem that unmeasured factors not accounted for in the PS model must be different between two patients with similar propensity scores for them to receive different treatments. Let patient utility associated with treatment U(T) and no treatment U(NT) be represented in terms of measured (XM) and unmeasured covariates (XU): (1) (2) The measured and unmeasured covariates in equations (1) and (2) represent any factors affecting the utility of treatment versus no treatment for the patient. These covariates could represent factors related to patient preferences over the outcome changes induced by treatment choice (e.g., an actor may value facial changes from cosmetic surgery more than a construction worker) or factors affecting the relative effectiveness of treatment (e.g., a child with an ear infection and a high fever will expect more benefit from an antibiotic than a child with an ear infection and a low fever). A patient will choose treatment if the net utility gain from treatment—NG(T)—is positive: (3) Based on equation (3), patient treatment choices depend on their respective values of XM and XU. If (α2 − β2) > 0 treated patients will tend to have higher average values of XU than untreated patients, but with XM also varying across patients it may be possible to find treated patients with low values of XU and untreated patients with high values of XU. If, however, two patients A and B are matched to have identical values of the measured covariate——and patient A chooses treatment and patient B does not, it must be that: (4) where for patient i NG(T)i equals the net gain of treatment and equals i's value of XU. With a fixed value of XM, for equation (4) to hold it must be that . If (α2 − β2) > 0 treated patients with matched XM values must have higher values of XU than untreated patients. Therefore, across a set of treated and nontreated patients matched on XM, we would expect greater average differences in XU than the average differences in XU between the population of treated and nontreated patients not matched by XM. In this study, we demonstrate the covariate balancing properties of PS-based algorithms through the lens of a simple treatment choice simulation model in which covariates affecting treatment choice are both measured and unmeasured. Prior simulation-based research showed that imbalance in unmeasured covariates related to treatment assignment remains after using PS-based algorithms (Austin, Grootendorst, and Anderson 2007). Others have described the extent in which treatment effect estimates from propensity score-based approaches are sensitive to imbalance in unobserved covariates (Rosenbaum and Rubin 1983a,b; Lin, Psaty et al. 1998). However, it has not been shown how PS-based algorithms affect the balance of unmeasured covariates between treated and untreated patients. In our simulations, we find properties that are problematic for researchers hoping to make treatment effect inferences relying only on the expectation that balancing measured covariate implies balanced unmeasured covariates. To yield treated and untreated patients with similar propensity scores, we find that PS algorithms require imbalance in the portion of the variation of the unmeasured covariates that affect treatment choice that is unrelated to the measured covariates. In addition, as compared with the full unweighted sample, PS algorithms exacerbate the imbalance in the portion of the unmeasured covariates unrelated to the measured covariates between treated and untreated patients. This result is directly counter to the assumption often relied on in applications of propensity score methods that balancing measured covariates implies balance in the unmeasured covariates that affected treatment choice (Shah et al. 2005).

Meaningful use of electronic health record systems and process quality of care: evidence from a panel data analysis of U.S. acute-care hospitals.

Abstract: Objective To estimate the incremental effects of transitions in electronic health record (EHR) system capabilities on hospital process quality. Data Source Hospital Compare (process quality), Health Information and Management Systems Society Analytics (EHR use), and Inpatient Prospective Payment System (hospital characteristics) for 2006–2010. Study Setting Hospital EHR systems were categorized into five levels (Level_0 to Level_4) based on use of eight clinical applications. Level_3 systems can meet 2011 EHR “meaningful use” objectives. Process quality was measured as composite scores on a 100-point scale for heart attack, heart failure, pneumonia, and surgical care infection prevention. Statistical analyses were conducted using fixed effects linear panel regression model for all hospitals, hospitals stratified on condition-specific baseline quality, and for large hospitals. Principal Findings Among all hospitals, implementing Level_3 systems yielded an incremental 0.35–0.49 percentage point increase in quality (over Level_2) across three conditions. Hospitals in bottom quartile of baseline quality increased 1.16–1.61 percentage points across three conditions for reaching Level_3. However, transitioning to Level_4 yielded an incremental decrease of 0.90–1.0 points for three conditions among all hospitals and 0.65–1.78 for bottom quartile hospitals. Conclusions Hospitals transitioning to EHR systems capable of meeting 2011 meaningful use objectives improved process quality, and lower quality hospitals experienced even higher gains. However, hospitals that transitioned to more advanced systems saw quality declines.

Accountable Care Organizations in the United States: market and demographic factors associated with formation.

Abstract: Implementation of the Affordable Care Act is no longer in doubt after the Supreme Court's June 2012 decision to largely uphold the Act and the 2012 elections. Delivery and payment system reforms included in the Act are proceeding, including implementation of accountable care organizations (Berwick and Hackbarth 2012). An accountable care organization (ACO) is a group of providers collectively held responsible for the overall cost and quality of care for a defined patient population. These and other value-based payment reforms are intended to address long-standing problems confronting U.S. health care: uneven quality, unsustainable costs, and care that is fragmented (Fisher et al. 2007; Fisher and Shortell 2010; Casalino and Shortell 2011; Fisher, McClellan, and Safran 2011). ACO implementation began in earnest in 2012: Medicare began the Pioneer ACO program and the Medicare Shared Savings Program to contract with ACOs (Berwick 2011), many organizations began commercial payer ACO contracts (Larson et al. 2012), and several states began negotiating or implementing ACO contracts under Medicaid programs (McGinnis and Small 2012). At this early stage of development, it is not known whether ACOs will be successful in spurring system-wide transformation of U.S. health care. Early evidence from Medicare's Physician Group Practice Demonstration and a similar commercial contract suggests that the model holds promise (Weeks et al. 2010; Song et al. 2011, 2012; Colla et al. 2012). However, many have raised concerns about the likelihood the ACO model will achieve its aims (Burns and Pauly 2012; Eddy and Shah 2012) and the potential for raised prices as providers consolidate to form ACOs (Berenson, Ginsburg, and Kemper 2010; Berenson et al. 2012; Scheffler, Shortell, and Wilensky 2012). In addition, there are concerns about the high cost and technical difficulty of establishing ACOs, including the work necessary to negotiate or apply for an ACO contract and the investment required to reorganize and redesign care (Pollack and Armstrong 2011; Burns and Pauly 2012; Evans 2012; Lewis et al. 2012). Thus, it is not certain whether a substantial number of provider organizations will undertake accountable care. Most important, nothing is known about the characteristics of the health care markets or local areas where ACOs are being implemented. Research has documented that diffusion of ACO precursors, such as managed care or health maintenance organizations (HMOs) and pay-for-performance, is uneven across local areas (Baker 1999; Rosenthal et al. 2006); thus, it is likely that ACO formation will also be uneven, particularly in the early phases of ACO development. In this article, we examine how many ACOs were established in the United States as of August 2012; where those ACOs are located; and what types of American communities are more or less likely to be served by an ACO. Our results indicate that over 55 percent of Americans reside in local areas where at least one ACO has a presence. We use information on demographic and health-system characteristics to identify factors associated with the local formation of ACOs. Our findings confirm concerns that the current ACO model may face barriers to implementation in many regions.

