The Role of Social Risk Factors in Dialysis Quality and Patient Outcomes Under a Medicare Quality Incentive Program.

BACKGROUND: There have been critical improvements in dialysis care and mortality in the last decade. These improvements track with the implementation of the end-stage renal disease (ESRD) Quality Incentive Program (QIP) beginning in 2012, which aligns Medicare payments to dialysis facilities with performance on quality measures. This study explores whether the improvements in dialysis care and patient outcomes under the ESRD QIP have been shared equally among patient groups. METHODS: Our analyses focus on 4 social risk factors: Black race, Hispanic ethnicity, dual eligibility for Medicare and Medicaid, and rurality. We estimated multivariable regressions using longitudinal Medicare and Consolidated Renal Operations in a Web-Enabled Network data. RESULTS: ESRD QIP payment reductions were more common at dialysis facilities with higher proportions of Black and dual-eligible ESRD patients. Patients with dual eligibility and Black race had persistently worse relative outcomes as the ESRD QIP was implemented. This finding was true for differences in outcomes when comparing patients within and across facilities and was not affected by the addition of specific quality measures to the ESRD QIP measure set. Hispanic patients and patients at rural facilities have generally not had worse outcomes since the start of the ESRD QIP. CONCLUSIONS: There is no evidence of widening disparities in dialysis care or patient outcomes across patient groups under the ESRD QIP, which is a longstanding and well-publicized concern with value-based purchasing programs. Relative changes between patient groups since the start of ESRD QIP have not favored any patient group. Many disparities in dialysis quality measures and assessment of dialysis facility payment reductions persist.

The business case for pediatric asthma quality improvement in low-income populations: examining a provider-based pay-for-reporting intervention.

OBJECTIVE: To measure the return on investment (ROI) for a pediatric asthma pay-for-reporting intervention initiated by a Medicaid managed care plan in New York State. DESIGN: Practice-level, randomized prospective evaluation. SETTING: Twenty-five primary care practices providing care to children enrolled in the Monroe Plan for Medical Care (the Monroe Plan). PARTICIPANTS: Practices were randomized to either treatment (13 practices, 11 participated) or control (12 practices). INTERVENTION: For each of its eligible members assigned to a treatment group practice, the Monroe plan paid a low monthly incentive fee to the practice. To receive the incentive, treatment group practices were required to conduct, and report to the Monroe Plan, the results of chart audits on eligible members. Chart audits were conducted by practices every 6 months. After each chart audit, the Monroe Plan provided performance feedback to each practice comparing its adherence to asthma care guidelines with averages from all other treatment group practices. Control practices continued with usual care. MAIN OUTCOME MEASURES: Intervention implementation and operating costs and per member, per month claims costs. ROI was measured by net present value (discounted cash flow analysis). RESULTS: The ROI to the Monroe Plan was negative, primarily due to high intervention costs and lack of reductions in spending on emergency department and hospital utilization for children in treatment relative to control practices. CONCLUSIONS: A pay-for-reporting, chart audit intervention is unlikely to achieve the meaningful reductions in utilization of high-cost services that would be necessary to produce a financial ROI in 2.5 years.

Mortality Risk of Patients Treated in Dialysis Facilities with Payment Reductions under ESRD Quality Incentive Program.

BACKGROUND: The Centers for Medicare & Medicaid Services End-Stage Renal Disease Quality Incentive Program (ESRD QIP) measures quality of care delivered by dialysis facilities and imposes Medicare payment reductions for quality lapses. We assessed the association between payment reductions and patient mortality, a quality indicator not included in the ESRD QIP measure set. METHODS: Association between mortality and ESRD QIP facility payment reduction based on the year of performance was expressed as the unadjusted rate and patient case-mix-adjusted hazard ratio. We also measured association between mortality and 1-year changes in payment reductions. Retrospective patient cohorts were defined by their treating dialysis facility on the first day of each year (2010-2018). RESULTS: Facility performance resulted in payment reductions for 5%-42% of dialysis facilities over the 9 study years. Patients experienced progressively higher mortality at each payment reduction level. Across all years, unadjusted mortality was 17.3, 18.1, 18.9, 20.3, and 23.9 deaths per 100 patient-years for patients in facilities that received 0%, 0.5%, 1%, 1.5%, and 2% payment reductions, respectively. The adjusted hazard ratio showed a similar stepwise pattern by the level of payment reduction: 1.0 (reference), 1.08 (95% confidence interval [CI], 1.07 to 1.09), 1.15 (95% CI, 1.13 to 1.16), 1.19 (95% CI, 1.16 to 1.21), and 1.34 (95% CI, 1.29 to 1.39). Strength of the association increased from 2010 to 2016. Patients treated in facilities that improved over 1 year generally experienced lower mortality; patients in facilities that performed worse on ESRD QIP measures generally experienced higher mortality. CONCLUSIONS: Patient mortality was associated with ESRD QIP facility payment reductions in dose-response and temporal patterns.

