Hospital inpatient and outpatient discharge files from the Florida Agency for Health Care Administration from 1997 through 2008.

This study focuses on inguinal hernia repair surgery, one of the most commonly performed surgical procedures in the United States. We estimate multivariate regressions of inguinal hernia surgery counts in the outpatient setting and in the inpatient setting. The key explanatory variable is the time-varying Medicare payment rate specific to the procedure and hospital. Control variables include time-varying hospital and county characteristics and hospital and year-fixed effects.

Outpatient hernia surgeries fell in response to OPPS-induced rate cuts. The volume of inpatient hernia repair surgeries did not increase in response to reductions in the outpatient reimbursement rate.

Potential substitution from the outpatient setting to the inpatient setting does not pose a serious threat to Medicare's efforts to contain hospital outpatient costs.

We used secondary data from the Early Childhood Longitudinal Study Birth Cohort (ECLS-B) between 2001 and 2006.

We estimated the predicted probability of treatment using tree-based methods and logistic regression (LR). We then modeled the exposure-outcome relation using weighted LR models while considering covariate balance and precision for each propensity score estimation method.

Among approximately 500 VLBW children, therapy receipt was associated with moderately improved preschool motor ability. Overall, ensemble methods produced the best covariate balance (Mean Squared Difference: 0.03–0.07) and the most precise effect estimates compared to LR (Mean Squared Difference: 0.11). The overall magnitude of the effect estimates was similar between RFC and LR estimation methods.

Propensity score estimation using RFC and bagging produced better covariate balance with increased precision compared to LR. Ensemble methods are a useful alterative to logistic regression to control confounding in observational studies.

AHRQ Healthcare Cost and Utilization Project State Inpatient Databases and complete hospital financial reporting of revenues mandated in 10 states for 2006.

Hospital discharge records and hospital financial information were merged to estimate revenue per stay by payer. Estimated prices were validated against other data sources.

Hospital prices can be reasonably estimated for 10 geographically diverse states. All-payer price-to-charge ratios, an intermediate step in estimating prices, compare favorably to cost-to-charge ratios. Estimated prices also compare well with Medicare, MarketScan private insurance, and the Medical Expenditure Panel Survey prices for major payers, given limitations of each dataset.

Public reporting of prices is a consumer resource in making decisions about health care treatment; for self-pay patients, they can provide leverage in negotiating discounts off of charges. Researchers can also use prices to increase understanding of the level and causes of price differentials among geographic areas. Prices by payer expand investigational tools available to study the interaction of inpatient hospital price setting among public and private payers—an important asset as the payer mix changes with the implementation of the Affordable Care Act.

Pennsylvania Trauma Outcome Study.

We performed an observational cohort study on 191,887 patients admitted to 28 Level 1 and Level II trauma centers. The main outcomes of interest were (1) death, (2) death or major complication, and (3) failure-to-rescue. Hospitals were categorized according to the proportion of black patients. Multivariate regression models were used to estimate trends in racial disparities and to assess whether the source of racial disparities was within or between hospitals.

Trauma patients admitted to hospitals with high concentrations of blacks (>20 percent) had a 45 percent higher odds of death (adj OR: 1.45, 95 percent CI: 1.09–1.92) and a 73 percent higher odds of death or major complication (adj OR: 1.73, 95 percent CI: 1.42–2.11) compared with patients admitted to hospitals treating low proportions of blacks. Blacks and whites admitted to the same hospitals had no difference in mortality (adj OR: 1.05, 95 percent CI: 0.87, 1.27) or death or major complications (adj OR: 1.01; 95 percent CI: 0.90, 1.13). The odds of overall mortality, and death or major complications have been reduced by 32 percent (adj OR: 0.68; 95 percent CI: 0.54–0.86) and 28 percent (adj OR: 0.72; 95 percent CI: 0.60–0.85) between 2000 and 2009, respectively. Racial disparities did not change over 10 years.

Despite the overall improvement in outcomes, the gap in quality of care between black and white trauma patients in Pennsylvania has not narrowed over the last 10 years. Racial disparities in trauma are due to the fact that black patients are more likely to be treated in lower quality hospitals compared with whites.

Surveillance, Epidemiology, and End Results data linked with Medicare claims.

We examined delays in adjuvant chemotherapy or radiation for women diagnosed with stage I–III breast cancer during 1992–2007. We used multivariable logistic regression to assess the probability of delay by race/ethnicity and included hospital fixed effects to assess whether hospitals explained disparities.

Among 54,592 women, black (11.9 percent) and Hispanic (9.9 percent) women had more delays than whites (7.8 percent, p < .0001). After adjustment, black (vs. white) women had higher odds of delay (odds ratio = 1.25, 95 percent confidence interval = 1.10–1.42), attenuated somewhat by including hospital fixed effects (OR = 1.17, 95 percent CI = 1.02–1.33).

Hospitals are the important contributors to racial disparities in treatment delay.

The data derive from the American Community Survey. Using a difference-in-difference design, we compare the target population (ages 19–25) with a control group (ages 26–29).

Net private health insurance coverage increased by 4.6 percentage points and overall coverage increased by 4.2 percentage points for people aged 19–25; more for Whites than non-White subgroups.