Case definitions for acute myocardial infarction in administrative databases and their impact on in-hospital mortality rates.

Abstract: Objective To identify validated ICD-9-CM/ICD-10 coded case definitions for acute myocardial infarction (AMI).

Racial/ethnic disparity trends in children's mental health care access and expenditures from 2002 to 2007.

Abstract: Noting significant racial/ethnic disparities in mental health care, the U.S. Surgeon General promoted a vision for reducing these disparities in 2001 (U.S. Department of Health and Human Services 2001). Evidence indicated that African American children were less likely than white children to use mental health services, even after adjustment for socioeconomic, family, and regional factors (Cuffe et al. 1995; Cunningham and Freiman 1996; Zahner and Daskalakis 1997). Limited findings for Latino children were mixed. Puerto Ricans received less mental health care than mainland Latino children (Leaf et al. 1996), and Latino children reported fewer lifetime counseling visits than whites (Pumariega et al. 1998). On the other hand, Latino children with both mental health disorder and impairment had less unmet need for care than white children (Flisher et al. 1997). Since then, other studies have emerged documenting disparities in children's mental health care, finding black–white disparities in psychotropic drug use nationwide (Chen and Chang 2002) and in the Medicaid population (Zito et al. 2005), and Latino–white and black–white disparities in psychotropic drug use in the child welfare population (Raghavan et al. 2005). Latino–white and black–white disparities were also found in children's antidepressant use (Kirby, Hudson, and Miller 2010) and stimulant use (Olfson et al. 2003; Hudson, Miller, and Kirby 2007). To our knowledge, previous studies have not assessed disparities in children's overall mental health care expenditures or outpatient mental health care use, or tracked trends in children's mental health care use and spending over time. To understand how disparities in children's mental health care use have changed since the Surgeon General's report, we use the Andersen behavioral model of health care utilization as a conceptual framework (Andersen 1995), recognizing the importance of predisposing factors (e.g., age and gender), enabling factors (e.g., income, education, and insurance), and need-based factors (e.g., mental health status and comorbid physical health status) on mental health care utilization. The IOM Definition of Health Care Disparities Health care disparities have been measured using multiple methods and disparity definitions, including assessing differences in unadjusted means as is done in the National Health care Disparities Reports (AHRQ 2008b), interpreting race coefficients in models that increasingly add available covariates in a regression context (Fiscella et al. 2002; Trivedi et al. 2005; Vaccarino et al. 2005; Guevara et al. 2006), and model-based estimations of disparities that adhere to the Institute of Medicine (IOM) definition of health care disparities (McGuire et al. 2006; Cook, McGuire, and Miranda 2007; Cook et al. 2010b). We use the latter approach in this article, building on the IOM definition put forth in Unequal Treatment (IOM 2002), that defines disparity to be any difference in health care not due to clinical appropriateness, need, or patient preferences for health care services (see Figure 1). Figure 1 The Institute of Medicine Definition of Racial/Ethnic Health Care Disparities Ideally, implementation of the IOM definition of racial/ethnic health care disparities in national health care datasets requires the identification of survey variables that match the constructs of clinical appropriateness, need, and patient preferences. To compute disparities, an analyst would adjust for differences due to clinical appropriateness, need, and patient preferences but include differences due to other variables (Cook, McGuire, and Zaslavsky 2012). Mental and physical health status variables are strong proxies for need for services, as are age and sex given large differences across these categories in prevalence of mental illness (Merikangas et al. 2010; Kessler et al. 2012). Differences due to patient preferences (e.g., perceived value of medical care and tolerance of risk) should also be adjusted for to the extent they are available in the data. Clinical appropriateness may be more difficult to account for in studies that use observational data, and we are unable to account for this in our analysis. Although clinical appropriateness, need, and patient preference variables should be equalized or adjusted across racial/ethnic groups according to the IOM definition, differences due to the operation of health care systems and the legal and regulatory climate should be considered part of the disparity (see Figure 1). In survey data, differences due to measures of socioeconomic status (SES) (e.g., income and education) can be considered to fit into the operation of health care systems category. For example, if lower SES families are less able to navigate the health care system or pay for their children's mental health care, and racial/ethnic minority families are disproportionately represented in lower SES categories, then the operation of the health care system may be affecting disparities through SES. Differences due to discrimination should also be considered to be part of the disparity according to the IOM definition (see Figure 1). In regression models of survey data, the independent effect of race/ethnicity on health care can be considered a proxy for discrimination (National Research Council 2004). Using Longitudinal Data to Assess Disparities in Mental Health Care Analysis of longitudinal data can identify underlying care-seeking behaviors and correlates of treatment behaviors that drive disparities in mental health care, fulfilling in part the need for research that is more directly translatable into disparities reduction policies (Alegria 2009). Unlike cross-sectional data analysis, longitudinal data analysis can distinguish whether disparities in any use of mental health care are due to whites’ greater initiation of care or whites’ longer treatment episodes. The policy implications will differ depending on the result. Also important, individual characteristics in cross-sectional data can be endogenous to the decision to seek treatment. For example, being insured may affect whether an individual seeks treatment, but once treatment is initiated, children may be more likely to be insured (either because providers encourage enrolling children in public programs or parents choose to enroll their children), thus biasing the effect of insurance on treatment downward. In longitudinal data, patient characteristics can be observed in the time period prior to treatment initiation, reflecting the circumstances of individuals when deciding to seek treatment. New Contribution This study measures recent national trends in disparities in children's mental health care and is the first study that we are aware of to measure disparities in children's use of outpatient mental health care. In addition, this article makes two important contributions over and above the prior literature. First, we ground our method of measuring disparities conceptually in the IOM definition of racial/ethnic health care disparities. Second, we capitalize on longitudinal panel data from the Medical Expenditure Panel Surveys (MEPS) and model the initiation of treatment among youth with unmet need for mental health care. This analysis allows us to identify mechanisms associated with disparities in initiation of mental health care.