Impact analysis of expanding anti-TNF therapy for Crohn's disease.

Aim: To estimate the impact of universal anti-TNF therapy in patients with moderate-to-severe Crohn's disease. Materials & methods: Developed a population-level Markov model to estimate the impact on health outcomes and medical expenditures of expanding anti-TNF therapy use versus current treatment practices. Results: Reductions in deaths (2600), hip fractures (980), major adverse cardiac events (2700) and patient out-of-pocket medical spending (2%) over 5 years. Total societal costs would be US$22,100 higher per patient per year, primarily due to the high cost of anti-TNF therapy. Conclusion: Expanding anti-TNF therapy use among US adult patients with moderate-to-severe Crohn's disease would reduce morbidity and mortality, decrease disease-related medical costs and increase treatment costs compared with current practice. Despite the higher costs, this approach could substantially benefit patients.

An economic evaluation of teledermatology care delivery for chronic skin diseases.

Aim: Analyze the impact of nationwide implementation of teledermatological care for psoriasis. Methods: Develop a Markov model that estimates the impact of telehealth technology for treatment of moderate-to-severe psoriasis on health and healthcare expenditures compared with in-person clinical care. Results: Lower medical costs by US$1.5 billion and total social costs of US$4.3 billion over 5 years. Patients save more than 67 million hours in work absenteeism and travel time, valued at US$598 million. Employers save US$1.2 billion over 5 years due to decreased employee absenteeism. Conclusion: National implementation of telehealth for psoriasis care has the potential to substantially reduce both formal healthcare costs and informal costs for families and patients, while maintaining equivalent clinical outcomes as traditional in-person care.

Lay abstract Recent innovations in telehealth (methods of accessing and managing healthcare using digital information and technologies, such as the computer or telephone) have allowed for convenient access to care for dermatology patients. A recent study found that psoriasis patients experienced similar quality of care outcomes when using telehealth services as they did when seeing a clinician in-person. This paper builds on previous work examining teledermatological care outcomes and aims to analyze the potential economic impacts of a nationwide implementation of telehealth for managing psoriasis. We model the impacts of healthcare utilization and costs, caregiver time loss, and work absenteeism costs between the telehealth and in-person care settings. These results produce cost savings to society of over US$4.3 billion over 5 years.

An impact evaluation of two modes of care for sickle cell disease crises.

Aim: To estimate the economic impacts of increased use of specialty care infusion centers for treating adults experiencing vaso-occlusive crises. Methods: A Markov model is developed to estimate the impact of expanding use of specialty care infusion centers to treat vaso-occlusive crises compared to emergency department care. Results: Access to infusion centers for sickle cell disease could result in savings over US$1.9 billion in formal medical costs and over US$2 billion in societal costs, based on uptake assumptions over 10 years. Conclusion: Expansion of adult sickle cell disease centers across the nation could lead to considerably better economic outcomes in the form of reduced costs and hospital length of stay in addition to improved clinical outcomes as reported in the existing literature.

Specialty care centers for sickle cell disease crises offer improved care and patient experience over typical emergency department care. This paper is based on a recent study, which compared the treatment for sickle cell disease crises at specialty centers and emergency departments by estimating the potential economic impacts of expanded specialty care centers. We estimate the impacts of increased specialty care use on healthcare costs, caregiver time, patient time and employer absenteeism costs. When 35% of the USA adult population acquires access to a specialty sickle cell disease center, a total savings over US$1.9 billion can be obtained after 10 years.

An economic evaluation of reducing colorectal cancer surveillance intensity.

