Agreement of Medicaid claims and electronic health records for assessing preventive care quality among adults

The validity of hospital diagnostic and procedure codes reflecting morbidity in preterm neonates born <32 weeks gestation

OBJECTIVE

To determine the validity of diagnostic hospital billing codes for complications of prematurity in neonates <32 weeks gestation.

STUDY DESIGN

Retrospective cohort data from discharge summaries and clinical notes (n = 160) were reviewed by trained, blinded abstractors for the presence of intraventricular hemorrhage (IVH) grades 3 or 4, periventricular leukomalacia (PVL), necrotizing enterocolitis (NEC), stage 3 or higher, retinopathy of prematurity (ROP), and surgery for NEC or ROP. Data were compared to diagnostic billing codes from the neonatal electronic health record.

RESULTS

IVH, PVL, ROP and ROP surgery had strong positive predictive values (PPV > 75%) and excellent negative predictive values (NPV > 95%). The PPVs for NEC (66.7%) and NEC surgery (37.1%) were low.

CONCLUSION

Diagnostic hospital billing codes were observed to be a valid metric to evaluate preterm neonatal morbidities and surgeries except in the instance of more ambiguous diagnoses such as NEC and NEC surgery.

Leveraging behavioral economics and implementation science to engage patients at risk for suicide in mental health treatment: a pilot study protocol

BACKGROUND

Primary care is an ideal setting to connect individuals at risk for suicide to follow-up care; however, only half of the patients referred from the primary care attend an initial mental health visit. We aim to develop acceptable, feasible, low-cost, and effective new strategies to increase treatment initiation among at-risk individuals identified in primary care.

METHODS

We will conduct a multi-phase, mixed-methods study. First, we will conduct a chart review study by using administrative data, including medical records, to identify characteristics of primary care patients at risk for suicide who do or do not attend an initial mental health visit following a referral. Second, we will conduct a mixed methods study by using direct observations and qualitative interviews with key stakeholders (N = 65) to understand barriers and facilitators to mental health service initiation among at-risk individuals. Stakeholders will include patients with suicidal ideation referred from primary care who do and do not attend a first mental health visit, primary care and behavioral health providers, and individuals involved in the referral process. We also will collect preliminary self-report and behavioral data regarding potential mechanisms of behavior change (i.e., self-regulation and social support) from patients. Third, we will leverage these findings, relevant frameworks, and the extant literature to conduct a multi-arm, non-randomized feasibility trial. During this trial, we will rapidly prototype and test strategies to support attendance at initial mental health visits. Strategies will be developed with subject matter experts (N = 10) and iteratively pilot tested (~5 patients per strategy) and refined. Research will be completed in the Penn Integrated Care Program (PIC), which includes fourteen primary care clinics in Philadelphia that provide infrastructure for electronic referrals, patient communication, and data access.

DISCUSSION

We will leverage frameworks and methods from behavioral economics and implementation science to develop strategies to increase mental health treatment initiation among individuals at risk for suicide identified in primary care. This project will lead to an evaluation of these strategies in a fully powered randomized trial and contribute to improvements in access to and engagement in mental health services for individuals at risk for suicide.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT05021224.

Patient-level factors associated with receipt of preventive care in the safety net

Prevention is critical to optimizing health, yet most people do not receive all recommended preventive services. As the complexity of preventive recommendations increases, there is a need for new measurements to capture the degree to which a person is up to date, and identify individual-level barriers and facilitators to receiving needed preventive care. We used electronic health record data from a national network of community health centers (CHCs) in the United States (US) during 2014-2017 to measure patient-level up-to-date status with preventive ratios (measuring up-to-date person-time denoted as a percent) for 12 preventive services and an aggregate preventive index. We use negative binomial regression to identify factors associated with up-to-date preventive care. We assessed 267,767 patients across 165 primary care clinics. Mean preventive ratios ranged from 8.7% for Hepatitis C screening to 83.3% for blood pressure screening. The mean aggregate preventive index was 43%. Lack of health insurance, smoking, and homelessness were associated with lower preventive ratios for most cancer and cardiovascular screenings (p < 0.05). Having more ambulatory visits, better continuity of care, and enrollment in the patient portal were positively associated with the aggregate preventive index (p < 0.05) and higher preventive ratios for all services (p < 0.05) except chlamydia and HIV screening. Overall, receipt of preventive services was low. CHC patients experience many barriers to receiving needed preventive care, but certain healthcare behaviors - regular visits, usual provider continuity, and patient portal enrollment - were consistently associated with more up-to-date preventive care. These associations should inform future efforts to improve preventive care delivery.

Racial and Ethnic Disparities in Acute Care Use for Pediatric Asthma

PURPOSE

Previous work has shown that asthma-related emergency department (ED) use is greatest among Black and Latine populations, but it is unknown whether health care use for exacerbations differs across settings (outpatient, ED, inpatient) and correlates with use of routine outpatient services. We aimed to measure disparities by race, ethnicity, and language in pediatric acute asthma care using data from US primary care community health centers.

METHODS

In an observational study using electronic health records from community health centers in 18 states, we compared non-Hispanic Black, English-preferring Latine, Spanish-preferring Latine, and non-Hispanic White children aged 3 to 17 years on visits for clinic-coded asthma exacerbations (2012-2018). We further evaluated asthma-related ED use and inpatient admissions in a subsample of Oregon-Medicaid recipients. Covariate-adjusted odds ratios (ORs) and rate ratios (RRs) were derived using logistic or negative binomial regression analysis with generalized estimating equations.

RESULTS

Among 41,276 children with asthma, Spanish-preferring Latine children had higher odds of clinic visits for asthma exacerbation than non-Hispanic White peers (OR = 1.10; 95% CI, 1.02-1.18). Among the subsample of 6,555 children insured under Oregon-Medicaid, non-Hispanic Black children had higher odds and rates of asthma-related ED use than non-Hispanic White peers (OR = 1.40; 95% CI, 1.04-1.89 and RR = 1.49; 95% CI, 1.09-2.04, respectively). We observed no differences between groups in asthma-related inpatient admissions.