Effect of Hospital–SNF Referral Linkages on Rehospitalization

Abstract: Over the last three decades, Medicare payment policies created silos that exacerbated health care fragmentation and increased health care transitions, including rehospitalizations. A growing number of hospitalized patients are discharged to postacute care (PAC) settings while hospital length of stay continues to drop (MedPAC 2011). The Affordable Care Act (ACA) instituted numerous provisions designed to break down these payment silos and to make hospitals accountable for their patients' PAC experiences and costs, including unnecessary rehospitalizations. One consequence of these policies may be to incentivize hospitals to enhance collaboration with PAC providers. Using Medicare claims data from beneficiaries newly transferred to skilled nursing facilities (SNFs), we examine the effect of hospitals' concentrating their discharges to particular SNFs on the 30-day rehospitalization rate. The last decade has seen a marked growth in rehospitalization of SNF patients. Between 2000 and 2006, 30-day rehospitalization rates of Medicare beneficiaries newly discharged to SNF rose from 16 to 20 percent. For prior nursing home residents, this increased from 22 to 27 percent, at an estimated total cost to Medicare of $4.34 billion in 2006 (Mor et al. 2010). The average rehospitalization rate during a Medicare-covered nursing home stay in the United States in 2006 was over 23 percent, and it has been climbing for the last decade (Saliba et al. 2000; Intrator, Zinn, and Mor 2004; Mor et al. 2010; Ouslander et al. 2010). Rehospitalizations are a symptom of dysfunction in the continuity of care, but we know little about the interorganizational structures that facilitate or complicate transitions between care settings (Feng et al. 2011a). Although there have been several randomized trials of interventions to reduce rehospitalizations among Medicare beneficiaries discharged home, no similarly rigorous studies of programs designed to reduce rehospitalization from PAC providers have been undertaken (Naylor et al. 1999; Coleman et al. 2006; Jack et al. 2009). Policy changes under the ACA, particularly the rehospitalization penalty which went into effect in 2012, have altered the landscape to the point that hospitals must now consider the clinical capabilities of the settings to which they discharge their patients (Mor and Besdine 2011). However, while hospitals have successfully extended their domain to cover outpatient care and to create “systems” that include physician practice associations, unless a hospital owns an SNF, partnering with one is less common. Nonetheless, it stands to reason that more frequent exchanges between hospital and SNF clinical staff could improve the efficiency and effectiveness of communication. One would assume that frequent referrals reinforce mutual learning, making it possible for hospital–SNF pairs to experience the benefits of a positive volume–outcome relationship. With repetition, easier communication at the administrative and clinical levels is routinized and personal relations overlay and reinforce administrative arrangements. We do know that closures of hospital-based SNFs in the last decade contributed to rising rehospitalization rates, suggesting that stronger organizational ties may be beneficial (Rahman, Zinn, and Mor 2013). In this study, we seek to determine the effect of hospitals concentrating their SNF discharges to a small number of providers on the risk that their discharged Medicare patients will be rehospitalized. We seek to test the impact of this “preferred provider” relationship with an SNF by estimating the impact of referral volume and concentration on the rate of rehospitalization. Our approach is based on the premise that the distance preference of patients acts as an important determinant of transfers from hospital to alternative SNFs. This allows us to identify portions of such transfers that are not due to unobserved quality of care or interorganizational arrangements, making it possible to estimate the net effect of hospital concentration of Medicare discharges on rehospitalizations.

The disparate impact of the ACA-dependent expansion across population subgroups.

Abstract: Young adults are less likely to have insurance than other age groups; 29.7 percent of people aged 19–25 were uninsured in 2010 as compared with 16.3 percent of all people (DeNavas-Walt, Proctor, and Smith 2011). Also, young adults are at the least risk of needing medical services; compared with the population 25 years and older, young adults are healthier and do not consume as many medical services (O'Hara and Caswell 2012). Therefore, at least in some health reform scenarios, expanding employer-provided coverage to dependent young adults is not very costly (Cantor et al. 2012). Thirty-seven states encouraged the expansion of private health insurance for younger adults before the national health insurance reform of 2010 for young adults. For instance, states regulated the private coverage market to increase dependents by using age caps, student status, or financial dependency (National Conference of State Legislatures 2010). Most of the studies that examined the state expansions found no statistical evidence that uninsurance decreased (Long, Yemane, and Stockley 2010; Levine, McKnight, and Heep 2011; Monheit et al. 2011; Blum et al. 2012). Section 2714 of the Patient Protection and Affordable Care Act of 2010 (ACA) targets this issue of insurance coverage among young adults aged 19–25 (P.L. 111-148 2010). The provision, which took effect on September 23, 2010, allows holders of private family plans to add children under the age of 26 to the policy, regardless of the child's circumstance. For example, a young adult can be in school, married, employed, or have his or her own children and also be on a parent's family plan (Kaiser Family Foundation 2010). Consistent with the law, private insurance coverage increased for young adults in 2011. Different survey data sources show wide variation in estimates of the magnitude of the policy's effect. The Annual Social and Economic Supplement to the Current Population Survey (CPS ASEC) estimated that between 2009 and 2010, there were 400,000 newly insured young adults (DeNavas-Walt, Proctor, and Smith 2011). The American Community Survey (ACS) estimated 700,000 newly insured between 2009 and 2011 (Rodean 2012). Turning to private insurance, research using the CPS ASEC data found that the change associated with the ACA was a 2.5 percentage point drop in own employer-sponsored coverage and a 4.3 percentage point gain in dependent employer-sponsored coverage for young adults aged 19–25 (Sommers and Kronick 2012). These changes amounted to a net 2.9 percentage point decrease in the uninsured rate and a 2.8 percentage point increase in private coverage. The National Health Interview Survey estimated 3 million newly insured young adults, comparing September 2010 versus December 2011 (Sommers 2012). Other research using the CPS ASEC found similar changes in dependent coverage, but slightly higher decreases in the uninsured rate (Cantor et al. 2012). This study also found greater impacts of the ACA in states with prior reforms. Prior research has also showed impacts for selected population subgroups, but sample size has limited the power to detect differences between groups (Sommers et al. 2013). In addition to expanding coverage to young adults, the ACA set a goal of reducing disparities in health care through provisions aimed at improving coverage, access, and outcomes. With this in mind, this research looks at how subgroups were differentially affected by the changes in dependent coverage using data from the ACS; the ACS is a large sample survey with power to detect these differences. Following Sommers and Kronick (2012) and Cantor et al. (2012), we focus on the coverage rate between two different age groups (aged 19–25 and aged 26–29) before and after the dependent coverage took effect. This article adds to the literature by examining coverage rates by state, gender, race, Hispanic origin, English-speaking ability, and citizenship status. In doing so, we provide group-specific and state-specific estimates of the policy's impact. Results are presented as both tabular and regression-based difference-in-difference estimates.

The Relationship between Older Americans Act Title III State Expenditures and Prevalence of Low-Care Nursing Home Residents

Abstract: Objective To test the relationship between older Americans Act (OAA) program expenditures and the prevalence of low-care residents in nursing homes (NHs). Data Sources and Collection Two secondary data sources: State Program Reports (state expenditure data) and NH facility-level data downloaded from LTCfocUS.org for 16,030 US NHs (2000–2009). Study Design Using a two-way fixed effects model, we examined the relationship between state spending on OAA services and the percentage of low-care residents in NHs, controlling for facility characteristics, market characteristics, and secular trends. Principal Findings Results indicate that increased spending on home-delivered meals was associated with fewer residents in NHs with low-care needs. Conclusions States that have invested in their community-based service networks, particularly home-delivered meal programs, have proportionally fewer low-care NH residents.

