Supplementing claims data with outpatient laboratory test results to improve confounding adjustment in effectiveness studies of lipid-lowering treatments

Dealing with missing data in laboratory test results used as a baseline covariate: results of multi-hospital cohort studies utilizing a database system contributing to MID-NET® in Japan

BACKGROUND

To evaluate missing data methods applied to laboratory test results used for confounding adjustment, utilizing data from 10 MID-NET®-collaborative hospitals.

METHODS

Using two scenarios, five methods dealing with missing laboratory test results were applied, including three missing data methods (single regression imputation (SRI), multiple imputation (MI), and inverse probability weighted (IPW) method). We compared the point estimates of adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs) between the five methods. Hospital variability in missing data was considered using the hospital-specific approach and overall approach. Confounding adjustment methods were propensity score (PS) weighting, PS matching, and regression adjustment.

RESULTS

In Scenario 1, the risk of diabetes due to second-generation antipsychotics was compared with that due to first-generation antipsychotics. The aHR adjusted by PS weighting using SRI, MI, and IPW by the hospital-specific-approach was 0.61 [95%CI, 0.39-0.96], 0.63 [95%CI, 0.42-0.93], and 0.76 [95%CI, 0.46-1.25], respectively. In Scenario 2, the risk of liver injuries due to rosuvastatin was compared with that due to atorvastatin. Although PS matching largely contributed to differences in aHRs between methods, PS weighting provided no substantial difference in point estimates of aHRs between SRI and MI, similar to Scenario 1. The results of SRI and MI in both scenarios showed no considerable changes, even upon changing the approaches considering hospital variations.

CONCLUSIONS

SRI and MI provide similar point estimates of aHR. Two approaches considering hospital variations did not markedly affect the results. Adjustment by PS matching should be used carefully.

Factors associated with higher hemoglobin A1c and type 2 diabetes-related costs: Secondary data analysis of adults 18 to 64 in Texas with commercial insurance

OBJECTIVE

This study will identify factors associated with higher hemoglobin A1c (A1c) values and diabetes-related costs among commercially insured adults in Texas diagnosed with type 2 diabetes.

RESEARCH DESIGN AND METHODS

This secondary data analysis was based on claims data from commercially insured individuals 18-64 years of age residing in Texas with diagnosed type 2 diabetes during the 2018-2019 study period. The final analysis sample after all the exclusions consisted of 34,992 individuals. Measures included hemoglobin A1c, diabetes-related costs, Charlson Comorbidity Index, diabetes-related complications, rurality and other socioeconomic characteristics. Longitudinal A1c measurements were modeled using age, sex, rurality, comorbidity, and diabetes-related complications in generalized linear longitudinal regression models adjusting the observation time, which was one of the 8 quarters in 2018 and 2019. The diabetes-related costs were similarly modeled in both univariable and multivariable generalized linear longitudinal regression models adjusting the observation time by calendar quarters and covariates.

RESULTS

The median A1c value was 7, and the median quarterly diabetes-related cost was $120. A positive statistically significant relationship (p = < .0001) was found between A1c levels and diabetes-related costs, although this trend slowed down as A1c levels exceeded 8.0%. Higher A1c values were associated with being male, having diabetes-related complications, and living in rural areas. Higher costs were associated with higher A1c values, older age, and higher Charlson Comorbidity Index scores.

CONCLUSION

The study adds updated analyses of the interrelationships among demographic and geographic factors, clinical indicators, and health-related costs, reinforcing the role of higher A1c values and complications as diabetes-related cost drivers.

Measuring concordance of data sources used for infectious disease research in the USA: a retrospective data analysis

OBJECTIVES

As highlighted by the COVID-19 pandemic, researchers are eager to make use of a wide variety of data sources, both government-sponsored and alternative, to characterise the epidemiology of infectious diseases. The objective of this study is to investigate the strengths and limitations of sources currently being used for research.

DESIGN

Retrospective descriptive analysis.

PRIMARY AND SECONDARY OUTCOME MEASURES

Yearly number of national-level and state-level disease-specific case counts and disease clusters for three diseases (measles, mumps and varicella) during a 5-year study period (2013-2017) across four different data sources: Optum (health insurance billing claims data), HealthMap (online news surveillance data), Morbidity and Mortality Weekly Reports (official government reports) and National Notifiable Disease Surveillance System (government case surveillance data).