CONCLUSIONS

This study is the first to show that patterns of clinic and ED acute-care use differ for non-Hispanic Black and Spanish-preferring Latine children when compared with non-Hispanic White peers. Non-Hispanic Black children had lower use of clinics, whereas Spanish-preferring Latine children had higher use, including for acute exacerbations. These patterns of clinic use were accompanied by higher ED use among Black children. Ensuring adequate care in clinics may be important in mitigating disparities in asthma outcomes.VISUAL ABSTRACT.

Comparing ascertainment of chronic condition status with problem lists versus encounter diagnoses from electronic health records

OBJECTIVE

To assess and compare electronic health record (EHR) documentation of chronic disease in problem lists and encounter diagnosis records among Community Health Center (CHC) patients.

MATERIALS AND METHODS

We assessed patient EHR data in a large clinical research network during 2012-2019. We included CHCs who provided outpatient, older adult primary care to patients age ≥45 years, with ≥2 office visits during the study. Our study sample included 1 180 290 patients from 545 CHCs across 22 states. We used diagnosis codes from 39 Chronic Condition Warehouse algorithms to identify chronic conditions from encounter diagnoses only and compared against problem list records. We measured correspondence including agreement, kappa, prevalence index, bias index, and prevalence-adjusted bias-adjusted kappa.

RESULTS

Overlap of encounter diagnosis and problem list ascertainment was 59.4% among chronic conditions identified, with 12.2% of conditions identified only in encounters and 28.4% identified only in problem lists. Rates of coidentification varied by condition from 7.1% to 84.4%. Greatest agreement was found in diabetes (84.4%), HIV (78.1%), and hypertension (74.7%). Sixteen conditions had <50% agreement, including cancers and substance use disorders. Overlap for mental health conditions ranged from 47.4% for anxiety to 59.8% for depression.

DISCUSSION

Agreement between the 2 sources varied substantially. Conditions requiring regular management in primary care settings may have a higher agreement than those diagnosed and treated in specialty care.

CONCLUSION

Relying on EHR encounter data to identify chronic conditions without reference to patient problem lists may under-capture conditions among CHC patients in the United States.

The Utilization and Costs of Grade D USPSTF Services in Medicare, 2007-2016

BACKGROUND

Low-value care, or patient care that offers no net benefit in specific clinical scenarios, is costly and often associated with patient harm. The US Preventive Services Task Force (USPSTF) Grade D recommendations represent one of the most scientifically sound and frequently delivered groups of low-value services, but a more contemporary measurement of the utilization and spending for Grade D services beyond the small number of previously studied measures is needed.

OBJECTIVE

To estimate utilization and costs of seven USPSTF Grade D services among US Medicare beneficiaries.

DESIGN

We conducted a cross-sectional study of data from the National Ambulatory Medical Care Survey (NAMCS) from 2007 to 2016 to identify instances of Grade D services.

SETTING/PARTICIPANTS

NAMCS is a nationally representative survey of US ambulatory visits at non-federal and non-hospital-based offices that uses a multistage probability sampling design. We included all visits by Medicare enrollees, which included traditional fee-for-service, Medicare Advantage, supplemental coverage, and dual-eligible Medicare-Medicaid enrollees.

MAIN MEASURES

We measured annual utilization of seven Grade D services among adult Medicare patients, using inclusion and exclusion criteria from prior studies and the USPSTF recommendations. We calculated annual costs by multiplying annual utilization counts by mean per-unit costs of services using publicly available sources.

KEY RESULTS

During the study period, we identified 95,121 unweighted Medicare patient visits, representing approximately 2.4 billion visits. Each year, these seven Grade D services were utilized 31.1 million times for Medicare beneficiaries and cost $477,891,886. Three services-screening for asymptomatic bacteriuria, vitamin D supplements for fracture prevention, and colorectal cancer screening for adults over 85 years-comprised $322,382,772, or two-thirds of the annual costs of the Grade D services measured in this study.

CONCLUSIONS

US Medicare beneficiaries frequently received a group of rigorously defined and costly low-value preventive services. Spending on low-value preventive care concentrated among a small subset of measures, representing important opportunities to safely lower US health care spending while improving the quality of care.

Enhanced access to healthcare utilization data through medical record review: Lessons learned from a community-based, multi-site project

Collecting accurate healthcare utilization (HCU) data on community-based interventions is essential to establishing their clinical effectiveness and cost-related impact. Strategies used to enhance receiving medical records for HCU data extraction in a multi-site longitudinal randomized control trial with urban adolescents are presented. Successful strategies included timely assessment of procedures and practice preferences for access to electronic health records and hardcopy medical charts. Repeated outreach to clinical practice sites to identify and accommodate their preferred procedure for medical record release and flexibility in obtaining chart information helped achieve a 75% success rate in this study. Maintaining participant contact, updating provider information, and continuously evaluating site-specific personnel needs are recommended.

The bird's-eye view: A data-driven approach to understanding patient journeys from claims data

OBJECTIVE

In preference-sensitive conditions such as back pain, there can be high levels of variability in the trajectory of patient care. We sought to develop a methodology that extracts a realistic and comprehensive understanding of the patient journey using medical and pharmaceutical insurance claims data.

MATERIALS AND METHODS

We processed a sample of 10 000 patient episodes (comprised of 113 215 back pain-related claims) into strings of characters, where each letter corresponds to a distinct encounter with the healthcare system. We customized the Levenshtein edit distance algorithm to evaluate the level of similarity between each pair of episodes based on both their content (types of events) and ordering (sequence of events). We then used clustering to extract the main variations of the patient journey.

RESULTS

The algorithm resulted in 12 comprehensive and clinically distinct patterns (clusters) of patient journeys that represent the main ways patients are diagnosed and treated for back pain. We further characterized demographic and utilization metrics for each cluster and observed clear differentiation between the clusters in terms of both clinical content and patient characteristics.