Changes in coverage for states appear driven by demographics rather than the existence of prior dependent expansions by the state. Disparities in health care coverage remain, but the absolute level of coverage is improving.

Medicare Advantage beneficiaries in 2008.

We examined three drug cost responses using multivariate logistic regression: cost-reducing behaviors (e.g., switching to generics), nonadherence (e.g., not refilling prescriptions), and financial stress (e.g., going without necessities).

Telephone interviews in a stratified random sample (N = 1,201, 70 percent response rate).

After adjustment, a comparable percentage of unsubsidized near-poor (26 percent) and higher income beneficiaries reported cost-reducing behaviors (23 percent, p = .63); fewer LIS beneficiaries reported cost-reducing behaviors (15 percent, p = .019 vs near-poor). Unsubsidized near-poor beneficiaries were more likely to reduce adherence (8.2 percent) than higher income (3.5 percent, p = .049) and LIS beneficiaries (3.1 percent, p = .027). Near-poor beneficiaries also more frequently experienced financial stress due to drug costs (20 percent) than higher income beneficiaries (11 percent, p = .050) and LIS beneficiaries (11 percent, p = .015).

Low-income subsidies provide protection from drug cost-related nonadherence and financial stress. Beneficiaries just above the LIS income threshold are most at risk for these potentially adverse behaviors.

A total of 3,881 work-related emergency department visit records drawn from the 2003–2006 National Hospital Ambulatory Medical Care Surveys.

Secondary, cross-sectional analyses of work-related emergency department visit data were performed. Odds ratios and 95 percent confidence intervals were modeled using logistic regression.

A substantial and increasing proportion of work-related emergency department visits in the United States were not expected to be paid by workers' compensation. Private insurance, Medicaid, Medicare, and workers themselves were expected to pay for 40 percent of the work-related emergency department visits with this percentage increasing annually. Work-related visits by blacks, in the South, to for-profit hospitals and for work-related illnesses were all more likely not to be paid by workers' compensation.

Emergency department-based surveillance and research that determine work-relatedness on the basis of expected payment by workers' compensation systematically underestimate the occurrence of occupational illness and injury. This has important methodological and policy implications.

Birth certificate data for women who gave birth in Washington D.C. and D.C. residents who gave birth in other jurisdictions.

Using propensity score modeling and instrumental variable analysis, we compare maternal and infant outcomes among women who receive prenatal care from birth center midwives and women who receive usual care. We match on observable characteristics available on the birth certificate, and we use distance to the birth center as an instrument.

Birth certificate data from 2005 to 2008.

Women who receive birth center care are less likely to have a C-section, more likely to carry to term, and are more likely to deliver on a weekend, suggesting less intervention overall. While less consistent, findings also suggest improved infant outcomes.

For women without medical complications who are able to be served in either setting, our findings suggest that midwife-directed prenatal and labor care results in equal or improved maternal and infant outcomes.

Medicare Provider of Services data, Cost Reports, and Census data from 1991 through 2010.

We examined market-level changes in entry and exit after payment reforms relative to a preexisting time trend. We also compared changes in high Medicare share markets relative to lower Medicare share markets and for freestanding relative to hospital-based facilities.

We calculated market-level entry, exit, and total stock of home health agencies, skilled nursing facilities, and inpatient rehabilitation facilities from Provider of Services files between 1992 and 2010. We linked these measures with demographic information from the Census and American Community Survey, information on Certificate of Need laws, and Medicare share of facilities in each market drawn from Cost Report data.

Payment reforms reducing average and marginal payments reduced entries and increased exits from the market. Entry effects were larger and more persistent than exit effects. Entry and exit rates fluctuated more for home health agencies than skilled nursing facilities. Effects on number of providers were consistent with entry and exit effects.

Payment reform affects market entry and exit, which in turn may affect market structure, access to care, quality and cost of care, and patient outcomes. Policy makers should consider potential impacts of payment reforms on post-acute care market structure when implementing these reforms.

Four years (2007–2011) of Medicaid claims data on blind and/or disabled beneficiaries, aged 20–64.

Randomized control trial with an intervention group (n = 3,540) that was enrolled in managed care with telephone care management and a control group (n = 1,524) who remained in fee-for-service system without care management services. Multi-disciplinary care coordination teams provided telephone services to the intervention group to address patients' medical and social needs.

Medicaid claims and encounter data for all participants were obtained from the state and the managed care organization.

There was no significant difference in use of primary care, specialist visits, hospital admissions, and emergency department between the intervention and the control group. Care managers experienced challenges in keeping members engaged in the intervention and maintaining contact by telephone.

The lack of success for Medicaid beneficiaries, along with other recent studies, suggests that more intensive and more targeted interventions may be more effective for the high-needs population.

We studied the effects of a comprehensive parity law enacted by Oregon in 2007. Using claims data, we compared expenditures for individuals in four Oregon commercial plans from 2005 through 2008 to a group of commercially insured individuals in Oregon who were exempt from parity.

We used difference-in-differences and difference-in-difference-in-differences analyses to estimate changes in spending, and quantile regression methods to assess changes in the distribution of expenditures associated with parity.