Aim: Analyze the impact of national implementation of 'low intensity' post-treatment colorectal cancer surveillance compared with current practices. Materials & methods: Create a population-level Markov model to estimate impacts of expansion of low versus high intensity surveillance post-treatment on healthcare utilization, costs and caregiver time loss. Results: Shifting to low intensity colorectal cancer surveillance would reduce patient burden by 301,830 h per patient annually over 5 years. Cost reductions over 5 years were US$43.5 million for Medicare and US$4.2 million for Medicaid. Total societal cost savings equaled US$104.2 million. Conclusion: National implementation of low intensity post-treatment colorectal cancer surveillance has the potential to significantly reduce burden and costs on patients and their caregivers with no added risks to health.

Impact analysis of expanding narrow-spectrum antibiotic use for children with ear, sinus and throat infections.

Aim: Estimate the impacts treating acute respiratory tract infections (ARTIs) in children aged 6 months through 12 years with narrow-spectrum antibiotics. Materials & methods: Decision-tree model to estimate children's health, healthcare utilization and costs, and caregiver's time and costs for using narrow-spectrum antibiotics in eligible children with an ARTI, compared with current use of narrow- and broad-spectrum antibiotics. Results: Reduced adverse drug reactions by 35,750 (14%) cases) and 4750 (12%) fewer emergency department visits, 300 (12%) fewer hospitalizations, and 50,500 (10%) avoided outpatient visits. Annual healthcare costs fell by US$120 million (22%). Total societal costs declined by US$131 million (20%). Conclusion: National implementation of narrow-spectrum antibiotics to treat ARTIs in children improves patient outcomes and reduces caregiver burden and annual healthcare costs.

Using medical records to supplement a claims-based comparative effectiveness analysis of antidepressants.

PURPOSE: Because health insurance claims lack clinical information, comparative effectiveness research studies that rely on these data may be challenging to interpret and may result in biased inference. We conducted an exploratory study to determine if medical information contained in patient charts could offer clinical details that would assist in interpreting the results of a claims-based comparative effectiveness study of selective serotonin reuptake inhibitors (SSRIs). METHODS: Retrospective review of 457 charts of patients initiating SSRI treatment. Descriptive data elements included patient diagnosis, symptoms of depressive and anxiety disorders, provider's assessment, and medication treatment and side effects. RESULTS: Most subjects were excluded from the study because their charts were not accessible (58.7%), they did not have a follow-up visit (55.6%), providers could not be contacted (58.0%), or providers refused participation in the study (36.5%). Among those included in the study, most patients were noted to have depression, but most charts lacked information on the majority of depression symptoms at baseline and follow-up. Few concomitant symptoms, side effects, and other important clinical and treatment characteristics were recorded. CONCLUSIONS: Inability to obtain charts due to plan or provider refusal, lack of available information in charts at key times in the course of illness, and missing data elements posed considerable challenges and prevented firm conclusions beyond those drawn from the parent, claims-based study.

Results of a retrospective claims database analysis of differences in antidepressant treatment persistence associated with escitalopram and other selective serotonin reuptake inhibitors in the United States.

BACKGROUND: Although previous studies have found no differences in response to antidepressant pharmacotherapy between selective serotonin reuptake inhibitors (SSRIs), some recent trials suggest benefits associated with more rapid onset of action. OBJECTIVE: The aim of this work was to compare the likelihood that patients initiating treatment with branded escitalopram, rather than with any of 3 SSRIs (ie, citalopram, fluoxetine, and paroxetine) that are available in generic or branded formulations, would continue therapy with the initial medication after 2 and 6 months. METHODS: We used propensity score-weighted logistic regression to assess the effect of antidepressant choice on the likelihood of continuing treatment, based on data from a large administrative claims database with information about US patients. We modeled the propensity to initiate treatment with escitalopram based on demographic, diagnostic, insurance, and service-use characteristics in the 6 months before treatment initiation and used the results to calculate weights for analysis of treatment continuation. The primary outcome measures were receipt of 2 prescriptions of the index drug in the first 2 months and, among those continuing at 2 months, 4 prescriptions in the first 6 months. Antidepressant choice, cost, and service-use characteristics during the treatment period were included as covariates. Patients who initiated therapy between July 2002 and April 2005 were eligible for inclusion. RESULTS: Based on data for 43,921 patients, at 2 months, escitalopram initiators were more likely to have continued initial medication than those receiving the other SSRIs (66.1% vs 61.9%, respectively; P < 0.01) and less likely to have switched or augmented treatment (4.8% vs 7.6%; P < 0.01). At 6 months, escitalopramtreated patients were also more likely to have continued initial medication (47.1% vs 41.0%; P < 0.01) and less likely to have switched or augmented treatment (9.4% vs 14.4%; P < 0.01). CONCLUSION: Patients initiating treatment with escitalopram were more likely to continue and less likely to switch or augment treatment at 2 and 6 months of therapy compared with those who initiated with alternative SSRIs.