DISCUSSION

Despite being a complex and often noisy data source, administrative claims provide a unique longitudinal overview of patient care across multiple service providers and locations. This methodology leverages claims to capture a data-driven understanding of how patients traverse the healthcare system.

CONCLUSIONS

When tailored to various conditions and patient settings, this methodology can provide accurate overviews of patient journeys and facilitate a shift toward high-quality practice patterns.

Evaluating the effect of Medicaid expansion on access to preventive reproductive care for women in Oregon

We evaluated the effect of the Affordable Care Act (ACA) Medicaid expansion on receipt of preventive reproductive services for women in Oregon. First, we compared service receipt among continuing Medicaid enrollees pre- and post-ACA. We then compared receipt among new post-ACA Medicaid enrollees to receipt by continuing enrollees after ACA implementation. Using Medicaid enrollment and claims data, we identified well-woman visits, contraceptive counseling, contraceptive services, sexually transmitted infection (STI) screening, and cervical cancer screening among women ages 15-44 in years when not pregnant. For pre-ACA enrollees, we assessed pre-ACA receipt in 2011-2013 (n = 83,719) and post-ACA receipt in 2014-2016 (n = 103,225). For post-ACA enrollees we similarly assessed post-ACA service receipt (n = 73,945) and compared this to service receipt by pre-ACA enrollees during 2014-2016. We estimated logistic regression models to compare service receipt over time and between enrollment groups. Among pre-ACA enrollees we found lower receipt of all services post-ACA. Adjusted declines ranged from 7.0 percentage points (95% CI: -7.5, -6.4) for cervical cancer screening to 0.4 percentage points [-0.6, -0.2] for STI screening. In 2014-2016, post-ACA enrollees differed significantly from pre-ACA enrollees in receipt of all services, but all differences were <2 percentage points. Despite small declines in receipt of several preventive reproductive services among prior enrollees, the ACA resulted in Medicaid financing of these services for a large number of newly enrolled low-income women in Oregon, which may eventually lead to population-level improvements in reproductive health. These findings among women in Oregon could inform Medicaid coverage efforts in other states.

Novel Data Linkages to Characterize Palliative and End-Of-Life Care: Challenges and Considerations

CONTEXT

Working groups have called for linkages of existing and diverse databases to improve quality measurement in palliative and end-of-life (EOL) care, but limited data are available on the challenges of using different data sources to measure such care.

OBJECTIVES

To assess concordance of data obtained from different sources in a novel linkage of death certificates, electronic health records (EHRs), cancer registry data, and insurance claims for patients who died with cancer.

METHODS

We joined a database of Washington State death certificates and EHR to a data repository of commercial health plan enrollment and claims files linked to registry records from Puget Sound Cancer Surveillance System. We assessed care in the last month including hospitalizations, intensive care unit (ICU) admissions, emergency department visits, imaging scans, radiation, and hospice, plus chemotherapy in the last 14 days. We used a Chi-squared test to compare differences between health care in EHR and claims.

RESULTS

Records of hospitalization, ICU use, and emergency department use were 33%, 15%, and 33% lower in EHR versus claims, respectively. Radiation, hospice, and imaging were 6%, 14%, and 28% lower, respectively, in EHR, but chemotherapy was 4% higher than that in claims. These differences were statistically different for hospice (P < 0.02), hospitalization, ICU, ER, and imaging (all P < 0.01) but not radiation (P = 0.12) or chemotherapy (P = 0.29).

CONCLUSION

We found substantial variation between EHR and claims for EOL health-care use. Reliance on EHR will miss some health-care use, while claims will not capture the complex clinical details in EHR that can help define the quality of palliative care and EOL health-care utilization.

Use of a Preventive Index to Examine Clinic-Level Factors Associated With Delivery of Preventive Care

INTRODUCTION

There is an increasing need for the development of new methods to understand factors affecting delivery of preventive care. This study applies a new measurement approach and assesses clinic-level factors associated with preventive care delivery.

METHODS

This retrospective longitudinal cohort study of 94 community health centers used electronic health record data from the OCHIN community health information network, 2014-2015. Clinic-level preventive ratios (time covered by a preventive service/time eligible for a preventive service) were calculated in 2017 for 12 preventive services with A or B recommendations from the U.S. Preventive Services Task Force along with an aggregate preventive index for all services combined. For each service, multivariable negative binomial regression modeling and calculated rate ratios assessed the association between clinic-level variables and delivery of care.

RESULTS

Of ambulatory community health center visits, 59.8% were Medicaid-insured and 10.4% were uninsured. Ambulatory community health centers served 16.9% patients who were Hispanic, 13.1% who were nonwhite, and 68.7% who had household incomes <138% of the federal poverty line. Clinic-level preventive ratios ranged from 3% (hepatitis C screening) to 93% (blood pressure screening). The aggregate preventive index including all screening measures was 47% (IQR, 42%-50%). At the clinic level, having a higher percentage of uninsured visits was associated with lower preventive ratios for most (7 of 12) preventive services.

CONCLUSIONS

Approaches that use individual preventive ratios and aggregate prevention indices are promising for understanding and improving preventive service delivery over time. Health insurance remains strongly associated with access to needed preventive care, even for safety net clinic populations.

Study protocol: a mixed-methods study of women's healthcare in the safety net after Affordable Care Act implementation - EVERYWOMAN

Background

Evidence-based reproductive care reduces morbidity and mortality for women and their children, decreases health disparities and saves money. Community health centres (CHCs) are a key point of access to reproductive and primary care services for women who are publicly insured, uninsured or unable to pay for care. Women of reproductive age (15–44 years) comprise just of a quarter (26%) of the total CHC patient population, with higher than average proportions of women of colour, women with lower income and educational status and social challenges (e.g. housing). Such factors are associated with poorer reproductive health outcomes across contraceptive, preventive and pregnancy-related services. The Affordable Care Act (ACA) prioritised reproductive health as an essential component of women’s preventive services to counter these barriers and increase women’s access to care. In 2012, the United States Supreme Court ruled ACA implementation through Medicaid expansion as optional, creating a natural experiment to measure the ACA’s impact on women’s reproductive care delivery and health outcomes.