California's minimum nurse staffing legislation: results from a natural experiment.

Abstract: In 1999, California became the first state in the United States to pass legislation requiring minimum nurse-to-patient staffing ratios in acute care hospitals. The legislation, for which nursing unions were outspoken advocates, was, in part, a response to a reported decline in hospitals' nurse staffing and skill mix induced by pressures from increasing managed care penetration. California Assembly Bill (AB 394) required the California State Department of Health Services to establish unit-specific minimum staffing levels for licensed nurses (registered nurses [RNs] and licensed vocational nurses [LVNs]) in acute care hospitals. The draft regulations were released in January 2002 and, after a period of highly contentious public comment, implemented in January 2004. The ratios were phased beginning January 1, 2004, the staffing ratio for medical-surgical areas was set at 1 : 6; in March 2005, the ratio was enriched to 1 : 5. In 2008, additional specialty units were subject to the regulations. Up to 50 percent of licensed nursing hours could be provided by LVNs (Spetz 2004). Studies have concluded that the legislation led to increases in nurse staffing (Donaldson et al. 2005; Spetz et al. 2009; Aiken et al. 2010; Donaldson and Shapiro 2010; Serratt et al. 2011; Cook et al. 2012). McHugh et al. (2011) found that California hospitals increased their nurse staffing significantly more after the legislation than did hospitals in other states and did not simultaneously reduce their skill mix—the ratio of RNs to total nursing staff. However, their study included hospitals in states that had adopted other approaches, for example, hospital-specific staffing requirements, public reporting of nurse staffing, potentially minimizing the extent to which the effects of the legislation could be seen (American Nurses Association [ANA] 2011). Although these studies demonstrated improvements in aggregate RN staffing, due to wide variability in hospitals' prelegislation staffing levels, it is likely that there were heterogeneous hospital responses to the regulations based on their prelegislation staffing levels, and that some hospitals met “minimum” requirements by differentially increasing their use of LVN/LPNs. The conclusions of research investigating whether quality improved following the legislation are mixed (Donaldson and Shapiro 2010). Neither Donaldson et al. (2005) nor Bolton et al. (2007) found significant decreases in falls, decubitus ulcers, or restraint use following implementation of the regulations. Using the Agency for Healthcare Research and Quality's (AHRQ) Patient Safety Indicators (PSIs), Spetz et al. (2009) found no significant improvement in postoperative sepsis, deep vein thrombosis, decubitus ulcers, mortality following pneumonia, or failure to rescue (FTR, death following a complication). Cook et al. (2012), using a panel of California hospitals from 2000 to 2006, found no statistically significant decrease in FTR or decubitus ulcers following implementation of the ratios, and Aiken et al. (2010), using cross-sectional data from California, New Jersey, and Pennsylvania, concluded that hospitals with staffing levels consistent with those mandated in California had significantly better nurse-reported quality and lower levels of mortality and FTR. More recently, Spetz et al. (2011), also using data from California hospitals from 2000 to 2006, found statistically significant decreases in postoperative respiratory failure and pressure ulcers, but no reductions in FTR or selected infections due to medical care, postoperative pulmonary embolism, or deep vein thrombosis. Finally, a trend analysis of state snapshots in AHRQ's most recent National Health Care Quality Report (http://www.ahrq.gov/qual/measureix.htm#quality) reveals that, from 2000 to 2007, overlapping the period of our study, rates of postoperative sepsis and infections due to medical care increased significantly more in California than in 25 other states (authors' unpublished analysis, 2012; see Appendix Table 1). However, the conclusions of these studies are limited by convenience sampling, cross-sectional designs, and failure to use a measure of nurse staffing that takes account of patients' needs for nursing care. Table 1 Variable Means (Standard Deviations) for California and Comparison States, by Preregulation Staffing Quartile* We address these concerns by relying on the natural experiment provided by the passage of California's minimum nurse staffing legislation and using a large panel of hospitals within a before-after comparison group design. We develop measures of acuity-adjusted nurse staffing, using Nursing Intensity Weights (NIWs) (ANA1997; ANA 2000; Needleman et al. 2002; Mark and Harless 2011) that provide a more accurate measure of staffing relative to patient need. We examined the following research questions: Following implementation of the nurse staffing legislation, did acuity-adjusted nurse staffing increase significantly more in California hospitals than in hospitals in comparison states? Did some California hospitals, especially those most affected by the staffing regulations, rely more on LVNs to meet the minimum staffing requirements than other hospitals? If California hospitals increased their acuity-adjusted nurse staffing significantly more than hospitals in comparison states, did quality of care improve significantly more in California hospitals than in hospitals in comparison states?

The effect of pay-for-performance in nursing homes: evidence from state Medicaid programs.

Abstract: Objective Pay-for-performance (P4P) is commonly used to improve health care quality in the United States and is expected to be frequently implemented under the Affordable Care Act. However, evidence supporting its use is mixed with few large-scale, rigorous evaluations of P4P. This study tests the effect of P4P on quality of care in a large-scale setting—the implementation of P4P for nursing homes by state Medicaid agencies. Data Sources/Study Setting 2001–2009 nursing home Minimum Data Set and Online Survey, Certification, and Reporting (OSCAR) datasets. Study Design Between 2001 and 2009, eight state Medicaid agencies adopted P4P programs in nursing homes. We use a difference-in-differences approach to test for changes in nursing home quality under P4P, taking advantage of the variation in timing of implementation across these eight states and using nursing homes in the 42 non-P4P states plus Washington, DC as contemporaneous controls. Principal Findings Quality improvement under P4P was inconsistent. While three clinical quality measures (the percent of residents being physically restrained, in moderate to severe pain, and developed pressure sores) improved with the implementation of P4P in states with P4P compared with states without P4P, other targeted quality measures either did not change or worsened. Of the two structural measures of quality that were tied to payment (total number of deficiencies and nurse staffing) deficiency rates worsened slightly under P4P while staffing levels did not change. Conclusions Medicaid-based P4P in nursing homes did not result in consistent improvements in nursing home quality. Expectations for improvement in nursing home care under P4P should be tempered.