RESULTS

Our study demonstrated drastic differences in reported infectious disease incidence across data sources. When compared with the other three sources of interest, Optum data showed substantially higher, implausible standardised case counts for all three diseases. Although there was some concordance in identified state-level case counts and disease clusters, all four sources identified variations in state-level reporting.

CONCLUSIONS

Researchers should consider data source limitations when attempting to characterise the epidemiology of infectious diseases. Some data sources, such as billing claims data, may be unsuitable for epidemiological research within the infectious disease context.

Data Source Concordance for Infectious Disease Epidemiology

Background

As highlighted by the COVID-19 pandemic, researchers are eager to make use of a wide variety of data sources, both government-sponsored and alternative, to characterize the epidemiology of infectious diseases. To date, few studies have investigated the strengths and limitations of sources currently being used for such research. These are critical for policy makers to understand when interpreting study findings.

Methods

To fill this gap in the literature, we compared infectious disease reporting for three diseases (measles, mumps, and varicella) across four different data sources: Optum (health insurance billing claims data), HealthMap (online news surveillance data), Morbidity and Mortality Weekly Reports (official government reports), and National Notifiable Disease Surveillance System (government case surveillance data). We reported the yearly number of national- and state-level disease-specific case counts and disease clusters according to each of our sources during a five-year study period (2013-2017).

Findings

Our study demonstrated drastic differences in reported infectious disease incidence across data sources. When compared against the other three sources of interest, Optum data showed substantially higher, implausible standardized case counts for all three diseases. Although there was some concordance in identified state-level case counts and disease clusters, all four sources identified variations in state-level reporting.

Interpretation

Researchers should consider data source limitations when attempting to characterize the epidemiology of infectious diseases. Some data sources, such as billing claims data, may be unsuitable for epidemiological research within the infectious disease context.

Worsening of kidney function in patients with atrial fibrillation and chronic kidney disease: evidence from the real-world CALLIPER study

OBJECTIVE

Evidence is needed on the impact of anticoagulation therapy on kidney function in patients with atrial fibrillation (AF). The objective of this analysis, which is part of the CALLIPER study, was to investigate the risk of worsening kidney function with rivaroxaban 15 mg once daily compared with warfarin in patients with AF and moderate-to-severe chronic kidney disease (CKD) in routine clinical practice in the United States.

METHODS

CALLIPER was an observational, retrospective, new-user cohort study. Adult patients with AF in the US IBM Watson MarketScan databases who newly initiated anticoagulation with rivaroxaban 15 mg once daily or warfarin between January 2013 and December 2017 were included. Comparative analysis was performed using Cox proportional hazards regression after adjustment for potential confounding by the stabilized inverse probability of treatment weighting approach and propensity score matching. One of the main study outcomes was worsening kidney function (composite of progression to CKD stage 5, kidney failure, or need for dialysis), besides traditional AF-related outcomes.

RESULTS

The cohort included 7368 patients: 5903 (80.1%) initiating warfarin and 1465 (19.9%) initiating rivaroxaban 15 mg once daily. Rivaroxaban 15 mg was associated with a significant 47% reduction in the risk of worsening kidney function versus warfarin (hazard ratio 0.53; 95% confidence interval 0.35-0.78). Similar results were observed in the subgroup of patients with type 2 diabetes.

CONCLUSIONS

Rivaroxaban 15 mg may be associated with a lower risk of worsening kidney function as compared with warfarin in the atrial fibrillation population with moderate-to-severe CKD.

TRIAL REGISTRATION NUMBER

NCT03359876.

Establishing a National Cardiovascular Disease Surveillance System in the United States Using Electronic Health Record Data: Key Strengths and Limitations

Cardiovascular disease surveillance involves quantifying the evolving population-level burden of cardiovascular outcomes and risk factors as a data-driven initial step followed by the implementation of interventional strategies designed to alleviate this burden in the target population. Despite widespread acknowledgement of its potential value, a national surveillance system dedicated specifically to cardiovascular disease does not currently exist in the United States. Routinely collected health care data such as from electronic health records (EHRs) are a possible means of achieving national surveillance. Accordingly, this article elaborates on some key strengths and limitations of using EHR data for establishing a national cardiovascular disease surveillance system. Key strengths discussed include the: (1) ubiquity of EHRs and consequent ability to create a more "national" surveillance system, (2) existence of a common data infrastructure underlying the health care enterprise with respect to data domains and the nomenclature by which these data are expressed, (3) longitudinal length and detail that define EHR data when individuals repeatedly patronize a health care organization, and (4) breadth of outcomes capable of being surveilled with EHRs. Key limitations discussed include the: (1) incomplete ascertainment of health information related to health care-seeking behavior and the disconnect of health care data generated at separate health care organizations, (2) suspect data quality resulting from the default information-gathering processes within the clinical enterprise, (3) questionable ability to surveil patients through EHRs in the absence of documented interactions, and (4) the challenge in interpreting temporal trends in health metrics, which can be obscured by changing clinical and administrative processes.