Methods

This paper describes a 5-year, mixed-methods study comparing women’s contraceptive, preventive, prenatal and postpartum care before and after ACA implementation and between Medicaid expansion and non-expansion states. Quantitative assessment will leverage electronic health record data from the ADVANCE Clinical Research Network, a network of over 130 CHCs in 24 states, to describe care and identify patient, practice and state-level factors associated with provision of recommended evidence-based care. Qualitative assessment will include patient, provider and practice level interviews to understand perceptions and utilisation of reproductive healthcare in CHC settings.

Discussion

To our knowledge, this will be the first study using patient level electronic health record data from multiple states to assess the impact of ACA implementation in conjunction with other practice and policy level factors such as Title X funding or 1115 Medicaid waivers. Findings will be relevant to policy and practice, informing efforts to enhance the provision of timely, evidence-based reproductive care, improve health outcomes and reduce disparities among women. Patient, provider and practice-level interviews will serve to contextualise our findings and develop subsequent studies and interventions to support women’s healthcare provision in CHC settings.

Exploration of Functional Limitation Codes for Outpatient Physical Therapy in the Medicare Population: A Retrospective Cohort Study

BACKGROUND

The Centers for Medicare & Medicaid Services (CMS) introduced functional limitation reporting (FLR) to capture patient progress in functional status in outpatient rehabilitation settings. FLR along with the severity modifier (SM) measure the effectiveness of the rehabilitation services at the physical therapist evaluation (initial examination [IE]) after 10 days of therapy and at discharge.

OBJECTIVE

The objective of this study was to explore the completeness of FLR codes and describe changes in SMs at scheduled checkpoints for patients receiving outpatient physical therapy.

DESIGN

The design was a retrospective cohort descriptive study.

METHODS

A 5% random sample of 2014 Part B fee-for-service Medicare claims for outpatient physical therapy was used. FLR codes with SMs were analyzed at required periods. The number of claims with FLR codes and SMs was calculated to describe the completeness for each period. Planned changes in SMs at the physical therapist IE (current status and projected goal status) and differences in SMs from the physical therapist IE (current status) to discharge (discharge status) were calculated.

RESULTS

For 114,588 beneficiaries, 166,572 physical therapist IE and 130,117 discharge claims were analyzed. Completion was greater than 90% for current status and projected goal status FLR codes at the physical therapist IE but was markedly lower for interim and discharge reporting (≤ 50% for all). More than 75% of claims had planned improvements in SMs at the physical therapist IE (projected goal status - current status), with variations by specific FLR codes. For the episodes with FLR codes at the physical therapist IE and discharge, improvements were reported in more than 2 of 3 episodes.

LIMITATIONS

Limitations for these analyses include a missing discharge claim on many outpatient physical therapy episodes and potential coding errors with Medicare claims.

CONCLUSIONS

Except for the physical therapist IE, FLR codes were not submitted consistent with regulations. Most physical therapy episodes showed improvements in FLR SMs from the physical therapist IE and discharge, although the low completion rate limited interpretation. Changes to the FLR program are warranted to understand whether changes in SMs correspond to changes in a patient's function.

Measurement Error and Misclassification in Electronic Medical Records: Methods to Mitigate Bias

PURPOSE OF REVIEW

We sought to: 1) examine common sources of measurement error in research using data from electronic medical records (EMR), 2) discuss methods to assess the extent and type of measurement error, and 3) describe recent developments in methods to address this source of bias.

RECENT FINDINGS

We identified eight sources of measurement error frequently encountered in EMR studies, the most prominent being that EMR data usually reflect only the health services and medications delivered within the specific health facility/system contributing to the EMR data. Methods for assessing measurement error in EMR data usually require gold standard or validation data, which may be possible using data linkage. Recent methodological developments to address the impact of measurement error in EMR analyses were particularly rich in the multiple imputation literature.

SUMMARY

Presently, sources of measurement error impacting EMR studies are still being elucidated, as are methods for assessing and addressing them. Given the magnitude of measurement error that has been reported, investigators are urged to carefully evaluate and rigorously address this potential source of bias in studies based in EMR data.

Leveraging electronic health records for clinical research

Electronic health records (EHRs) can be a major tool in the quest to decrease costs and timelines of clinical trial research, generate better evidence for clinical decision making, and advance health care. Over the past decade, EHRs have increasingly offered opportunities to speed up, streamline, and enhance clinical research. EHRs offer a wide range of possible uses in clinical trials, including assisting with prestudy feasibility assessment, patient recruitment, and data capture in care delivery. To fully appreciate these opportunities, health care stakeholders must come together to face critical challenges in leveraging EHR data, including data quality/completeness, information security, stakeholder engagement, and increasing the scale of research infrastructure and related governance. Leaders from academia, government, industry, and professional societies representing patient, provider, researcher, industry, and regulator perspectives convened the Leveraging EHR for Clinical Research Now! Think Tank in Washington, DC (February 18-19, 2016), to identify barriers to using EHRs in clinical research and to generate potential solutions. Think tank members identified a broad range of issues surrounding the use of EHRs in research and proposed a variety of solutions. Recognizing the challenges, the participants identified the urgent need to look more deeply at previous efforts to use these data, share lessons learned, and develop a multidisciplinary agenda for best practices for using EHRs in clinical research. We report the proceedings from this think tank meeting in the following paper.