Global Comparators Project: International Comparison of Hospital Outcomes Using Administrative Data

Abstract: There has been substantial discussion on differing quality of health care between countries with different approaches to financing that care. However, there are few direct comparisons of clinical outcomes at hospital level between the various health care systems, although many countries have regional or national indicator projects (Groene, Skau, and Frolich 2008). The first step in determining the optimal approach for health care delivery from among these systems is the creation of an international database of clinical outcomes. This task is not trivial. Administrative data are generally the only feasible resource for this but use a number of coding schemes that require reconciliation. Although the recording of hospital discharge information may be standardized within countries, it is not so across international borders. Differences in diagnostic coding systems create challenges as the relative importance placed on the accuracy of discharge coding will vary between health systems. We established a mechanism for collecting discharge data from hospitals in the United Kingdom, Europe, and the United States. We then reconciled the differing coding systems and entered the harmonized data into statistical risk-adjustment models. Risk-adjusted outcomes of care were then compared using a purpose-built web interface. Comparing outcomes across international boundaries has a number of challenges—including logistical, IT-related, and cultural—as the project partners get to know each other, share data, decide on the clinical areas of common interest, and exchange information on processes of care and other relevant issues. In this article, we describe the data and modeling challenges and our approach to tackling them.

The urban-rural disparity in nursing home quality indicators: the case of facility-acquired contractures.

Abstract: As the trends in population aging continues in the United States, the demand for all forms of long-term care is likely to increase in the coming decades. Meeting this rising demand is more challenging for rural areas because rural elderly have greater long-term care needs than their urban counterparts (Fennell and Campbell 2007) and there are well-documented differences in urban and rural long-term care providers (Phillips, Hawes, and Williams 2003; Hutchinson, Hawes, and Williams 2005).1 One dimension in which there are significant differences between urban-rural providers is in the quality of care provided by nursing homes. A series of papers have generally found that rural nursing homes have lower quality than their urban counterparts after adjusting for case-mix (Buchanan et al. 2004; Phillips et al. 2004; Bolin, Phillips, and Hawes 2006; Kang, Meng, and Miller 2011; Temkin-Greener, Zheng, and Mukamel 2012). While these studies have identified potential sources of these differences in the quality of care, few have investigated the relative importance of these sources in explaining the urban-rural disparity. This study aims to quantify and further our understanding of the sources of the urban-rural disparity in the quality of care provided by nursing homes. The Blinder–Oaxaca decomposition technique is used to determine the extent to which the disparity is attributable to differences in facility and aggregate resident characteristics between rural and urban nursing homes. Decomposition techniques have been used to study racial and ethnic disparities in a variety of health service settings (Zuvekas and Taliaferro 2003; Grabowski and McGuire 2009; Bowblis and Yun 2010). To our knowledge, this is the first paper that uses decomposition techniques to study urban-rural disparities in the quality of care provided by nursing homes. In this paper, quality of care is defined as the proportion of residents within a facility that acquire a contracture after admission to the nursing home. A contracture is an abnormal muscle shortening and joint fixation commonly seen among persons with immobility or central nervous system disorders (Adams and Victor 1993; Fergusson, Hutton, and Drodge 2007). There are several reasons for focusing on contractures in this study. First, contractures are considered a measure of quality of care because the development of a contracture is considered a failure on the part of the nursing home to meet federal quality of care standards (§483.25, Centers for Medicare and Medicaid Services (CMS) [CMS] 2010). Also, contractures are often preventable with proper supervision and intervention (Wagner et al. 2008; Jamshed and Schneider 2010). Second, contractures are a highly prevalent condition among nursing home residents. In 1999, 24 percent of nursing home residents had a contracture regardless of its presence at admission, whereas contracture rates are between 28 and 29 percent for the years 2005 to 2009 (Harrington, Carrillo, and LaCava 2006; Harrington et al. 2010). Third, contractures are an understudied measure of nursing home quality that have not received as much attention as other quality measures (Wagner and Clevenger 2010). In addition, contractures can lead to pain, pressure ulcers, infections, and functional disability, and they have a significant impact on quality of life; therefore, contractures are closely linked with other nursing home quality measures (Kane et al. 2008). Finally, contractures are not directly targeted by Quality Improvement Organizations in the recent quality improvement efforts by the CMS. In summary, this study aims to address the gap in our understanding of an important yet understudied aspect of nursing home quality.

Creating and supporting a mixed methods health services research team.

Abstract: Objective. To use the experience from a health services research evaluation to provide guidance in team development for mixed methods research. Methods. The Research Initiative Valuing Eldercare (THRIVE) team was organized by the Robert Wood Johnson Foundation to evaluate The Green House nursing home culture change program. This article describes the development of the research team and provides insights into how funders might engage with mixed methods research teams to maximize the value of the team. Results. Like many mixed methods collaborations, the THRIVE team consisted of researchers from diverse disciplines, embracing diverse methodologies, and operating under a framework of nonhierarchical, shared leadership that required new collaborations, engagement, and commitment in the context of finite resources. Strategies to overcome these potential obstacles and achieve success included implementation of a Coordinating Center, dedicated time for planning and collaborating across researchers and methodologies, funded support for in-person meetings, and creative optimization of resources. Conclusions. Challenges are inevitably present in the formation and operation of effective mixed methods research teams. However, funders and research teams can implement strategies to promote success.

Hospital and Surgeon Variation in Complications and Repeat Surgery Following Incident Lumbar Fusion for Common Degenerative Diagnoses

Abstract: Low back pain is a condition for which expanding treatments and surgical innovation have outpaced supporting scientific evidence of their effectiveness (Deyo et al. 2009). Policy makers are questioning the value of lumbar fusion surgery for certain indications, and insurance companies have recently tightened coverage for this common procedure (BlueCross & BlueShield of North Carolina 2010). Recent reviews of surgical efficacy suggest that fusion surgery is no better than multidisciplinary, intensive nonsurgical treatment for discogenic back pain (pain due to degenerative discs, without sciatica), but with a worse safety profile and greater cost (Mirza and Deyo 2007; Washington State Health Care Authority 2007). Poor outcomes of lumbar fusion may be particularly pronounced in workers' compensation populations (Maghout-Juratli et al. 2006). Lumbar fusion may have a clearer role for treating deformities such as degenerative spondylolisthesis and scoliosis (Herkowitz and Kurz 1991; Fischgrund et al. 1997). However, even when there is an indication for less invasive surgery, such as decompressive laminectomy for spinal stenosis, complex fusion procedures that increase the risk of a complication may be performed (Deyo et al. 2010). Multilevel fusions and circumferential approaches are often performed without strong evidence of corresponding improvements in pain or physical functioning. A greater understanding of factors associated with lumbar fusion would help inform current debates. Postoperative complications may be influenced by the choice of surgical technique (Fritzell, Hagg et al. 2002; Deyo et al. 2010; Cizik et al. 2012), underscoring the imperative to rigorously evaluate the safety of surgical treatments. However, population-based measures of safety have been only sparsely reported, and little is currently known about hospital and surgeon variation in rates of postoperative complications following spinal fusion. Using a statewide inpatient discharge registry that allowed us to link successive episodes of care for the same patient across multiple years and institutions, we sought to determine the rates of postoperative complications following fusion for degenerative disease, assess the variation in these rates across individual hospitals and surgeons, and identify how much of the variation is accounted for by operative features.

The importance of a high-performance work environment in hospitals.