Among 2,195 individuals with SMI, parity was associated with increased expenditures for behavioral health services of $333 (95 percent CI $67, $615), without corresponding increases in out-of-pocket spending. The increase in expenditures was primarily attributable to shifts in the right tail of the distribution.

Oregon's parity law led to higher average expenditures for individuals with SMI. Parity may allow individuals with high mental health needs to receive services that may have been limited without parity regulations.

Veterans Affairs (VA) administrative data and Medicare claims in 2001–2004 for 11,123 Medicare-eligible VA primary care users in 2000.

We compared means and distributions of VA reliance (the proportion of all VA/Medicare primary care visits occurring in VA) predicted from beta-binomial, binomial, and ordinary least-squares (OLS) models.

Beta-binomial model fits the bimodal distribution of VA reliance better than binomial and OLS models due to the nondependence on normality and the greater flexibility in shape parameters.

Increased awareness of beta-binomial regression may help analysts apply appropriate methods to outcomes with bimodal or U-shaped distributions.

A nationwide random sample of long-stay nursing home residents with data elements from Medicare claims and enrollment files, the Minimum Data Set, the Online Survey Certification and Reporting System, and Area Resource File. The sample consisted of 22,200,089 person-quarters from 754,592 individuals who became long-stay residents in 17,149 nursing homes over the period beginning January 1, 2000 through December 31, 2005.

Linear regression models using a pre/post design adjusted for resident, nursing home, market, and state characteristics. Nursing home and year-quarter fixed effects were included to control for time-invariant facility influences and temporal trends associated with hospitalization of long-stay residents.

Adoption of a Medicaid bed-hold policy was associated with an absolute increase of 0.493 percentage points (95% CI: 0.039–0.946) in hospitalizations of long-stay nursing home residents, representing a 3.883 percent relative increase over the baseline mean.

Medicaid bed-hold policies may increase the likelihood of hospitalization of long-stay nursing home residents and increase costs for the federal Medicare program.

2010 Visiting Medical Consultant Database for the state of Iowa (Carver College of Medicine) and the Iowa Physicians Information System (Carver College of Medicine).

We compared shortest driving times to the nearest medical oncologist for all Iowa census tracts under two scenarios: including only primary practice locations and adding monthly visiting consultant clinic locations.

For rural Iowans, the median driving time to the closest site for medical oncology care falls from 51.6 to 19.2 minutes when monthly visiting consultant clinics are considered.

Including visiting consultant clinics has a significant impact on measures of geographic access to cancer care.

Study participants were community-dwelling Medicare beneficiaries in four different regions in the United States.

During personal interviews, participants considered four EOL scenarios, each presenting a choice between a less intense and more intense care option.

Participants selected their initially favored option. Depending on that choice, in the subsequent TT the length of life offered by the more intense option was systematically increased or decreased until the participant “switched” to his or her initially rejected option.

Participants were able to select an initially favored option (in 3 of the 4 scenarios; this was the less intense option). The majority of participants were able to engage with the subsequent TT. In all scenarios, regardless of the increase/decrease in the length of life offered by the more intense option, the majority of participants were unwilling to “switch” to their initially rejected option.

In surveys of populations' preferential attitudes toward EOL care options, the TT was a feasible elicitation method, engaging most participants and measuring the strength of their attitudes. Further methodological work is merited, involving (1) populations with various participant characteristics, and (2) different attributes in the TT task itself.

MEDLINE, the National Guidelines Clearinghouse, the National Quality Measures Clearinghouse, and supplemental sources.

Systematic review of clinical standards (guidelines, quality metrics, or pay-for-performance programs) for glycemic control in adult diabetes patients. The primary outcome was discussion of increased risk for hypoglycemia in vulnerable populations.

Manuscripts identified were abstracted by two independent reviewers using prespecified inclusion/exclusion criteria and a standardized abstraction form.

We screened 1,166 titles, and reviewed 220 manuscripts in full text. Forty-four guidelines, 17 quality metrics, and 8 pay-for-performance programs were included. Five (11 percent) guidelines and no quality metrics or pay-for-performance programs met the primary outcome.

Clinical standards do not substantively incorporate evidence about increased risk for hypoglycemia in vulnerable populations.

The SEER registry and Area Resource File provided county-level data describing cancer care resources and socioeconomic conditions of Kentucky residents.

The outcome variable was rate of BCS without radiotherapy in each county for 2000–2007. Eight-year weighted average rates of radiation therapy providers and hospitals per 100,000 residents were explanatory variables of interest. Exploratory spatial data analyses and spatial econometric models were estimated.

Appalachian counties in Kentucky had significantly fewer radiation oncologists, hospitals with radiation therapy facilities, and surgeons per 100,000 residents than non-Appalachian counties. The likelihood of BCS without radiation was significantly higher among Appalachian compared to non-Appalachian women (42.5 percent vs. 29.0 percent, p < .001). Higher proportions of women not receiving recommended radiotherapy after BCS were clustered in Eastern Kentucky around Lexington. This geographic disparity was partially explained by significantly fewer radiation therapy facilities in Appalachian Kentucky in adjusted analyses.

Scarce radiation therapy resources in Appalachian Kentucky are associated with disparities in receipt of guideline-concordant radiotherapy, suggesting that policy action is needed to improve the cancer treatment infrastructure in disadvantaged mountainous areas.