Medicaid coverage accuracy in electronic health records

Health insurance coverage facilitates access to preventive screenings and other essential health care services, and is linked to improved health outcomes; therefore, it is critical to understand how well coverage information is documented in the electronic health record (EHR) and which characteristics are associated with accurate documentation. Our objective was to evaluate the validity of EHR data for monitoring longitudinal Medicaid coverage and assess variation by patient demographics, visit types, and clinic characteristics. We conducted a retrospective, observational study comparing Medicaid status agreement between Oregon community health center EHR data linked at the patient-level to Medicaid enrollment data (gold standard). We included adult patients with a Medicaid identification number and ≥1 clinic visit between 1/1/2013-12/31/2014 [>1 million visits (n = 135,514 patients)]. We estimated statistical correspondence between EHR and Medicaid data at each visit (visit-level) and for different insurance cohorts over time (patient-level). Data were collected in 2016 and analyzed 2017-2018. We observed excellent agreement between EHR and Medicaid data for health insurance information: kappa (>0.80), sensitivity (>0.80), and specificity (>0.85). Several characteristics were associated with agreement; at the visit-level, agreement was lower for patients who preferred a non-English language and for visits missing income information. At the patient-level, agreement was lower for black patients and higher for older patients seen in primary care community health centers. Community health center EHR data are a valid source of Medicaid coverage information. Agreement varied with several characteristics, something researchers and clinic staff should consider when using health insurance information from EHR data.

Screening for Latent Tuberculosis Infection Among HIV-Infected Medicaid Enrollees

OBJECTIVES

In the United States, universal screening for latent tuberculosis (TB) infection among people with HIV is recommended, but the percentage receiving screening is unknown. This study assessed screening for latent TB infection among people with HIV enrolled in Medicaid during 2006-2010.

METHODS

We used nationwide fee-for-service Medicaid records to identify people with HIV, measure screening for latent TB infection, and examine associated demographic, social, and clinical factors. We used logistic regression analysis to calculate odds ratios (ORs) and 95% confidence intervals (CIs). We created 2 multivariate models to prevent collinearity between variables for length of HIV infection.

RESULTS

Of 152 831 people with HIV, 26 239 (17.2%) were screened for latent TB infection. The factor most strongly associated with screening was TB exposure or suspected TB (OR = 3.78; 95% CI, 3.27-4.37). Significant demographic characteristics associated with screening included being African American (OR = 1.28; 95% CI, 1.24-1.32) or ≤20 years of age (OR = 1.35; 95% CI, 1.28-1.42). Significant clinical and social factors associated with screening included poor housing conditions, low body mass index, chemotherapy treatment, and use of certain substances (ORs ranged from 1.24 [95% CI, 1.20-1.27] to 1.47 [95% CI, 1.22-1.76]). The screening rate for latent TB infection was higher among people with newly diagnosed HIV infection than among those with established infection (OR = 1.37; 95% CI, 1.32-1.41) and among people with a longer established HIV infection than among those with shorter HIV infection (OR = 1.24; 95% CI, 1.23-1.26 for each additional year).

CONCLUSION

Screening for latent TB infection among fee-for-service Medicaid beneficiaries with HIV was suboptimal, despite the presence of demographic, social, or clinical characteristics that should have increased the likelihood of screening. The lack of certain data in Medicaid may have resulted in an underestimation of screening.

The Pediatric Emergency Care Applied Research Network Registry: A Multicenter Electronic Health Record Registry of Pediatric Emergency Care

BACKGROUND

Electronic health record (EHR)-based registries allow for robust data to be derived directly from the patient clinical record and can provide important information about processes of care delivery and patient health outcomes.

METHODS

A data dictionary, and subsequent data model, were developed describing EHR data sources to include all processes of care within the emergency department (ED). ED visit data were deidentified and XML files were created and submitted to a central data coordinating center for inclusion in the registry. Automated data quality control occurred prior to submission through an application created for this project. Data quality reports were created for manual data quality review.

RESULTS

The Pediatric Emergency Care Applied Research Network (PECARN) Registry, representing four hospital systems and seven EDs, demonstrates that ED data from disparate health systems and EHR vendors can be harmonized for use in a single registry with a common data model. The current PECARN Registry represents data from 2,019,461 pediatric ED visits, 894,503 distinct patients, more than 12.5 million narrative reports, and 12,469,754 laboratory tests and continues to accrue data monthly.

CONCLUSION

The Registry is a robust harmonized clinical registry that includes data from diverse patients, sites, and EHR vendors derived via data extraction, deidentification, and secure submission to a central data coordinating center. The data provided may be used for benchmarking, clinical quality improvement, and comparative effectiveness research.

Use of a prescription opioid registry to examine opioid misuse and overdose in an integrated health system

Strategies are needed to identify at-risk patients for adverse events associated with prescription opioids. This study identified prescription opioid misuse in an integrated health system using electronic health record (EHR) data, and examined predictors of misuse and overdose. The sample included patients from an EHR-based registry of adults who used prescription opioids in 2011 in Kaiser Permanente Northern California, a large integrated health care system. We characterized time-at-risk for opioid misuse and overdose, and used Cox proportional hazard models to model predictors of these events from 2011 to 2014. Among 396,452 patients, 2.7% were identified with opioid misuse and 1044 had an overdose event. Older patients were less likely to meet misuse criteria or have an overdose. Whites were more likely to be identified with misuse, but not to have an overdose. Alcohol and drug disorders were related to higher risk of misuse and overdose, with the exception that marijuana disorder was not related to opioid misuse. Higher daily opioid dosages and benzodiazepine use increased the risk of both opioid misuse and overdose. We characterized several risk factors associated with misuse and overdose using EHR-based data, which can be leveraged relatively quickly to inform preventive strategies to address the opioid crisis.

Biases introduced by filtering electronic health records for patients with "complete data"

Objective

One promise of nationwide adoption of electronic health records (EHRs) is the availability of data for large-scale clinical research studies. However, because the same patient could be treated at multiple health care institutions, data from only a single site might not contain the complete medical history for that patient, meaning that critical events could be missing. In this study, we evaluate how simple heuristic checks for data "completeness" affect the number of patients in the resulting cohort and introduce potential biases.

Materials and Methods

We began with a set of 16 filters that check for the presence of demographics, laboratory tests, and other types of data, and then systematically applied all 216 possible combinations of these filters to the EHR data for 12 million patients at 7 health care systems and a separate payor claims database of 7 million members.

Results

EHR data showed considerable variability in data completeness across sites and high correlation between data types. For example, the fraction of patients with diagnoses increased from 35.0% in all patients to 90.9% in those with at least 1 medication. An unrelated claims dataset independently showed that most filters select members who are older and more likely female and can eliminate large portions of the population whose data are actually complete.