Constructing Epidemiologic Cohorts from Electronic Health Record Data

In the United States, electronic health records (EHR) are increasingly being incorporated into healthcare organizations to document patient health and services rendered. EHRs serve as a vast repository of demographic, diagnostic, procedural, therapeutic, and laboratory test data generated during the routine provision of health care. The appeal of using EHR data for epidemiologic research is clear: EHRs generate large datasets on real-world patient populations in an easily retrievable form permitting the cost-efficient execution of epidemiologic studies on a wide array of topics. Constructing epidemiologic cohorts from EHR data involves as a defining feature the development of data machinery, which transforms raw EHR data into an epidemiologic dataset from which appropriate inference can be drawn. Though data machinery includes many features, the current report focuses on three aspects of machinery development of high salience to EHR-based epidemiology: (1) selecting study participants; (2) defining “baseline” and assembly of baseline characteristics; and (3) follow-up for future outcomes. For each, the defining features and unique challenges with respect to EHR-based epidemiology are discussed. An ongoing example illustrates key points. EHR-based epidemiology will become more prominent as EHR data sources continue to proliferate. Epidemiologists must continue to improve the methods of EHR-based epidemiology given the relevance of EHRs in today’s healthcare ecosystem.

Lab-based and diagnosis-based chronic kidney disease recognition and staging concordance

BACKGROUND

Chronic kidney disease (CKD) is often under-recognized and poorly documented via diagnoses, but the extent of under-recognition is not well understood among Medicare beneficiaries. The current study used claims-based diagnosis and lab data to examine patient factors associated with clinically recognized CKD and CKD stage concordance between claims- and lab-based sources in a cohort of Medicare beneficiaries.

METHODS

In a cohort of fee-for-service (FFS) beneficiaries with CKD based on 2011 labs, we examined the proportion with clinically recognized CKD via diagnoses and factors associated with clinical recognition in logistic regression. In the subset of beneficiaries with CKD stage identified from both labs and diagnoses, we examined concordance in CKD stage from both sources, and factors independently associated with CKD stage concordance in logistic regression.

RESULTS

Among the subset of 206,036 beneficiaries with lab-based CKD, only 11.8% (n = 24,286) had clinically recognized CKD via diagnoses. Clinical recognition was more likely for beneficiaries who had higher CKD stages, were non-elderly, were Hispanic or non-Hispanic Black, lived in core metropolitan areas, had multiple chronic conditions or outpatient visits in 2010, or saw a nephrologist. In the subset of 18,749 beneficiaries with CKD stage identified from both labs and diagnoses, 70.0% had concordant CKD stage, which was more likely if beneficiaries were older adults, male, lived in micropolitan areas instead of non-core areas, or saw a nephrologist.

CONCLUSIONS

There is significant under-diagnosis of CKD in Medicare FFS beneficiaries, which can be addressed with the availability of lab results.

Propensity score prediction for electronic healthcare databases using Super Learner and High-dimensional Propensity Score Methods

The optimal learner for prediction modeling varies depending on the underlying data-generating distribution. Super Learner (SL) is a generic ensemble learning algorithm that uses cross-validation to select among a "library" of candidate prediction models. While SL has been widely studied in a number of settings, it has not been thoroughly evaluated in large electronic healthcare databases that are common in pharmacoepidemiology and comparative effectiveness research. In this study, we applied and evaluated the performance of SL in its ability to predict the propensity score (PS), the conditional probability of treatment assignment given baseline covariates, using three electronic healthcare databases. We considered a library of algorithms that consisted of both nonparametric and parametric models. We also proposed a novel strategy for prediction modeling that combines SL with the high-dimensional propensity score (hdPS) variable selection algorithm. Predictive performance was assessed using three metrics: the negative log-likelihood, area under the curve (AUC), and time complexity. Results showed that the best individual algorithm, in terms of predictive performance, varied across datasets. The SL was able to adapt to the given dataset and optimize predictive performance relative to any individual learner. Combining the SL with the hdPS was the most consistent prediction method and may be promising for PS estimation and prediction modeling in electronic healthcare databases.