Impacts of a disease management program for dually eligible beneficiaries.

The LifeMasters Supported SelfCare demonstration program provides disease management (DM) services to Florida Medicare beneficiaries who are also enrolled in Medicaid and have congestive heart failure (CHF), diabetes, or coronary artery disease (CAD). The population-based program provides primarily telephonic patient education and monitoring services. Findings from the randomized, intent-to-treat design over the first 18 months of operations show virtually no overall impacts on hospital or emergency room (ER) use, Medicare expenditures, quality of care, or prescription drug use for the 33,000 enrollees. However, for beneficiaries with CHF who resided in high-cost South Florida counties, the program reduced Medicare expenditures by 9.6 percent.

Prevalence of unclaimed prescriptions at military pharmacies.

BACKGROUND: Prescriptions that are ordered by physicians but not picked up by patients represent a potential quality improvement opportunity in health systems. Previous research has demonstrated that anywhere from as little as 0.28% to as much as 30.0% of prescriptions are unclaimed, and that 0.45% to 22.0% of patients fail to claim prescriptions. In the Military Health System (MHS), prescriptions filled at military pharmacies are dispensed with no copayment, providing an opportunity to examine the factors that contribute to unclaimed prescriptions other than out-of-pocket cost. OBJECTIVES: To estimate the prevalence of unclaimed prescriptions in the MHS, investigate reasons for unclaimed prescriptions, and compare self-reported noncompliance, defined as the failure to pick up at least 1 prescription in a 12-month period, with evidence from an administrative database of prescription orders and dispensings. METHODS: Research methods included pharmacy staff interviews at 6 military pharmacies, a telephone survey of beneficiaries who filled prescriptions at these pharmacies, descriptive analysis of survey data, and comparison of administrative pharmacy data with self-reported survey data. Beneficiary interviews, conducted from May through July 2004, covered background characteristics, medical conditions, and unclaimed prescriptions, relying on 12 months of recall regarding noncompliance. Interviews with pharmacy staff covered day-to-day operations, factors that alleviate or exacerbate noncompliance, and the burden that noncompliance places on pharmacies. Administrative data from the Pharmacy Data Transaction Service (pharmacy claims) and Composite Health Care System (CHCS: prescription orders and dispensings) databases were used to select a random sample for the beneficiary survey. Survey respondents' CHCS data were matched to their responses to determine the degree of agreement between self-reports and administrative data. RESULTS: Pharmacy interviews were completed with 30 staff members at 6 military pharmacies, and telephone interviews were completed with 1,214 beneficiaries (60.6% response rate). Beneficiary respondents filled an average of 7 prescriptions in the 5 months approximately surrounding the survey administration time frame (from March to July 2004). More than half (56.8%) of respondents were female, and nearly 60.6% were retired military or their dependents. Among all respondents at all study pharmacies, 8.0% reported failing to claim at least 1 prescription during the prior 12 months. Among survey respondents deemed compliant by CHCS data, 93.8% correctly identified themselves as compliant. However, among patients identified as noncompliant using CHCS data, only 16.0% selfidentified as noncompliant. The administrative data were not concordant with self-report data: of 105 survey respondents identifying themselves as noncompliant in the prior year and matched to administrative data (CHCS), only 58.1% were noncompliant per administrative data, and of 1,065 selfidentifying as compliant, only 61.1% were compliant per administrative data. The most common reasons cited by respondents for not picking up their prescriptions were: no perceived need for the prescription (18.5% of the noncompliant), forgot to pick it up (17.3%), the prescription was not in stock (14.8%), long wait time (11.1%), the prescription was not yet available (10.5%), was out of town (9.9%), and was too busy to pick up the prescription (6.2%). Factors associated with unclaimed prescriptions were: younger age, active duty military status, lower educational levels, and the absence of certain chronic medical conditions (i.e., no claims for cardiovascular medications, no self-reported arthritis). CONCLUSIONS: The present study's survey findings of an 8.0% selfreported noncompliance rate fall in the midrange of noncompliance rates reported in previous literature: between 0.45% and 22.0% in nonmilitary populations. Although reported reasons for noncompliance were generally consistent with those identified in previously published studies, they were only partially consistent with previous military pharmacy literature, which also found that patients did not know they had a prescription waiting or had some of the prescribed medicine at home. Concordance between measures of noncompliance, comparing administrative data with patient self-report based on 12-month recall, was poor.