Discussion and Conclusion

As investigators design studies, they need to balance their confidence in the completeness of the data with the effects of placing requirements on the data on the resulting patient cohort.

Comparison of Electronic Health Record-Based and Claims-Based Diabetes Care Quality Measures: Causes of Discrepancies

OBJECTIVE

To investigate magnitude and sources of discrepancy in quality metrics using claims versus electronic health record (EHR) data.

STUDY DESIGN

Assessment of proportions of HbA1c and LDL testing for people ascertained as diabetic from the respective sources. Qualitative interviews and review of EHRs of discrepant cases.

DATA COLLECTION/EXTRACTION

Claims submitted to Rhode Island Medicaid by three practice sites in 2013; program-coded EHR extraction; manual review of selected EHRs.

PRINCIPAL FINDINGS

Of 21,030 adult Medicaid beneficiaries attributed to a primary care patient at a site by claims or EHR data, concordance on assignment ranged from 0.30 to 0.41. Of patients with concordant assignment, the ratio of patients ascertained as diabetic by EHR versus claims ranged from 1.06 to 1.14. For patients with concordant assignment and diagnosis, the ratio based on EHR versus claims ranged from 1.08 to 18.34 for HbA1c testing, and from 1.29 to 14.18 for lipid testing. Manual record review of 264 patients discrepant on diagnosis or testing identified problems such as misuse of ICD-9 codes, failure to submit claims, and others.

CONCLUSIONS

Claims data underestimate performance on these metrics compared to EHR documentation, by varying amounts. Use of claims data for these metrics is problematic.

Medicaid's Impact on Chronic Disease Biomarkers: A Cohort Study of Community Health Center Patients

BACKGROUND

Understanding the impact of health insurance is critical, particularly in the era of Affordable Care Act Medicaid expansion. The electronic health record (EHR) provides new opportunities to quantify health outcomes.

OBJECTIVE

To assess changes in biomarkers of chronic disease among community health center (CHC) patients who gained Medicaid coverage with the Oregon Medicaid expansion (2008-2011).

DESIGN

Prospective cohort. Patients were followed for 24 months, and rate of mean biomarker change was calculated. Time to a controlled follow-up measurement was compared using Cox regression models.

SETTING/PATIENTS

Using EHR data from OCHIN (a non-profit network of CHCs) linked to state Medicaid data, we identified three cohorts of patients with uncontrolled chronic conditions (diabetes, hypertension, and hyperlipidemia). Within these cohorts, we included patients who gained Medicaid coverage along with a propensity score-matched comparison group who remained uninsured (diabetes n = 608; hypertension n = 1244; hyperlipidemia n = 546).

MAIN MEASURES

Hemoglobin A1c (HbA1c) for the diabetes cohort, systolic and diastolic blood pressure (SBP and DBP, respectively) for the hypertension cohort, and low-density lipoprotein (LDL) for the hyperlipidemia cohort.

KEY RESULTS

All cohorts improved over time. Compared to matched uninsured patients, adults in the diabetes and hypertension cohorts who gained Medicaid coverage were significantly more likely to have a follow-up controlled measurement (hazard ratio [HR] =1.26, p = 0.020; HR = 1.35, p < 0.001, respectively). No significant difference was observed in the hyperlipidemia cohort (HR = 1.09, p = 0.392).

CONCLUSIONS

OCHIN patients with uncontrolled chronic conditions experienced objective health improvements over time. In two of three chronic disease cohorts, those who gained Medicaid coverage were more likely to achieve a controlled measurement than those who remained uninsured. These findings demonstrate the effective care provided by CHCs and the importance of health insurance coverage within a usual source of care setting.

CLINICAL TRIALS REGISTRATION

NCT02355132 [ https://clinicaltrials.gov/ct2/show/NCT02355132 ].

Assessing Community Cancer care after insurance ExpanSionS (ACCESS) study protocol

Background

Cancer is the second most common cause of mortality in the United States. Cancer screening and prevention services have contributed to improved overall cancer survival rates in the past 40 years. Vulnerable populations (i.e., uninsured, low-income, and racial/ethnic minorities) are disproportionately affected by cancer, receive significantly fewer cancer prevention services, poorer healthcare, and subsequently lower survival rates than insured, white, non-Hispanic populations. The Affordable Care Act (ACA) aims to provide health insurance to all low-income citizens and legal residents, including an expansion of Medicaid eligibility for those earning ≤138% of federal poverty level. As of 2012, Medicaid was expanded in 32 states and the District of Columbia, while 18 states did not expand, creating a ‘natural experiment’ to assess the impact of Medicaid expansion on cancer prevention and care.

Methods

We will use electronic health record data from up to 990 community health centers available up to 24-months before and at least one year after Medicaid expansion. Primary outcomes include health insurance and coverage status, and type of insurance. Additional outcomes include healthcare delivery, number and types of encounters, and receipt of cancer prevention and screening for all patients and preventive care and screening services for cancer survivors.

Discussion

Cancer morbidity and mortality is greatly reduced through screening and prevention, but uninsured patients are much less likely than insured patients to receive these services as recommended. This natural policy experiment will provide valuable information about cancer-related healthcare services as the US tackles the distribution of healthcare resources and future health reform.

Trial Registration

Clinicaltrails.gov identifier NCT02936609.

Lack of Lipid Screening Disparities in Obese Latino Adults at Health Centers

INTRODUCTION

In cross-sectional survey studies, obese Latinos are less likely to be screened for elevated serum cholesterol, despite their higher risk for hyperlipidemia and coronary artery disease. This study evaluated insurance and racial/ethnic disparities in lipid screening receipt between obese Latino and non-Hispanic white patients in Oregon community health centers (CHCs) over 5 years, using electronic health record data.