Risk of serious infections in tocilizumab versus other biologic drugs in patients with rheumatoid arthritis: a multidatabase cohort study

OBJECTIVE

To investigate the rate of serious bacterial, viral or opportunistic infection in patients with rheumatoid arthritis (RA) starting tocilizumab (TCZ) versus tumour necrosis factor inhibitors (TNFi) or abatacept.

METHODS

Using claims data from US Medicare from 2010 to 2015, and IMS and MarketScan from 2011 to 2015, we identified adults with RA who initiated TCZ or TNFi (primary comparator)/abatacept (secondary comparator) with prior use of ≥1 different biologic drug or tofacitinib. The primary outcome was hospitalised serious infection (SI), including bacterial, viral or opportunistic infection. To control for >70 confounders, TCZ initiators were propensity score (PS)-matched to TNFi or abatacept initiators. Database-specific HRs were combined by a meta-analysis.

RESULTS

The primary cohort included 16 074 TCZ PS-matched to 33 109 TNFi initiators. The risk of composite SI was not different between TCZ and TNFi initiators (combined HR 1.05, 95% CI 0.95 to 1.16). However, TCZ was associated with an increased risk of serious bacterial infection (HR 1.19, 95% CI 1.07 to 1.33), skin and soft tissue infections (HR 2.38, 95% CI 1.47 to 3.86), and diverticulitis (HR 2.34, 95% CI 1.64 to 3.34) versus TNFi. An increased risk of composite SI, serious bacterial infection, diverticulitis, pneumonia/upper respiratory tract infection and septicaemia/bacteraemia was observed in TCZ versus abatacept users.

CONCLUSIONS

This large multidatabase cohort study found no difference in composite SI risk in patients with RA initiating TCZ versus TNFi after failing ≥1 biologic drug or tofacitinib. However, the risk of serious bacterial infection, skin and soft tissue infections, and diverticulitis was higher in TCZ versus TNFi initiators. The risk of composite SI was higher in TCZ initiators versus abatacept.

Measuring prevalence and incidence of chronic conditions in claims and electronic health record databases

Background

Health care databases are natural sources for estimating prevalence and incidence of chronic conditions, but substantial variation in estimates limits their interpretability and utility. We evaluated the effects of design choices when estimating prevalence and incidence in claims and electronic health record databases.

Methods

Prevalence and incidence for five chronic diseases at increasing levels of expected frequencies, from cystic fibrosis to COPD, were estimated in the Clinical Practice Research Datalink (CPRD) and MarketScan databases from 2011 to 2014. Estimates were compared using different definitions of lookback time and contributed person-time.

Results

Variation in lookback time substantially affected estimates. In 2014, for CPRD, use of an all-time vs a 1-year lookback window resulted in 4.3–8.3 times higher prevalence (depending on disease), reducing incidence by 1.9–3.3 times. All-time lookback resulted in strong temporal trends. COPD prevalence between 2011 and 2014 in MarketScan increased by 25% with an all-time lookback but stayed relatively constant with a 1-year lookback. Varying observability did not substantially affect estimates.

Conclusion

This framework draws attention to the underrecognized potential for widely varying incidence and prevalence estimates, with implications for care planning and drug development. Though prevalence and incidence are seemingly straightforward concepts, careful consideration of methodology is required to obtain meaningful estimates from health care databases.

Evaluation of Electronic Healthcare Databases for Post-Marketing Drug Safety Surveillance and Pharmacoepidemiology in China

INTRODUCTION

Electronic healthcare databases (EHDs) are used increasingly for post-marketing drug safety surveillance and pharmacoepidemiology in Europe and North America. However, few studies have examined the potential of these data sources in China.

METHODS

Three major types of EHDs in China (i.e., a regional community-based database, a national claims database, and an electronic medical records [EMR] database) were selected for evaluation. Forty core variables were derived based on the US Mini-Sentinel (MS) Common Data Model (CDM) as well as the data features in China that would be desirable to support drug safety surveillance. An email survey of these core variables and eight general questions as well as follow-up inquiries on additional variables was conducted. These 40 core variables across the three EHDs and all variables in each EHD along with those in the US MS CDM and Observational Medical Outcomes Partnership (OMOP) CDM were compared for availability and labeled based on specific standards.