Utilization of, and adherence to, drug therapy among medicaid beneficiaries with congestive heart failure.

BACKGROUND: Congestive heart failure (CHF) affects 4.8 million adult Americans, particularly those aged >65 years, and has been described as a "new epidemic" due to the high annual incidence of the disease (an estimated 550,000 new cases per year). OBJECTIVES: The goal of this research was to determine the number of Medicaid beneficiaries with CHF, identify the rate of CHF drug use, estimate adherence rates, examine factors associated with CHF drug use and adherence, and explore policy implications of the research findings. METHODS: Methods used included identifying noninstitutionalized beneficiaries with >or=1 inpatient claim or >or=2 ambulatory claims with a CHF diagnosis and claims for CHF drugs using 1998 State Medicaid Research Files and 1999 Medicaid Analytic eXtract data for Arkansas, California, Indiana, and New Jersey. Patient adherence was estimated using the medication possession ratio (MPR) and days of medication persistence. Multivariate regression models were used to identify factors associated with CHF drug use and adherence. RESULTS: Overall, 84.8% of beneficiaries had claims for at least 1 CHF medication; 15.2% of beneficiaries were not using any CHF medications. Among those with a claim, the mean number of claims per month was 1.4, and 25.8% had >or=4 claims per month. Mean MPR was 71.9% and mean days of medication persistence were 24.8 per month. Persons aged <65 years, men, ethnic minorities, patients with hospital admissions for conditions other than CHF, and beneficiaries with high Chronic Illness and Disability Payment System scores were less likely to have a CHF drug claim and had lower adherence rates. CONCLUSIONS: State Medicaid agencies and Medicare prescription drug plans should consider designing targeted interventions that encourage better adherence among Medicaid beneficiaries with CHF, particularly men, those aged <65 years, ethnic minorities, and patients with poor overall health status.

Prescription drug use and expenditures among dually eligible beneficiaries.

Using Medicaid Analytic eXtract (MAX) claims files for 1999 and 2001, the authors describe patterns of prescription drug use and expenditures among dually eligible Medicare and Medicaid beneficiaries for all Medicaid full dually eligible beneficiaries and three important subgroups: (1) aged, (2) disabled, and (3) full-year nursing home residents. The analyses indicate great variation in use and expenditures across States that cannot be explained through differences in use of cost containment strategies. Further, the findings suggest that Medicare Part D plans may achieve significant savings by providing incentives for greater use of generic drugs.

Integrating health care for high-need medicaid beneficiaries with serious mental illness and chronic physical health conditions at managed care, provider, and consumer levels.

OBJECTIVE: Policies supporting value-based care and alternative payment models, notably in the Affordable Care Act and the Medicare Access & CHIP Reauthorization Act of 2015, offer hope to advance care integration for individuals with behavioral and chronic physical health conditions. The potential for integration to improve quality while managing costs for individuals with high needs, coupled with the remaining financial, operational, and policy challenges, underscores a need for continued discussion of integration programs' preliminary outcomes and lessons. The authors describe the early efforts of the HealthChoices HealthConnections pilot program for adult Medicaid beneficiaries with serious mental illness and co-occurring chronic conditions, which used a navigator model in 3 southeastern Pennsylvania counties. METHOD: The authors conducted a difference-in-differences analysis of emergency department (ED) visits, hospitalizations, and readmissions using Medicaid claims data and collected data about program implementation. RESULTS: ED visits decreased 4% among study group members (n = 4,788) while increasing almost 6% in the comparison group (n = 7,039) during the intervention period (p = .036); there were no statistically significant differences in hospitalizations or readmissions. This pilot demonstrated the promise of nurse navigators (care managers) to bridge gaps between the physical and mental health care systems, and the success of a private-public partnership developing a member profile to share behavioral and physical health information in the absence of an interoperable health information technology system. CONCLUSIONS AND IMPLICATIONS FOR PRACTICE: The implementation lessons can inform state Medicaid Health Home models as well as accountable care organizations considering incorporation of behavioral health care. (PsycINFO Database Record

Lessons from comparative effectiveness research methods development projects funded under the Recovery Act.