Abstract: Work environment, sometimes also called “work climate” or “culture,” has become an important factor in health services research, shown in numerous studies to be associated with positive outcomes for workers, patients, and organizations. However, what do we mean when we say an organization has a good work environment or culture or climate? Does an organization have multiple cultures or work environments, for example, on different units or among different professions? If so, whose work environment matters for understanding what an organization does or how it performs? These questions motivated the current study, which examines the diversity of work environments in hospitals as well as the implications for processes and outcomes for organizations, employees, and patients. Drawing on the literature on high-performance work systems, we define work environment as being comprised of a bundle of practices designed to promote broader worker engagement and organizational commitment. This bundle includes but extends beyond sufficient material resources and support for the work itself. It also encompasses managerial practices, such as an emphasis on worker discretion and participation in decision making; facilitation of communication and information sharing; and human resource management practices focused on developing workers’ skills and recruiting and retaining qualified workers (Baron and Kreps 1999; Appelbaum, Bailey et al. 2000; Guthrie 2001; Doeringer, Evans-Klock et al. 2002; Bartel 2004; Evans and Davis 2005). Together, this bundle of management practices comprises what we term a high-performance work environment (HPWE). The HPWE measures we use resemble those identified by Aiken and colleagues in a series of studies describing a supportive work environment for nurses. That research suggests that organizational arrangements that promote nurses’ professional status and discretion yield greater job satisfaction for nurses and better patient outcomes, including greater satisfaction and lower mortality rates (see, e.g., Aiken, Smith et al. 1994; Aiken, Sloane et al. 1997a, b; Aiken and Patrician 2000; Clarke, Sloane et al. 2002; Vahey, Aiken et al. 2004). The strength and consistency of these findings begs the question of whether the benefits of a supportive work environment extend to the experience and performance of other care providers as well. The Institute of Medicine (IOM) recommends that health care be analyzed and understood as a system (Kohn, Corrigan et al. 1999) in which patient safety and quality of care require collaborative, interdisciplinary teamwork focused on patient-centered care (Corrigan, Donaldson et al. 2001). This framework suggests the fruitfulness of expanding the scope of investigation—beyond nurses or any single occupational group for that matter—to include all providers involved in delivering patient care. Such an approach is consistent with the work on high-performance work system in other industries, which emphasizes the importance of engaging and empowering the entire workforce regardless of education, job title, or experience, such that, for example, even the assembly-line worker in a manufacturing plant becomes a crucial partner in organizational performance (Macduffie 1995; Appelbaum, Bailey et al. 2000). In this study, we adopt this same democratic approach and test its applicability in the hospital setting. We hypothesize that a high-performance work environment, as characterized by the perceptions of a broad range of occupations engaged in care, will relate positively to desirable work processes, such as collaboration and empowerment; retention as measured by intent to quit, job satisfaction, and actual turnover; and care quality as measured by patient ratings and adverse events.

Duration of breastfeeding and childhood obesity: a generalized propensity score approach.

Abstract: Objective To estimate the effect of breastfeeding duration on childhood obesity. Data Source The Child Development Supplement (CDS) of the Panel Study of Income Dynamics (PSID). The PSID provides extensive data on the income and well-being of a representative sample of U.S. families from 1968 to the present. The CDS collects information on the children in PSID families ranging from cognitive, behavioral, and health status to their family and neighborhood environment. The first two waves of the CDS were conducted in 1997 and 2002, respectively. The data provide information on 3,271 children and their mothers. Study Design We use the generalized propensity score to adjust for confounding based on continuous treatment, and the general additive model to analyze the adjusted association between treatment and the outcome conditional on the propensity score. The main outcome is the body mass index (BMI) directly assessed during the in-person interview in 2002. Covariates include family, maternal, and child characteristics, many of which were measured in the year the child was born. Principal Findings After using propensity scores to adjust for confounding, the relationship between breastfeeding duration and childhood BMI is trivially small across a range of model specifications, and none of them is statistically significant except the unadjusted model. Conclusions The causal link between duration of breastfeeding and childhood obesity has not been established. Any recommendation of promoting breastfeeding to reduce childhood obesity is premature.

An examination of pay-for-performance in general practice in Australia.

Abstract: Research has documented deficiencies in health care quality in many industrialized countries (Seddon et al. 2001; McGlynn et al. 2003; The Study Group of Diagnosis of the Working Group on Heart Failure of the European Society of Cardiology et al. 2003; Hussey et al. 2004; Tomio et al. 2010). Health policy experts often attribute the suboptimal quality to fee-for-service compensation, which incentivizes visit quantity rather than quality (Harris and Zwar 2007; Collier 2009; Stremikis, Davis, and Guterman 2010). Pay-for-performance (P4P) programs, which pay clinicians based upon achieving or improving specific quality metrics, are increasingly being used to improve quality of care. P4P programs range from offering small bonuses (approximately $400) to clinicians who reach threshold performance on a few indicators, such as breast and cervical cancer screening rates, to the Quality and Outcomes Framework in the United Kingdom that ties approximately one-fourth of clinicians' incomes to reaching over 100 quality indicators (Pearson et al. 2008; Campbell et al. 2009). Recent reviews of the impact of P4P in primary care suggest that the programs generally have had limited, positive impacts (Petersen et al. 2006; Van Herck et al. 2010; de Bruin, Baan, and Struijs 2011; Scott et al. 2011). Van Herck and colleagues estimated that P4P programs have resulted in a 5 percent improvement in quality indicators on average, although they document substantial variation—from no impact to large improvements (Van Herck et al. 2010). Many questions remain unanswered about P4P programs (Petersen et al. 2006; Christianson, Leatherman, and Sutherland 2008). Little is known, for example, about the characteristics that distinguish highly successful programs, the factors that limit the impact of programs, and which clinicians are more responsive to P4P. In addition, while there is evidence from the United Kingdom that P4P programs can narrow socio-economic disparities in health care, there is broad concern that P4P programs may widen disparities (Casalino et al. 2007; Doran et al. 2008; Australian Department of Health and Ageing 2009). This study uses mixed methods to evaluate the impact of a voluntary P4P program that incentivizes Australian general practitioners (GPs) to provide recommended diabetes and asthma care, as well as cervical cancer screening. The goals of this study were to (1) examine the impact of the P4P program on incentivized quality measures; (2) investigate whether there is a differential program impact based upon GP characteristics; and (3) explore GPs' perception of the program's impact on their practice.

Reporting CAHPS and HEDIS data by race/ethnicity for Medicare beneficiaries.

Abstract: Because many beneficiaries may have difficulty understanding health plan performance data when reported for subgroups, this information should be reported as supplemental data and only for plans meeting sample size requirements.