RESULTS

All of the EHDs' custodians confirmed their willingness to share their databases with academic institutions after appropriate approval was obtained. The regional community-based database contained 1.19 million people in 2015 with 85% of core variables. Resampled annually nationwide, the national claims database included 5.4 million people in 2014 with 55% of core variables, and the EMR database included 3 million inpatients from 60 hospitals in 2015 with 80% of core variables. Compared with MS CDM or OMOP CDM, the proportion of variables across the three EHDs available or able to be transformed/derived from the original sources are 24-83% or 45-73%, respectively.

CONCLUSIONS

These EHDs provide potential value to post-marketing drug safety surveillance and pharmacoepidemiology in China. Future research is warranted to assess the quality and completeness of these EHDs or additional data sources in China.

Cardiovascular Safety of Tocilizumab Versus Tumor Necrosis Factor Inhibitors in Patients With Rheumatoid Arthritis: A Multi-Database Cohort Study

Objective

While tocilizumab (TCZ) is known to increase low‐density lipoprotein (LDL) cholesterol levels, it is unclear whether TCZ increases cardiovascular risk in patients with rheumatoid arthritis (RA). This study was undertaken to compare the cardiovascular risk associated with receiving TCZ versus tumor necrosis factor inhibitors (TNFi).

Methods

To examine comparative cardiovascular safety, we conducted a cohort study of RA patients who newly started TCZ or TNFi using claims data from Medicare, IMS PharMetrics, and MarketScan. All patients were required to have previously used a different TNFi, abatacept, or tofacitinib. The primary outcome measure was a composite cardiovascular end point of hospitalization for myocardial infarction or stroke. TCZ initiators were propensity score matched to TNFi initiators with a variable ratio of 1:3 within each database, controlling for >65 baseline characteristics. A fixed‐effects model combined database‐specific hazard ratios (HRs).

Results

We included 9,218 TCZ initiators propensity score matched to 18,810 TNFi initiators across all 3 databases. The mean age was 72 years in Medicare, 51 in PharMetrics, and 53 in MarketScan. Cardiovascular disease was present at baseline in 14.3% of TCZ initiators and 13.5% of TNFi initiators. During the study period (mean ± SD 0.9 ± 0.7 years; maximum 4.5 years), 125 composite cardiovascular events occurred, resulting in an incidence rate of 0.52 per 100 person‐years for TCZ initiators and 0.59 per 100 person‐years for TNFi initiators. The risk of cardiovascular events associated with TCZ use versus TNFi use was similar across all 3 databases, with a combined HR of 0.84 (95% confidence interval 0.56–1.26).

Conclusion

This multi‐database population‐based cohort study showed no evidence of an increased cardiovascular risk among RA patients who switched from a different biologic drug or tofacitinib to TCZ versus to a TNFi.

Conducting Privacy-Preserving Multivariable Propensity Score Analysis When Patient Covariate Information Is Stored in Separate Locations

Distributed networks of health-care data sources are increasingly being utilized to conduct pharmacoepidemiologic database studies. Such networks may contain data that are not physically pooled but instead are distributed horizontally (separate patients within each data source) or vertically (separate measures within each data source) in order to preserve patient privacy. While multivariable methods for the analysis of horizontally distributed data are frequently employed, few practical approaches have been put forth to deal with vertically distributed health-care databases. In this paper, we propose 2 propensity score-based approaches to vertically distributed data analysis and test their performance using 5 example studies. We found that these approaches produced point estimates close to what could be achieved without partitioning. We further found a performance benefit (i.e., lower mean squared error) for sequentially passing a propensity score through each data domain (called the "sequential approach") as compared with fitting separate domain-specific propensity scores (called the "parallel approach"). These results were validated in a small simulation study. This proof-of-concept study suggests a new multivariable analysis approach to vertically distributed health-care databases that is practical, preserves patient privacy, and warrants further investigation for use in clinical research applications that rely on health-care databases.