METHODS

This retrospective cohort study evaluated obese (BMI ≥30), low-income, adult patients (aged 21-79 years) with at least one visit at an Oregon CHC during 2009-2013 (n=11,095). Odds of lipid screening in the study period (clinical data collected in 2009-2013) were measured, adjusting for age, sex, primary clinic, and comorbidities, stratified by utilization in the study period. Analysis was done in 2016.

RESULTS

Sixty percent of the study population received at least one lipid screening in 2009-2013. There were no significant differences in screening between insured Latinos and insured non-Hispanic whites, except those with more than five visits over 5 years (OR=0.75, 95% CI=0.60, 0.94). Uninsured Latinos had higher odds of screening versus insured non-Hispanic whites among the low visit strata (OR=1.65, 95% CI=1.18, 2.30). Among Latinos, Spanish preference resulted in higher screening odds versus English preference in the two- to five-visit stratum (OR=1.63, 95% CI=1.12, 2.35).

CONCLUSIONS

Obese, low-income patients at CHCs underutilize cholesterol screening. However, screening differences by race/ethnicity and preferred language are minimal. Further research is necessary to understand how care delivered by CHCs may mitigate previously reported disparities in lipid screening.

For Diabetes Shared Savings Programs, 1 Year of Data Is Not Enough

Fee-for-service payment models are moving toward pay-for-performance designs, many of which rely on shared savings for financial sustainability. Shared savings programs divide the cost savings between health care purchaser and provider based on provider performance. Often, these programs measure provider performance as the delivery of agreed-upon clinical practice guidelines that usually are represented as evidence-based medicine (EBM). Multiyear studies show a negative relationship between total cost and EBM, indicating that long-term shared savings can be substantial. This study explores expectations for the rewards in the first year of a shared savings program. It also indicates the effectiveness of using 1 year of claims to assess cost savings from evidence-based care, especially in a patient population with high turnover. This study analyzed 1956 adults with diabetes insured through Medicaid. Results of linear regression showed that the relationship between total cost of care and each element of evidence-based medical care during a 1-year period was positive (higher cost) or insignificant. The results indicate that diabetes EBM programs cannot expect to see significant cost savings if the evaluation lasts only 1 year or less. The study concludes that improvements in EBM incentive programs could come from investigating the length of time needed to realize cost savings from each element of diabetes EBM. Investigating other factors that could affect the expected amount of cost savings also would benefit these programs, especially factors derived from sources external to insurance program information such as the medical record and care management data.

Identification of Long Bone Fractures in Radiology Reports Using Natural Language Processing to support Healthcare Quality Improvement

BACKGROUND

Important information to support healthcare quality improvement is often recorded in free text documents such as radiology reports. Natural language processing (NLP) methods may help extract this information, but these methods have rarely been applied outside the research laboratories where they were developed.

OBJECTIVE

To implement and validate NLP tools to identify long bone fractures for pediatric emergency medicine quality improvement.

METHODS

Using freely available statistical software packages, we implemented NLP methods to identify long bone fractures from radiology reports. A sample of 1,000 radiology reports was used to construct three candidate classification models. A test set of 500 reports was used to validate the model performance. Blinded manual review of radiology reports by two independent physicians provided the reference standard. Each radiology report was segmented and word stem and bigram features were constructed. Common English "stop words" and rare features were excluded. We used 10-fold cross-validation to select optimal configuration parameters for each model. Accuracy, recall, precision and the F1 score were calculated. The final model was compared to the use of diagnosis codes for the identification of patients with long bone fractures.

RESULTS

There were 329 unique word stems and 344 bigrams in the training documents. A support vector machine classifier with Gaussian kernel performed best on the test set with accuracy=0.958, recall=0.969, precision=0.940, and F1 score=0.954. Optimal parameters for this model were cost=4 and gamma=0.005. The three classification models that we tested all performed better than diagnosis codes in terms of accuracy, precision, and F1 score (diagnosis code accuracy=0.932, recall=0.960, precision=0.896, and F1 score=0.927).

CONCLUSIONS

NLP methods using a corpus of 1,000 training documents accurately identified acute long bone fractures from radiology reports. Strategic use of straightforward NLP methods, implemented with freely available software, offers quality improvement teams new opportunities to extract information from narrative documents.

Measuring Preventive Care Delivery: Comparing Rates Across Three Data Sources

INTRODUCTION

Preventive care delivery is an important quality outcome, and electronic data reports are being used increasingly to track these services. It is highly informative when electronic data sources are compared to information manually extracted from medical charts to assess validity and completeness.

METHODS

This cross-sectional study used a random sample of Medicaid-insured patients seen at 43 community health centers in 2011 to calculate standard measures of correspondence between manual chart review and two automated sources (electronic health records [EHRs] and Medicaid claims), comparing documentation of orders for and receipt of ten preventive services (n=150 patients/service). Data were analyzed in 2015.

RESULTS

Using manual chart review as the gold standard, automated EHR extraction showed near-perfect to perfect agreement (κ=0.96-1.0) for services received within the primary care setting (e.g., BMI, blood pressure). Receipt of breast and colorectal cancer screenings, services commonly referred out, showed moderate (κ=0.42) to substantial (κ=0.62) agreement, respectively. Automated EHR extraction showed near-perfect agreement (κ=0.83-0.97) for documentation of ordered services. Medicaid claims showed near-perfect agreement (κ=0.87) for hyperlipidemia and diabetes screening, and substantial agreement (κ=0.67-0.80) for receipt of breast, cervical, and colorectal cancer screenings, and influenza vaccination. Claims showed moderate agreement (κ=0.59) for chlamydia screening receipt. Medicaid claims did not capture ordered or unbilled services.

CONCLUSIONS

Findings suggest that automated EHR and claims data provide valid sources for measuring receipt of most preventive services; however, ordered and unbilled services were primarily captured via EHR data and completed referrals were more often documented in claims data.

Effect of Gaining Insurance Coverage on Smoking Cessation in Community Health Centers: A Cohort Study

BACKGROUND

Community health center (CHC) patients have high rates of smoking. Insurance coverage for smoking cessation assistance, such as that mandated by the Affordable Care Act, may aid in smoking cessation in this vulnerable population.