Leadership of CHCs in Massachusetts.

We used purposeful snowball sampling of CHC leaders for in-depth exploration of reimbursement policies and practices, experiences with WC, and decisions about using WC. We quantified the prevalence of perceived barriers to using WC through a mail survey of all CHCs in Massachusetts.

Emergent coding was used to elaborate themes and processes related to use of WC. Numbers and percentages of survey responses were calculated.

Few CHCs formally discourage use of WC, but underutilization emerged as a major issue: “We see an awful lot of work-related injury, and I would say that most of it doesn't go through workers' comp.” Barriers include lack of familiarity with WC, uncertainty about work-relatedness, and reliance on patients to identify work-relatedness of their conditions. Reimbursement delays and denials lead patients and CHCs to absorb costs of services.

Follow-up studies should fully characterize barriers to CHC use of WC and experiences in other states to guide system changes in CHCs and WC agencies. Education should target CHC staff and workers about WC.

Nursing home data from the Online Survey Certification and Reporting (OSCAR) System merged with state nurse staffing requirements.

Facility-level housekeeping, food service, and activities staff levels are regressed on nurse staffing requirements and other controls using fixed effect panel regression.

OSCAR surveys from 1999 to 2004.

Increases in state direct care and licensed nurse staffing requirements are associated with decreases in the staffing levels of all types of support staff.

Increased nursing home nurse staffing requirements lead to input substitution in the form of reduced support staffing levels.

From pharmacy claims data in a retrospective cohort study, we identified 6,383 prevalent users of oral diabetes medications from four VA Medical Centers. Patients were included in this prevalent cohort if they had one fill both 3 months prior and 4–12 months prior to the index date, defined as the month in which medication copayments increased. To determine whether these historic medication use criteria introduced bias in the estimated response to a $5 medication copayment increase, we compared adherence trends from cohorts defined from different medication use criteria and from different index dates of copayment change. In an attempt to validate the prior observation of an upward trend in adherence prior to the date of the policy change, we replicated time series analyses varying the index dates prior to and following the date of the policy change, hypothesizing that the trend line associated with the policy change would differ from the trend lines that were not.

Medication adherence trends differed when different medication use criteria were applied. Contrary to our expectations, similar adherence trends were observed when the same medication use criteria were applied at index dates when no copayment changes occurred.

To avoid introducing bias due to study design in outcomes assessments of medication policy changes, historic medication use inclusion criteria must be chosen carefully when constructing cohorts of prevalent users. Furthermore, while pharmacy data have enormous potential for population research and monitoring, there may be inherent logical flaws that limit cohort identification solely through administrative pharmacy records.

A microsimulation model was designed to evaluate osteoporosis treatment, estimating cost and utility measures for patient cohorts at high risk of osteoporosis-related fractures. Incremental cost-effectiveness ratios (ICERs) were estimated using a full implementation of CRNs, a partial implementation of CRNs, and no CRNs. A modification to traditional probabilistic sensitivity analysis (PSA) was used to determine how variance reduction can impact a decision maker's view of treatment efficacy and costs.

The full use of CRNs provided a 93.6 percent reduction in variance compared to simulations not using the technique. The use of partial CRNs provided a 5.6 percent reduction. The PSA results using full CRNs demonstrated a substantially tighter range of cost-benefit outcomes for teriparatide usage than the cost-benefits generated without the technique.

CRNs provide substantial variance reduction for cost-effectiveness studies. By reducing variability not associated with the treatment being evaluated, CRNs provide a better understanding of treatment effects and risks.

Kids' Inpatient Databases (KID) for 2003, 2006, and 2009.

KID 2006 subpopulation of neonatal discharges was analyzed by weighted frequency distribution and multivariable logistic regression analyses for the outcome of death, adjusted for insurance status and other variables. Multivariable linear regression analyses were conducted for the outcomes mean adjusted length of stay and hospital charges. The death analysis was repeated with KID 2003 and 2009.

Of 4,318,121 estimated discharges in 2006, 5.4 percent were uninsured. There were 17,892 deaths; 9.5 percent were uninsured. The largest risks of death were five clinical conditions with adjusted odds ratios (AOR) of 13.7–3.1. Lack of insurance had an AOR of 2.6 (95 percent CI: 2.4, 2.8), greater than many clinical conditions; AOR estimates in alternate models were 2.1–2.7. Compared with insureds, uninsureds were less likely to have been admitted in transfer, more likely to have died in rural hospitals and to have received fewer resources. Similar death outcome results were observed for 2003 and 2009.

Uninsured neonates had decreased care and increased risk of dying.

2001–2009 nursing home Minimum Data Set and Online Survey, Certification, and Reporting (OSCAR) datasets.

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.

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.

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.

Three approaches were used to triangulate the program's impact: (1) analysis of trends in national claims for incentivized services pre- and postprogram implementation; (2) fixed effects panel regression models examining the impact of GPs' P4P program participation on provision of incentivized services; and (3) in-depth interviews to explore GPs' perceptions of their own response to the program.