Comparative effectiveness of standard versus patient-centered collaborative care interventions for depression among African Americans in primary care settings: the BRIDGE Study

Abstract: In the United States, most individuals with mental disorders are untreated or poorly treated, and this is particularly true for ethnic minorities (Wang et al. 2005; Cook, McGuire, and Miranda 2007). Despite the proven efficacy of pharmacotherapy and psychotherapy, African Americans with depressive disorders receive lower quality of care (Young et al. 2001; Alegria et al. 2008; Stockdale et al. 2008). When they do seek mental health care, African Americans are seen mostly in primary care settings (Cooper-Patrick et al. 1999), where disparities persist in diagnosis (Borowsky et al. 2000; Stockdale et al. 2008), pharmacotherapy, and psychotherapy referrals (Leo, Sherry, and Jones 1998; Sirey et al. 1999; Young et al. 2001; Stockdale et al. 2008). Disparities between African Americans and whites in the adequacy of treatment with antidepressant medications (Harman, Edlund, and Fortney 2004; Miranda and Cooper 2004) are not entirely explained by differences in education, income, and health insurance coverage (Padgett et al. 1994; Charbonneau et al. 2003; Alegria et al. 2008). Physician knowledge (e.g., impact of race and other social determinants on health and health care), attitudes (e.g., respect for variations in cultural norms, awareness of their own biases), and skills (e.g., patient-centered communication, prescribing behaviors), and patient access barriers (e.g., cultural beliefs, attitudes, and preferences), social context (e.g., experiences of discrimination), and relationships with health professionals are intervention targets for improving outcomes and reducing disparities in depression care (Cooper, Hill, and Powe 2002; Cooper et al. 2006). Primary care physicians discuss depression and engage in rapport-building less frequently with African Americans than whites (Ghods et al. 2008); African Americans also rate their decision making with physicians as less participatory (Cooper-Patrick et al. 1999). Communication disparities may explain lower recognition and treatment rates for African American patients. Compared with whites, African Americans express stronger preferences for counseling (Dwight-Johnson and Lagomasino 2007) and spiritual approaches (Cooper et al. 2001), lower trust in physicians (Boulware et al. 2003), and more negative attitudes toward antidepressant medication (Cooper et al. 2003; Givens et al. 2007), the most common form of treatment of depression used by primary care physicians. Standard collaborative care (CC) strategies (e.g., structured approaches to care based on chronic disease management principles and using depression care managers working in conjunction with a primary care physician and a mental health specialist to monitor mood and medications, coordinate care, and facilitate patient engagement) have demonstrated effectiveness for depression care (Gilbody et al. 2006). Two studies of standard CC interventions show similar improvements (Arean et al. 2005) or better responses (Davis et al. 2011) in treatment and clinical and functional outcomes for minorities versus whites. Another CC intervention that included attention to cultural issues in clinician and patient intervention materials was more effective among minorities than whites at improving depression status, but not at eliminating disparities in treatment or functional outcomes within 12 months (Miranda et al. 2003). In the same study, reductions in treatment and outcome disparities by race/ethnicity were observed after 5 years (Wells et al. 2004). Patient-centeredness and cultural competence are approaches to improve health care quality that are promoted extensively (Institute of Medicine 2002). At the core of both approaches is the ability of health care providers to see patients as unique persons, build effective rapport, use the bio-psychosocial model to explore patient beliefs, values, and meaning of illness, and to find common ground regarding treatment plans. Similarly, both patient-centeredness and cultural competence emphasize the ability of the health care system to align services to meet patients’ needs and preferences (Saha, Beach, and Cooper 2008). Interventions targeting patient-centered communication have shown improvements in patient adherence, satisfaction, and some mental health outcomes (Griffin et al. 2004). Those focused on cultural issues increase patients’ knowledge, decrease access barriers, and improve providers’ cultural competence (Beach et al. 2005; Fisher et al. 2007). Yet few standard care strategies for depression target the contribution of patients’ cultural and social barriers to disparities in health care (Miranda et al. 2003; Loh et al. 2007). Most target provider knowledge of treatment guidelines and disease-oriented management of patients, rather than the quality of patient–clinician communication or cultural relevance of treatment approaches (Rost et al. 2001; Unutzer et al. 2002; Katon et al. 2004; Rubenstein et al. 2006). Interventions aimed at improving patient–clinician relationships and making health care systems more responsive to patients’ needs may provide opportunities to improve outcomes in ethnic minority patients with depression beyond those achieved by standard CC interventions. The Blacks Receiving Interventions for Depression and Gaining Empowerment (BRIDGE) Study is a cluster randomized trial comparing a standard CC intervention for patients (disease management) and clinicians (review of guidelines and mental health consultation) to a patient-centered and culturally tailored CC intervention for patients (care management focused on access barriers, social context, and patient–provider relationships) and clinicians (participatory communication skills training and mental health consultation), hereafter referred to as the patient-centered intervention. We hypothesized that patients in the patient-centered group would have a greater reduction in their depression symptoms, higher rates of depression remission, and greater improvements in mental health functioning at 6, 12, and 18 months than patients in the standard group. We also compared patient ratings of care and receipt of guideline-concordant treatment for depression over time between the two groups and expected better patient ratings of care in the patient-centered group.

Nursing home quality and financial performance: does the racial composition of residents matter?

Abstract: Between 2010 and 2050, the minority population age 65 and older is expected to increase from 20 to 42 percent of the total population of those 65 and older (U.S. Census Bureau 2008). This suggests that the racial composition of nursing home residents will become more diverse over time. While recent studies point to increased nursing home use among blacks (Ness, Ahmed, and Aronow 2004; Smith et al. 2008; Feng et al. 2011), Konetzka and Werner (2009) have documented through a systematic literature review that there are disparities in the quality of nursing home care for minority residents. Racial/ethnic disparities in quality of care may arise from minorities being concentrated in lower performing nursing homes in terms of both quality and financial performance. Although overt segregation, such as the Jim Crow laws, has been abolished in the United States for more than 40 years, de facto segregation continues to exist in the U.S. health care system. Prior research indicates that nursing home segregation exists (Smith, 1990; Smith et al. 2007; Fennell et al., 2000) and various factors may contribute to U.S. nursing home segregation. One contributing factor may be geographic/residential segregation, which may limit nursing home choice for minorities. Black nursing home residents tend to follow residential housing patterns, residing in nursing homes located in their communities (Reed, Andes, and Tobias 2001; Smith et al. 2007, 2008; and Fennell et al., 2000). However, these nursing homes may be lower performing facilities in part as a result of lower socioeconomic conditions of minority communities. Another potential contributor of nursing home segregation is the admission process of residents into nursing homes. Better performing nursing homes may selectively admit residents based on payer status and/or race. Medicaid reimbursement is generally less attractive than private pay reimbursement to nursing homes. As minorities are disproportionately covered by Medicaid, this may further limit nursing home choice and further contribute to nursing home segregation. Nursing homes that disproportionately serve black residents tend to be heavily dependent on Medicaid. As such, they are considered part of the lower tier in what has been described as a two-tiered system of nursing homes in the United States (Mor et al. 2004). Medicaid-dependent facilities are more likely to encounter financial challenges due to lack of other revenue sources (e.g., other payers or philanthropy) needed to overcome Medicaid shortfalls (Weech-Maldonado et al. 2012). As a result, nursing homes with higher proportions of black residents may lack the resources needed to invest in staffing, training, and quality improvement initiatives to promote quality of care in nursing homes. The financial resources available to nursing homes may contribute to racial/ethnic disparities, as the quality of nursing home care has been associated with the availability of resources. Previous studies have used payer mix as a proxy to describe the financial resources available to nursing homes (Mor et al. 2004; Smith et al. 2007; Cai, Mukamel, and Temkin-Greener 2010). However, this approach does not provide a complete assessment of financial performance (revenues and costs) of nursing homes. Although studies suggest that a higher proportion of black residents are associated with lower quality and financial performance in nursing homes, little is known about how financial performance affects the relationship between the racial composition of residents and the quality of care. The study reported here therefore expands on the current literature using actual measures of financial performance (revenues, expenses, operating margin, and total margin) to examine the relationship between the racial composition of nursing home residents and nursing home financial performance. In addition, the study examined whether financial performance mediates the relationship between the racial composition of residents and the quality of care in nursing homes.