Gout and the Risk of Non-vertebral Fracture

Prior studies suggest an association between osteoporosis, systemic inflammation, and pro-inflammatory cytokines such as interleukin (IL)-1 and IL-6. Conflicting findings exist on the association between hyperuricemia and osteoporosis. Furthermore, it remains unknown whether gout, a common inflammatory arthritis, affects fracture risk. Using data from a US commercial health plan (2004-2013), we evaluated the risk of non-vertebral fracture (ie, forearm, wrist, hip, and pelvis) in patients with gout versus those without. Gout patients were identified with ≥2 diagnosis codes and ≥1 dispensing for a gout-related drug. Non-gout patients, identified with ≥2 visits coded for any diagnosis and ≥1 dispensing for any prescription drugs, were free of gout diagnosis and received no gout-related drugs. Hip fracture was the secondary outcome. Fractures were identified with a combination of diagnosis and procedure codes. Cox proportional hazards models compared the risk of non-vertebral fracture in gout patients versus non-gout, adjusting for more than 40 risk factors for osteoporotic fracture. Among gout patients with baseline serum uric acid (sUA) measurements available, we assessed the risk of non-vertebral fracture associated with sUA. We identified 73,202 gout and 219,606 non-gout patients, matched on age, sex, and the date of study entry. The mean age was 60 years and 82% were men. Over the mean 2-year follow-up, the incidence rate of non-vertebral fracture per 1,000 person-years was 2.92 in gout and 2.66 in non-gout. The adjusted hazard ratio (HR) was 0.98 (95% confidence interval [CI] 0.85-1.12) for non-vertebral fracture and 0.83 (95% CI 0.65-1.07) for hip fracture in gout versus non-gout. Subgroup analysis (n = 15,079) showed no association between baseline sUA and non-vertebral fracture (HR = 1.03, 95% CI 0.93-1.15), adjusted for age, sex, comorbidity score, and number of any prescription drugs. Gout was not associated with a risk of non-vertebral fracture. Among patients with gout, sUA was not associated with the risk of non-vertebral fracture. © 2016 American Society for Bone and Mineral Research.

Few Disparities in Baseline Laboratory Testing After the Diuretic or Digoxin Initiation by Medicare Fee-For-Service Beneficiaries

BACKGROUND

Despite the persistence of significant disparities, few evaluations examine disparities in laboratory testing by race/ethnicity, age, sex, Medicaid eligibility, and number of chronic conditions for Medicare fee-for-service beneficiaries' newly prescribed medications. In Medicare beneficiaries initiating diuretics or digoxin, this study examined disparities in guideline-appropriate baseline laboratory testing and abnormal laboratory values.

METHODS AND RESULTS

To evaluate guideline-concordant testing for serum creatinine and serum potassium within 180 days before or 14 days after the index prescription fill date, we constructed retrospective cohorts from 10 states of 99 711 beneficiaries who had heart failure or hypertension initiating diuretic in 2011 and 8683 beneficiaries who had heart failure or atrial fibrillation initiating digoxin. Beneficiaries initiating diuretics were less likely to have testing if they were non-Hispanic Black (relative risk [RR], 0.99; 95% confidence interval [CI], 0.98-0.99) than non-Hispanic White. Beneficiaries initiating diuretics and beneficiaries initiating digoxin were more likely to have testing if they had multiple chronic conditions relative to 0 to 1 conditions. Beneficiaries initiating diuretics with laboratory values were more likely to have an abnormal serum creatinine value at baseline if they were non-Hispanic Black (RR, 2.57; 95% CI, 1.91-3.44), other race (RR, 2.11; 95% CI, 1.08-4.10), or male (RR, 2.75; 95% CI, 2.14-3.52) or an abnormal serum potassium value if they were aged ≥76 years (RR, 1.29; 95% CI, 1.09-1.51) or male (RR, 1.17; 95% CI, 1.03-1.33).

CONCLUSIONS

Testing rates were consistently high, so there were negligible disparities in guideline-concordant testing of creatinine and potassium after the initiation of digoxin or diuretics by Medicare beneficiaries.

Appropriate baseline laboratory testing following ACEI or ARB initiation by Medicare FFS beneficiaries

BACKGROUND

Laboratory testing to identify contraindications and adverse drug reactions is important for safety of patients initiating angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs). Rates and predictors of appropriate testing among Medicare fee-for-service beneficiaries are unknown.