There was a short-term increase in diabetes testing and cervical cancer screens after program implementation. The increase, however, was for all GPs. Neither signing onto the program nor claiming incentive payments was associated with increased diabetes testing or cervical cancer screening. GPs reported that the incentive did not influence their behavior, largely due to the modest payment and the complexity of tracking patients and claiming payment.

Monitoring and evaluating P4P programs is essential, as programs may not spark the envisioned impact on quality improvement.

Medicare claims for a random 5 percent sample of 2009 Medicare beneficiaries linked to American Medical Association Group Practice data identifying provider groups sufficiently large to be eligible for ACO program participation.

We calculated the fraction of community-dwelling beneficiaries whose assignment shifted, as a consequence of including post-acute E&M services, from the group providing their outpatient primary care to a different group providing their inpatient post-acute care.

Assignment shifts occurred for 27.6 percent of 25,992 community-dwelling beneficiaries with at least one post-acute skilled nursing facility stay, and they were more common for those incurring higher Medicare spending. Those whose assignment shifted constituted only 1.3 percent of all community-dwelling beneficiaries cared for by large ACO-eligible organizations (n = 535,138), but they accounted for 8.4 percent of total Medicare spending for this population.

Under current Medicare assignment rules, ACOs may not be accountable for an influential group of post-acute patients, suggesting missed opportunities to improve care coordination and reduce inappropriate readmissions.

The Los Angeles County Participant Reporting System with multicross-sectional annual data (2006–2009) for adult participants (n = 16,637) who received treatment from publicly funded programs (n = 276) for the first time.

Retrospective analyses of county discharge and admission data. Hierarchical linear regressions models were used to test the hypotheses.

Client data were collected during personal interviews at admission and discharge for most participants.

African Americans and Latinos reported lower odds of completing treatment compared with Whites. Within-group analysis revealed significant heterogeneity within racial and ethnic groups, highlighting primary drug problem, days of drug use before admission, and homelessness as significant factors affecting treatment completion. Service factors, such as referral by the criminal justice system, enabled completion among Latinos and Whites only.

These findings have implications for reducing health disparities among members of racial and ethnic minorities by identifying individual and service factors associated with treatment adherence, particularly for first-time clients.

Pooled data from the 2006–2008 Medical Expenditure Panel Survey (MEPS), a U.S. health survey representative of community-dwelling civilians.

We compared the health and service utilization of two groups of people with disabilities to a contrast group without disability. We modeled ED visits on the basis of disability status, measures of health and health conditions, access to care, and sociodemographics.

These variables were aggregated from the household component, the medical condition, and event files to provide average annual estimates for the period spanning 2006–2008.

People with disabilities accounted for almost 40 percent of the annual visits made to U.S. EDs each year. Three key factors affect their ED use: access to regular medical care (including prescription medications), disability status, and the complexity of individuals’ health profiles.

Given the volume of health conditions among people with disabilities, the ED will always play a role in their care. However, some ED visits could potentially be avoided if ongoing care were optimized.

National InPatient Sample sponsored by the Agency for Healthcare Research and Quality Project for the years from 2000 to 2008 using 18 states.

A difference-in-difference methodology compares costs and number of emergency admissions for Washington, Texas, and Georgia, which implemented disease management programs between 2000 and 2008, to states that did not undergo the transition to managed care (N = 103).

Costs and emergency admissions were extracted for diabetic Medicaid enrollees diagnosed in the reform and non-reform states and collapsed into state and year cells.

In the three treatment states, the implementation of disease management programs did not have statistically significant impacts on the outcome variables when compared to the control states.

States that implemented disease management programs did not achieve improvements in costs or the number of emergency of admissions; thus, these programs do not appear to be an effective way to reduce the burden of this chronic disease.

Secondary nursing home level data over the period 2000 through 2008 obtained from the Minimum Data Set, OSCAR, and Medicare claims.

We employ an instrumental variables approach to address the potential endogeneity of the share of SMI residents in nursing homes in a model including nursing home and year fixed effects.

An increase in the share of SMI nursing home residents positively affected the hospitalization rate among non-SMI residents and negatively affected staffing skill mix and level. We did not observe a statistically significant effect on inspection-based health deficiencies or the hospitalization rate for SMI residents.

Across the majority of indicators, a greater SMI share resulted in lower nursing home quality. Given the increased prevalence of nursing home residents with SMI, policy makers and providers will need to adjust practices in the context of this new patient population. Reforms may include more stringent preadmission screening, new regulations, reimbursement changes, and increased reporting and oversight.

2006–2008 administrative claims and enrollment data for the 5 percent Medicare sample.

Sample included 346,583 beneficiary-year observations of statin users enrolled in stand-alone prescription drug plans, excluding low-income subsidy recipients. We estimated the association between a plan's OOP statin costs and statin utilization using an instrumental variable approach to account for potential bias due to plan selection. Adherence was defined as percentage of days covered (PDC) of at least 80 percent. Plan OOP costs were constructed for a representative market basket of statin medications. Analyses controlled for demographic characteristics, cardiovascular disease risk, co-morbidity presence, and regional characteristics.

About 67 percent of the sample had a PDC of at least 80 percent. An increase in annual statin OOP from $200 (50th percentile) to $240 (75th percentile) was associated with a reduction in the rate of adherent beneficiaries from 67 percent to 56 percent (p < .001).