BACKGROUND: The American Recovery and Reinvestment Act of 2009 (ARRA) directed nearly US$29.2 million to comparative effectiveness research (CER) methods development. AIM: To help inform future CER methods investments, we describe the ARRA CER methods projects, identify barriers to this research and discuss the alignment of topics with published methods development priorities. METHODS: We used several existing resources and held discussions with ARRA CER methods investigators. RESULTS & CONCLUSION: Although funded projects explored many identified priority topics, investigators noted that much work remains. For example, given the considerable investments in CER data infrastructure, the methods development field can benefit from additional efforts to educate researchers about the availability of new data sources and about how best to apply methods to match their research questions and data.

Public perceptions of comparative effectiveness research and use of evidence in healthcare decision-making.

AIMS: This study elaborates on the public's understanding of comparative effectiveness research (CER) or patient-centered outcomes research (PCOR), attitudes toward CER/PCOR and use of evidence in healthcare decision-making. MATERIALS & METHODS: We conducted six focus groups with the general public - three with individuals actively engaged in healthcare decision-making and three with individuals more passive in their approach. RESULTS: The general public has little knowledge of CER/PCOR, and its perceptions of certain CER/PCOR concepts are inconsistent with those of researchers and policy-makers. Active healthcare consumers value information more than passive consumers and are likely to use evidence in decision-making. CONCLUSION: Providers are an important source for disseminating and communicating CER/PCOR evidence to active and passive consumers.

Workforce development for comparative effectiveness research: training programs funded by the American Recovery and Reinvestment Act.

AIM: We conducted a midstream assessment of the comparative effectiveness research (CER) training programs funded by the American Recovery and Reinvestment Act (ARRA) by examining program characteristics, planned curriculum development activities and core competencies. MATERIALS & METHODS: We examined all 43 training projects funded by the US$46 million ARRA CER investment, collecting data from key informant discussions and a technical expert panel. RESULTS: The majority of projects leveraged institutional resources to provide an individualized combination of didactic and experiential learning supported by strong mentorship. Core competencies included skills in statistical modeling, evidence synthesis (systematic reviews and meta-analysis) and general research design skills. CONCLUSION: ARRA-supported CER training programs enhanced workforce capacity by developing curricula and preparing CER researchers to apply emerging methods and utilize new CER infrastructure.

Association between outpatient visits following hospital discharge and readmissions among Medicare beneficiaries with atrial fibrillation and other chronic conditions.

Atrial fibrillation (AF) afflicts nearly 3 million people in the United States annually, the large majority of whom are Medicare beneficiaries with other chronic illnesses. Beneficiaries with multiple chronic conditions have high hospitalization and readmission rates but evidence on factors associated with readmissions is limited, and little is known about differences in rates between beneficiaries with and without AF. In a retrospective analysis of Medicare claims data, the relationship between outpatient visits within 14 days after hospital discharge and readmission was examined for beneficiaries with AF or other chronic conditions. About half of those beneficiaries with a hospitalization had an outpatient visit within 14 days of discharge. Readmission rates were 11% to 24% lower for beneficiaries with an outpatient visit than for those without one (P < .01). These findings suggest that follow-up care shortly after discharge may lower readmissions for patients with AF or other chronic conditions.

American Recovery and Reinvestment Act investments in data infrastructure.

AIM: This article describes American Reinvestment and Recovery Act comparative effectiveness research data infrastructure (DI) investments and identifies facilitators and barriers to implementation. MATERIALS & METHODS: We reviewed original project proposals, conducted an investigator survey and interviewed project officers and principal investigators. RESULTS: DI projects assembled or enhanced existing clinical datasets, established linkages between public and private data sources and built infrastructure. Facilitators included building on existing relationships across organizations and making collection as seamless as possible for clinicians. CONCLUSION: To sustain DI, investigators should reduce the burden of comparative effectiveness research data collection on practices, adequately address data privacy and security issues, resolve or lessen the impact of data-linking issues and build research capacity for other investigators and clinicians.