PURPOSE

The study's purpose was to examine baseline laboratory testing rates, identify predictors of suboptimal testing, and assess the prevalence of abnormal creatinine and potassium among beneficiaries initiating ACE inhibitors or ARBs.

DESIGN AND SUBJECTS

Retrospective cohort of 101 376 fee-for-service beneficiaries from 10 eastern US states in 1 July to 30 November 2011.

MAIN MEASURES

Appropriate monitoring for serum creatinine or serum potassium was defined as evidence of an outpatient claim within 180 days before or 14 days after the index prescription fill date.

KEY RESULTS

Thirty-eight percent of beneficiaries were men, 78% were White race, 26% had prevalent heart failure, and 89% had prevalent hypertension. Rates of appropriate baseline laboratory testing were 82.7% for potassium, 83.2% for creatinine, and 82.6% for both potassium and creatinine 180 days prior to initiation. In logistic regression, men (odds ratio [OR] = 1.15, 95% confidence interval [CI]: 1.11, 1.19), African-Americans (OR = 1.26, 95%CI: 1.20, 1.32), and beneficiaries with Alzheimer's disease and related disorders (OR = 1.22, 95%CI: 1.15, 1.28) or stroke (OR = 1.34, 95%CI: 1.26, 1.43) were more likely to experience suboptimal testing. At baseline, hyperkalemia was relatively uncommon (5.8%), and elevated creatinine values were rare (1.4%).

CONCLUSIONS

Appropriate monitoring could be improved for African-American beneficiaries and beneficiaries with a history of stroke or Alzheimer's disease and related disorders initiating ACE inhibitors or ARBs. Copyright © 2016 John Wiley & Sons, Ltd.

Incorporating linked healthcare claims to improve confounding control in a study of in-hospital medication use

INTRODUCTION

The Premier Perspective hospital billing database provides a promising data source for studies of inpatient medication use. However, in-hospital recording of confounders is limited, and incorporating linked healthcare claims data available for a subset of the cohort may improve confounding control. We investigated methods capable of adjusting for confounders measured in a subset, including complete case analysis, multiple imputation of missing data, and propensity score (PS) calibration.

METHODS

Methods were implemented in an example study of adults in Premier undergoing percutaneous coronary intervention (PCI) in 2004-2008 and exposed to either bivalirudin or heparin. In a subset of patients enrolled in UnitedHealth for at least 90 days before hospitalization, additional confounders were assessed from healthcare claims, including comorbidities, prior medication use, and service use intensity. Diagnostics for each method were evaluated, and methods were compared with respect to the estimates and confidence intervals of treatment effects on repeat PCI, bleeding, and in-hospital death.

RESULTS

Of 210,268 patients in the hospital-based cohort, 3240 (1.5 %) had linked healthcare claims. This subset was younger and healthier than the overall study population. The linked subset was too small for complete case evaluation of two of the three outcomes of interest. Multiple imputation and PS calibration did not meaningfully impact treatment effect estimates and associated confidence intervals.

CONCLUSIONS

Despite more than 98 % missingness on 24 variables, PS calibration and multiple imputation incorporated confounders from healthcare claims without major increases in estimate uncertainty. Additional research is needed to determine the relative bias of these methods.

Linkage of Laboratory Results to Medicare Fee-for-Service Claims

BACKGROUND

Medicare is the single largest purchaser of laboratory testing in the United States, yet test results associated with Medicare laboratory claims have historically not been available.

OBJECTIVE

The purpose of this study was to describe both the linkage of laboratory results data to Medicare claims and the completeness of these results data. In a subgroup of beneficiaries initiating angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers, we also demonstrate the generalizability of Medicare beneficiaries with laboratory values compared with those without laboratory values. We end with a discussion of the limitations and potential uses of these linked data.

METHODS

We obtained information about laboratory orders and results for all Medicare fee-for-service beneficiaries in 10 eastern states who had outpatient laboratory tests conducted by a large national laboratory services vendor in 2011. Using a combination of direct identifiers and patient demographic characteristics, we linked patients in these laboratory data to Medicare beneficiaries, enabling us to associate test results with existing claims.

RESULTS

Nearly all patients in the laboratory data were able to be linked to Medicare beneficiaries. There were over 2 million distinct beneficiaries with nearly 125 million specific test results in the laboratory data. For specific tests ordered in an outpatient or office setting in these 10 states, between 5% and 15% of them had linked laboratory data. Medicare beneficiaries initiating angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers who had laboratory results data had similar patient characteristics to those without results data.