Greater OOP costs for statins are associated with reductions in statin utilization.

Nursing home complaints to the North Carolina Long-Term Care Ombudsman Program and state certification agency from October 2002 through September 2006 were matched with Online Survey Certification and Reporting data and Minimum Data Set Quality Indicators (MDS-QIs).

We examine the association between the number of complaints filed against a facility and measures of inspection violations, staffing levels, and MDS-QIs.

One observation per facility per quarter is constructed by matching quarterly data on complaints to OSCAR data from the same or most recent prior quarter and to MDS-QIs from the same quarter. One observation per inspection is obtained by matching OSCAR data to complaint totals from both the same and the immediate prior quarter.

There is little relationship between MDS-QIs and complaints. Ombudsman complaints and inspection violations are generally unrelated, but there is a positive relationship between state certification agency complaints and inspection violations.

Ombudsman and state certification agency complaint data are resident- and caregiver-derived quality measures that are distinctive from and complement traditional quality measures.

A simulation model of treatment choice and outcome.

Simulation.

Eight simulation scenarios varied with the values placed on measured and unmeasured covariates and the strength of the relationships between the measured and unmeasured covariates. The balance of both measured and unmeasured covariates was compared across patients either grouped or reweighted by propensity scores methods.

Propensity score algorithms require unmeasured covariate variation that is unrelated to measured covariates, and they exacerbate the imbalance in this variation between treated and untreated patients relative to the full unweighted sample.

The balance of measured covariates between treated and untreated patients has opposite implications for unmeasured covariates in randomized and observational studies. Measured covariate balance between treated and untreated patients in randomized studies reinforces the notion that all covariates are balanced. In contrast, forced balance of measured covariates using propensity score methods in observational studies exacerbates the imbalance in the independent portion of the variation in the unmeasured covariates, which can be likened to squeezing a balloon. If the unmeasured covariates affecting treatment choice are confounders, propensity score methods can exacerbate the bias in treatment effect estimates.

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.

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.

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.

Primary survey data from 40,356 registered nurses in 665 hospitals in four states in 2006 were linked with American Hospital Association and inpatient mortality data from state agencies for approximately 1.3 million patients.

Logistic regression models were used to examine the association between SRN use and 30-day in-hospital mortality and FTR, controlling for patient and hospital characteristics, nurse staffing, the proportion of nurses with bachelor's degrees, and quality of the work environment.

Before controlling for multiple nurse characteristics of hospitals, higher proportions of agency-employed SRNs in hospitals appeared to be associated with higher mortality (OR = 1.06) and FTR (OR = 1.05). Hospitals with higher proportions of SRNs have poorer work environments, however, and the significant relationships between SRNs and mortality outcomes were rendered insignificant when work environments were taken into account.

Higher use of SRNs does not appear to have deleterious consequences for patient mortality and may alleviate nurse staffing problems that could produce higher mortality.

Medicare dialysis population.

Data from the U.S. Renal Data System (USRDS) were used to identify 3,884 freestanding dialysis facilities and 37,942 Medicare patients incident to end-stage renal disease (ESRD) in 2006. The role of injectable medications was evaluated during a 2-year follow-up period by mediational analyses using mixed-effect regression models.

USRDS data were matched with Dialysis Facility Report data from Centers for Medicare and Medicaid Services (CMS) and census data.

There was a strong association found between organizational status and use of injectable drugs. Large for-profit chains used significantly higher injectable medications compared with nonprofit chains and independent facilities. However, the relationship between facility organizational status and patient mortality was not found to be mediated through the higher use of injectable drugs.

Large for-profit chain facilities administered higher IV epoetin, iron, and vitamin D dosages, but this did not result in improved survival. Given the associated costs and lack of a survival benefit, the overuse of injectable medications among the U.S. dialysis patients will likely end under the recent bundling of injectable medications without jeopardizing patient outcomes.

State hospital discharge data, Multi-State Nursing Care and Patient Safety Survey, and hospital characteristics from the AHA Annual Survey.

Cross-sectional analysis of linked patient outcomes for individuals undergoing AAA repair in four states.

Secondary data sources.

Favorable nursing practice environments and higher hospital volumes of AAA repair are associated with lower mortality and fewer failures-to-rescue in main-effects models. Furthermore, nurse staffing interacts with volume such that there is no mortality advantage observed in high-volume hospitals with poor nurse staffing. When hospitals have good nurse staffing, patients in low-volume hospitals are 3.4 times as likely to die and 2.6 times as likely to die from complications as patients in high-volume hospitals (p < .001).

Nursing is part of the explanation for lower mortality after AAA repair in high-volume hospitals. Importantly, lower mortality is not found in high-volume hospitals if nurse staffing is poor.

Nationally representative data from the 2009 Health Center Patient Survey.

Cross-sectional analyses were limited to adults (n = 2,683). We compared sociodemographic characteristics, health conditions, access to health care, and utilization of services among homeless and nonhomeless patients. We also examined the independent effect of homelessness on health care access and utilization, as well as factors that influenced homeless patients' health care experiences.

Computer-assisted personal interviews were conducted with health center patients.