Hospital Use of Agency-Employed Supplemental Nurses and Patient Mortality and Failure to Rescue

Abstract: Objective To determine the association between the use of agency-employed supplemental registered nurses (SRNs) to staff hospitals and patient mortality and failure to rescue (FTR).

Favorable selection, risk adjustment, and the Medicare Advantage program.

Abstract: Objectives To examine the effects of changes in payment and risk adjustment on (1) the annual enrollment and switching behavior of Medicare Advantage (MA) beneficiaries, and (2) the relative costliness of MA enrollees and disenrollees. Data From 1999 through 2008 national Medicare claims data from the 5 percent longitudinal sample of Parts A and B expenditures. Study Design Retrospective, fixed effects regression analysis of July enrollment and year-long switching into and out of MA. Similar regression analysis of the costliness of those switching into (out of) MA in the 6 months prior to enrollment (after disenrollment) relative to nonswitchers in the same county over the same period. Findings Payment generosity and more sophisticated risk adjustment were associated with substantial increases in MA enrollment and decreases in disenrollment. Claims experience of those newly switching into MA was not affected by any of the policy reforms, but disenrollment became increasingly concentrated among high-cost beneficiaries. Conclusions Enrollment is very sensitive to payment levels. The use of more sophisticated risk adjustment did not alter favorable selection into MA, but it did affect the costliness of disenrollees.

Simulating the Impact of Long‐Term Care Policy on Family Eldercare Hours

Abstract: Objective To understand the effect of current and future long-term care (LTC) policies on family eldercare hours for older adults (60 years of age and older) in Singapore.

Why Do Some Primary Care Practices Engage in Practice Improvement Efforts Whereas Others Do Not

Abstract: Why do some primary care practices manage to transform themselves into new models of health care delivery whereas others do not? This is the central question of our research because the quality of care received by many Americans is often suboptimal (Schoen et al. 2007; Anderson and Marcovich 2010). Quality issues plaguing primary care include patients' lack of access to services (Huynh et al. 2006), inconsistencies in providing evidence-based medicine (Grol and Grimshaw 2003; McGynn et al. 2003), poor coordination of care across health system components (MacKinney, Ullrich, and Mueller 2011), and complexity involved in caring for individuals with chronic illnesses (Von Korff et al. 1997). Recognizing this problem, the Affordable Care Act emphasizes patient-centered care that is reliable, accessible, and safe; improves the health of the population; and reduces costs to deliver care. Primary care transformation is seen as a key element in meeting these goals. Knowing which practices have adopted new primary care approaches, like the patient-centered medical home (PCMH) model, and contrasting them with those that have not is an important step toward knowing which policies to select to remedy the overall capabilities of primary care delivery. Thus, in this study we differentiate between primary care practices that are and are not transforming to deliver evidence-based medicine, implementing new models of care delivery such as the PCMH, improving transparency through performance measurement and reporting, and creating strategic alliances for advanced integrated care models like accountable care organizations (ACOs). Pressures external to the organization that favor these transformations come via pay-for-performance (P4P) compensation methods, public reporting of performance, government requirements for adoption and meaningful use of electronic health records (EHRs), board recertification processes, and increased expectations from patients and other stakeholder groups. However, primary care practices also experience pressures not to change. For example, payment systems encourage high volume and episodic care, which runs counter to key features of the PCMH and ACO models. Primary care practices are therefore caught in a cross fire of contradictory forces. Recent literature has identified various internal and external factors that may influence practices' ability to transform (Milstein and Gilbertson 2009). Adoption of PCMH components was greatest for large medical groups and for those owned by large health systems—all more likely to have greater resources (Rittenhouse et al. 2008; Goldberg and Mick 2010). The National Demonstration Project identified access to resources as a facilitator of practice transformation, as well as having a supportive infrastructure and management model, facilitative leadership, and an empowering and responsive culture (Nutting et al. 2010). Wise et al. (2011) found that transformation to PCMHs correlated with perceived value of the change, understanding PCMH requirements, leadership and staff commitment, and financial incentives. Reid et al. (2011) reported lack of financial incentives as the primary reason why residency practices discontinued transformation efforts. Fernald et al. (2011) found that embedded culture from historical events, such as previous failed attempts at transformation, a lack of meeting structure, and lack of participation by key practice members influenced practices' ability to transform. They also identified barriers to practice transformation, including a lack of support by leadership and affiliated organizations, and nonsupportive organizational structures and processes. Although these studies present various influences on practice transformation, they do not provide an exploration of both pressures and internal practice characteristics affecting change. The present study begins to fill this gap. There are three critical aspects of current practice transformation efforts (Hoff 2010). First, is added payment for care coordination or case management to break the cycle of “15-minute medicine” caused by volume-driven fee-for-service reimbursement. Second is a “minimum level” of health information technology (HIT) capacity in every practice. And, third, is the transformation of existing patient care and administrative work into team-based care models, in which physicians become team leaders and nurses have increased roles and responsibilities for patient care. The problem is that: It cannot nor should it be expected that after a decade or more of forcing PCPs [primary care physicians] to practice in an assembly-line-like manner provides an immediately favorable environment for practices to innovate…. PCP mindsets are attuned to the demands of high-volume medicine. (Hoff 2010, p. 181) Given forces arrayed against practice transformation efforts, our basic question was what enables a practice to transform itself. Building on previous research was another goal of our study. Our aim was to gain additional knowledge from in-depth case studies to develop a framework explaining the mechanisms of influence and contextual modifiers on performance improvement in physician practices. We studied physician practices in their natural environment to understand performance improvement efforts or their lack and real-life complications, issues, and solutions.