OBJECTIVE

We aimed to determine if uninsured CHC patients who gain Medicaid coverage experience greater primary care utilization, receive more cessation medication orders, and achieve higher quit rates, compared to continuously uninsured smokers.

DESIGN

Longitudinal observational cohort study using electronic health record data from a network of Oregon CHCs linked to Oregon Medicaid enrollment data.

PATIENTS

Cohort of patients who smoke and who gained Medicaid coverage in 2008-2011 after ≥ 6 months of being uninsured and with ≥ 1 smoking assessment in the 24-month follow-up period from the baseline smoking status date. This group was propensity score matched to a cohort of continuously uninsured CHC patients who smoke (n = 4140 matched pairs; 8280 patients).

INTERVENTION

Gaining Medicaid after being uninsured for ≥ 6 months.

MAIN MEASURES

'Quit' smoking status (baseline smoking status was 'current every day' or 'some day' and status change to 'former smoker' at a subsequent visit), smoking cessation medication order, and ≥ 6 documented visits (yes/no variables) at ≥ 1 smoking status assessment within the 24-month follow-up period.

KEY RESULTS

The newly insured had 40 % increased odds of quitting smoking (aOR = 1.40, 95 % CI:1.24, 1.58), nearly triple the odds of having a medication ordered (aOR = 2.94, 95 % CI:2.61, 3.32), and over twice the odds of having ≥ 6 follow-up visits (aOR = 2.12, 95 % CI:1.94, 2.32) compared to their uninsured counterparts.

CONCLUSIONS

Newly insured patients had increased odds of quit smoking status over 24 months of follow-up than those who remained uninsured. Providing insurance coverage to vulnerable populations may have a significant impact on smoking cessation.

Receipt of Preventive Services After Oregon's Randomized Medicaid Experiment

INTRODUCTION

It is predicted that gaining health insurance via the Affordable Care Act will result in increased rates of preventive health services receipt in the U.S., primarily based on self-reported findings from previous health insurance expansion studies. This study examined the long-term (36-month) impact of Oregon's 2008 randomized Medicaid expansion ("Oregon Experiment") on receipt of 12 preventive care services in community health centers using electronic health record data.

METHODS

Demographic data from adult (aged 19-64 years) Oregon Experiment participants were probabilistically matched to electronic health record data from 49 Oregon community health centers within the OCHIN community health information network (N=10,643). Intent-to-treat analyses compared receipt of preventive services over a 36-month (2008-2011) period among those randomly assigned to apply for Medicaid versus not assigned, and instrumental variable analyses estimated the effect of actually gaining Medicaid coverage on preventive services receipt (data collected in 2012-2014; analysis performed in 2014-2015).

RESULTS

Intent-to-treat analyses revealed statistically significant differences between patients randomly assigned to apply for Medicaid (versus not assigned) for 8 of 12 assessed preventive services. In intent-to-treat analyses, Medicaid coverage significantly increased the odds of receipt of most preventive services (ORs ranging from 1.04 [95% CI=1.02, 1.06] for smoking assessment to 1.27 [95% CI=1.02, 1.57] for mammography).

CONCLUSIONS

Rates of preventive services receipt will likely increase as community health center patients gain insurance through Affordable Care Act expansions. Continued effort is needed to increase health insurance coverage in an effort to decrease health disparities in vulnerable populations.

Community Health Center Use After Oregon's Randomized Medicaid Experiment

PURPOSE

There is debate about whether community health centers (CHCs) will experience increased demand from patients gaining coverage through Affordable Care Act Medicaid expansions. To better understand the effect of new Medicaid coverage on CHC use over time, we studied Oregon's 2008 randomized Medicaid expansion (the "Oregon Experiment").

METHODS

We probabilistically matched demographic data from adults (aged 19-64 years) participating in the Oregon Experiment to electronic health record data from 108 Oregon CHCs within the OCHIN community health information network (originally the Oregon Community Health Information Network) (N = 34,849). We performed intent-to-treat analyses using zero-inflated Poisson regression models to compare 36-month (2008-2011) usage rates among those selected to apply for Medicaid vs not selected, and instrumental variable analyses to estimate the effect of gaining Medicaid coverage on use. Use outcomes included primary care visits, behavioral/mental health visits, laboratory tests, referrals, immunizations, and imaging.

RESULTS

The intent-to-treat analyses revealed statistically significant differences in rates of behavioral/mental health visits, referrals, and imaging between patients randomly selected to apply for Medicaid vs those not selected. In instrumental variable analyses, gaining Medicaid coverage significantly increased the rate of primary care visits, laboratory tests, referrals, and imaging; rate ratios ranged from 1.27 (95% CI, 1.05-1.55) for laboratory tests to 1.58 (95% CI, 1.10-2.28) for referrals.

CONCLUSIONS

Our results suggest that use of many different types of CHC services will increase as patients gain Medicaid through Affordable Care Act expansions. To maximize access to critical health services, it will be important to ensure that the health care system can support increasing demands by providing more resources to CHCs and other primary care settings.

Using electronic health record data to evaluate preventive service utilization among uninsured safety net patients

OBJECTIVE

This study compared the preventive service utilization of uninsured patients receiving care at Oregon community health centers (CHCs) in 2008 through 2011 with that of continuously insured patients at the same CHCs in the same period, using electronic health record (EHR) data.

METHODS

We performed a retrospective cohort analysis, using logistic mixed effects regression modeling to calculate odds ratios and rates of preventive service utilization for patients without insurance, or with continuous insurance.

RESULTS

CHCs provided many preventive services to uninsured patients. Uninsured patients were less likely than continuously insured patients to receive 5 of 11 preventive services, ranging from OR 0.52 (95% CI: 0.35-0.77) for mammogram orders to 0.75 (95% CI: 0.66-0.86) for lipid panels. This disparity persisted even in patients who visited the clinic regularly.

CONCLUSION

Lack of insurance is a barrier to preventive service utilization, even in patients who can access care at a CHC. Policymakers in the United States should continue to address this significant prevention disparity.