Homeless patients had worse health status—lifetime burden of chronic conditions, mental health problems, and substance use problems—compared with housed respondents. In adjusted analyses, homeless patients had twice the odds as housed patients of having unmet medical care needs in the past year (OR = 1.98, 95 percent CI: 1.24–3.16) and twice the odds of having an ED visit in the past year (OR = 2.00, 95 percent CI: 1.37–2.92).

There is an ongoing need to focus on the health issues that disproportionately affect homeless populations. Among health center patients, homelessness is an independent risk factor for unmet medical needs and ED use.

Patient-reported data were collected prospectively from 337 hospitalized adult patients with pain on medical/surgical oncology units in four hospitals in three states.

This methodological study used a cross-sectional survey design. Each consenting patient completed two PainCQ^© Surveys, the Brief Pain Inventory-Short Form, and demographic questions. Clinical data were extracted from the medical record.

All data were double entered into a Microsoft Access database, cleaned, and then extracted into SPSS, AMOS, and Mplus for analysis.

Confirmatory factor analysis using Structural Equation Modeling supported the initial factor structure. Modification indices guided decisions that resulted in a superior, parsimonious model for the PainCQ-Interdisciplinary Care Survey (six items, two subscales) and the PainCQ-Nursing Care Survey (14 items, three subscales). Cronbach's alpha coefficients all exceeded .80.

Cumulative evidence supports the reliability and validity of the companion PainCQ^© Surveys in hospitalized patients with pain in the oncology setting. The tools may be relevant in both clinical research and quality improvement. Future research is recommended in other populations, settings, and with more diverse groups.

From 1999 through 2008 national Medicare claims data from the 5 percent longitudinal sample of Parts A and B expenditures.

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.

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.

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.

Centers for Medicare and Medicaid Services' Chronic Condition Data Warehouse (2006, 2007).

Retrospective cross-sectional comparison of drug use and proportion of days covered (PDC) for oral-antidiabetics, ACE-inhibitors/ARBs, and antihyperlipidemics among PDP and MAPD enrollees with diabetes. We estimated “naïve” regression models assuming exogenous plan choice and two-stage residual inclusion (2SRI) models to study endogeneity in choice of Part D plan type.

We identified 111,290 diabetics based on ICD-9 codes in Medicare claims from a random 5 percent sample of Medicare beneficiaries in 2005 excluding dual eligibles.

The naïve regression models indicated lower probability of drug use for oral-antidiabetics (−4 percent; p < .001) and ACE-inhibitors/ARBS (−2 percent; p = .004) among PDP enrollees, but their PDC was higher (3–5 percent) for all drug classes (p < .001). 2SRI models produced no significant differences in any-use equations, but significantly higher PDC values for PDP enrollees for oral-antidiabetics and ACE-inhibitors/ARBs.

We found similar overall use of recommended drugs in diabetes treatment and no consistent evidence of favorable or adverse selection into PDPs and MAPDs.

Retrospective analyses of Medicare claims for initial AMI admissions between January and September 2007. Hospital service availability was obtained from annual survey data. Logistic regression estimated the associations of mental illness with admission to hospitals with any invasive cardiac services (diagnostic catheterization, coronary angioplasty, or bypass surgery) and post-admission care patterns and outcomes.

Eighty-two percent of mentally ill AMI patients (n = 28,888) versus 87 percent of other AMI patients (n = 73,895) were initially admitted to hospitals with invasive cardiac facilities [adjusted odds ratio (OR) = 0.81, p < .001]. Admission to such hospitals was associated with overall higher rate of procedure use within 90 days of admission and improved 30-days readmission and mortality rates. However, irrespective of on-site service availability of the admitting hospital, mentally ill patients were one half as likely to receive invasive procedures (adjusted OR approximately 0.5, p < .001).

Among Medicare patients with AMI, those with comorbid mental illness were less likely to be admitted to hospitals with on-site invasive cardiac services. Mental illness was associated with reduced cardiac procedure use within each type of admitting hospitals (with on-site invasive cardiac services or not).

General Internist Community Tracking Study Physician Survey respondents (2000–2001, 2004–2005) linked with residency program evaluations and American Board of Internal Medicine board certification examination score records; n = 2,331.

Cross-sectional regressions of career outcome and practice characteristic measures on board examination scores/success, residency evaluations interacted with residency type, and potential confounding variables.

Failure to achieve board certification was associated with $27,206 (18 percent, p < .05) less income and 14.9 percent more minority patients relative to physicians scoring in the bottom quartile on their initial examination who eventually became certified (p < .01). Other skill measures were not associated with income. Scoring in the top rather than bottom quartile on the board certification examination was associated with 9 percent increased likelihood of reporting high career satisfaction (p < .05). Among physicians trained in community hospital residency programs, lower evaluations were associated with 14.5 percent higher share of minority patients (p < .05). Both skill measures were associated with practice type.

There are associations between early career skill measures and career outcomes. In addition, minority patients are more likely to be treated by physicians with lower early career clinical skills measures than nonminority patients.

Primary data were collected from primary care physicians in North and South Carolina.

In a balanced factorial experimental design, primary care physicians viewed one of 16 (2⁴) video vignette presentations of CHD and provided detailed information about how they would manage the case.

256 MD primary care physicians were interviewed face-to-face in North and South Carolina.

We identify three clusters depicting unique styles of CHD management that are robust to controls for physician (gender and level of experience) and patient characteristics (age, gender, socioeconomic status, and race) as well as key organizational features of physicians' work settings. Physicians in Cluster 1 “Cardiac” (N = 92) were more likely to focus on cardiac issues compared with their counterparts; physicians in Cluster 2 “Talkers” (N = 93) were more likely to give advice and take additional medical history; whereas physicians in Cluster 3 “Minimalists” (N = 71) were less likely than their counterparts to take action on any of the types of management behavior.

Variations in styles of decision making, which encompass multiple outcome variables and extend beyond individual-level demographic predictors, may add to our understanding of disparities in health quality and outcomes.

Obtained SEER-Medicare linked dataset for 4,233 non-Hispanic White, non-Hispanic African American, Hispanic of any race, and Non-Hispanic Asian women aged ≥66 years old diagnosed with ovarian cancer during 2000–2002 from 17 SEER registries. Physician specialty was identified by linking data to the AMA master file using Unique Physician Identification Numbers.

Retrospective claims data analysis for 1999–2006. Logistic regression models were used to analyze the association between GO utilization and race/ethnicity in the initial, continuing, and final phases of care.

GO use decreased from the initial to final phase of care (51.4–28.8 percent). No racial/ethnic differences were found overall and by phase of cancer care. Women >70 years old and those with unstaged disease were less likely to receive GO care compared to their counterparts. GO use was lower in some SEER registries compared to the Atlanta registry.

GO use for the initial ovarian cancer treatment or for longer term care was low but not different across racial/ethnic groups. Future research should identify factors that affect GO utilization and understand why use of these specialists remains low.

Nationally representative internet panel survey conducted in February 2011.

Screener questions identified a sample of 7,812 individuals in private health insurance plans with provider networks who utilized health services within the prior 12 months. Participants reported details of their inpatient and outpatient contacts with out-of-network physicians. An inpatient out-of-network contact was defined as involuntary if: (1) it was due to a medical emergency; (2) the physician's out-of-network status was unknown at the time of the contact; or (3) an attempt was made to find an in-network physician in the hospital but none was available. Outpatient contacts were only defined as involuntary if the physician's out-of-network status was unknown at the time of the contact.

Eight percent of respondents used an out-of-network physician. Approximately 40 percent of individuals using out-of-network physicians experienced involuntary out-of-network care. Among out-of-network physician contacts, 58 percent of inpatient contacts and 15 percent of outpatient contacts were involuntary. The majority of inpatient involuntary contacts were due to medical emergencies (68 percent). In an additional 31 percent, the physician's out-of-network status was unknown at the time of the contact. Half (52 percent) of individuals using out-of-network services experienced at least one contact with an out-of-network physician where cost was not transparent at the time of care.

The frequency of involuntary out-of-network care is not inconsequential. Policy interventions can increase receipt of cost information prior to using out-of-network physician services, but they may be less helpful when patients have constrained physician choice due to emergent problems or limited in-hospital physician networks.

We use a convenience sample of enrollees with employer-provided insurance from the MarketScan database for the years 2003 to 2007. Population weights that consider the age, sex, and geographic distribution of enrollees are assigned to construct representative estimates.

We compare two types of price indexes: (1) a Service Price Index (SPI) that is similar to the BLS index, which holds services fixed and measures the prices of the underlying treatments; (2) a Medical Care Expenditure Index (MCE) that measures the cost of treating diseases and allows for utilization shifts.

Over the entire period of study the CAGR of the SPI grows 0.7 percentage points faster than the preferred MCE index.

Our findings suggest that the health component of inflation may be overstated by 0.7 percentage points per year, and real GDP growth may be understated by a similar amount. However, more work may be necessary to precisely replicate the indexes of the BLS to obtain a more accurate measure of these price differences.

To determine whether the mortality benefit shown for drug-eluting stents (DES) over bare metal stents (BMS) in observational studies persists after controls for/tests for confounding.

Retrospective observational study involving 38,019 patients, 65 years or older admitted for an index percutaneous coronary intervention receiving DES or BMS in Pennsylvania in 2004–2005 followed up for death through 3 years.

Analysis was at the patient level. Mortality was analyzed with Cox proportional hazards models allowing for stratification by disease severity or DES use propensity, accounting for clustering of patients. Instrumental variables analysis used lagged physician stent usage to proxy for the focal stent type decision. A method originating in work by Cornfield and others in 1954 and popularized by Greenland in 1996 was used to assess robustness to confounding.

DES was associated with a significantly lower adjusted risk of death at 3 years in Cox and in instrumented analyses. An implausibly strong hypothetical unobserved confounder would be required to fully explain these results.

Confounding by indication can bias observational studies. No strong evidence of such selection biases was found in the reduced risk of death among elderly patients receiving DES instead of BMS in a Pennsylvanian state-wide population.

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).

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.

Results indicate that increased spending on home-delivered meals was associated with fewer residents in NHs with low-care needs.

States that have invested in their community-based service networks, particularly home-delivered meal programs, have proportionally fewer low-care NH residents.