CONCLUSIONS

This novel linkage of laboratory results data to Medicare claims creates unprecedented opportunities for conducting comparative effectiveness research related to patient safety and quality.

Disease-modifying antirheumatic drug use and the risk of incident hyperlipidemia in patients with early rheumatoid arthritis: a retrospective cohort study

OBJECTIVE

To compare the risk of incident hyperlipidemia in early rheumatoid arthritis (RA) patients after initiation of various disease-modifying antirheumatic drugs (DMARDs).

METHODS

We conducted a cohort study using insurance claims data (2001-2012) in early RA patients. Early RA was defined by the absence of any RA diagnosis or DMARD prescriptions for 12 months. Four mutually exclusive groups were defined based on DMARD initiation: tumor necrosis factor α (TNFα) inhibitors ± nonbiologic (nb) DMARDs, methotrexate (MTX) ± nonhydroxycholorquine nbDMARDs, hydroxychloroquine ± non-MTX nbDMARDs, and other nbDMARDs only. The primary outcome was incident hyperlipidemia, defined by a diagnosis and a prescription for a lipid-lowering agent. For the subgroup of patients with laboratory results available, change in lipid levels was assessed. Multivariable Cox proportional hazard models and propensity score (PS) decile stratification with asymmetric trimming were used to control for confounding.

RESULTS

Of the 17,145 early RA patients included in the study, 364 developed incident hyperlipidemia. The adjusted hazard ratios (HRs; 95% confidence intervals [95% CIs]) for hyperlipidemia were 1.41 (95% CI 0.99, 2.00) for TNFα inhibitors, 0.81 (95% CI 0.63, 1.04) for hydroxychloroquine, and 1.33 (95% CI 0.95, 1.84) for other nbDMARDs compared with MTX in the full cohort, while HRs for the PS trimmed cohort were 1.18 (95% CI 0.80, 1.73), 0.75 (95% CI 0.58, 0.98), and 1.41 (95% CI 1.01, 1.98), respectively. In the subgroup analysis, hydroxychloroquine use showed significant reduction in low-density lipoprotein (-8.9 mg/dl, 95% CI -15.8, -2.0), total cholesterol (-12.3 mg/dl, 95% CI -19.8, -4.8) and triglyceride levels (-19.5 mg/dl, 95% CI -38.7, -0.3) from baseline compared with MTX.

CONCLUSION

Use of hydroxychloroquine may be associated with a lower risk of hyperlipidemia among early RA patients.

Effects of xanthine oxidase inhibitors on cardiovascular disease in patients with gout: a cohort study

BACKGROUND

Hyperuricemia and gout are associated with an increased risk of cardiovascular disease (CVD). It is unknown whether treating hyperuricemia with xanthine oxidase inhibitors (XOIs), including allopurinol and febuxostat, modifies cardiovascular risks.

METHODS

We used US insurance claims data to conduct a cohort study among gout patients, comparing XOI initiators with non-users with hyperuricemia defined as serum uric acid level ≥6.8 mg/dL. We calculated incidence rates of a composite nonfatal cardiovascular outcome that included myocardial infarction, coronary revascularization, stroke, and heart failure. Propensity score (PS)-matched Cox proportional hazards regression compared the risk of composite cardiovascular endpoint in XOI initiators vs those with untreated hyperuricemia, controlling for baseline confounders. In a subgroup of patients with uric acid levels available, PS-matched Cox regression further adjusted for baseline uric acid levels.

RESULTS

There were 24,108 PS-matched pairs with a mean age of 51 years and 88% male. The incidence rate per 1000 person-years for composite CVD was 24.1 (95% confidence interval [CI] 22.6-26.0) in XOI initiators and 21.4 (95% CI, 19.8-23.2) in the untreated hyperuricemia group. The PS-matched hazard ratio for composite CVD was 1.16 (95% CI, 0.99-1.34) in XOI initiators vs those with untreated hyperuricemia. In subgroup analyses, the PS-matched hazard ratio for composite CVD adjusted for serum uric acid levels was 1.10 (95% CI, 0.74-1.64) among XOI initiators.

CONCLUSIONS

Among patients with gout, initiation of XOI was not associated with an increased or decreased cardiovascular risk compared with those with untreated hyperuricemia. Subgroup analyses adjusting for baseline uric acid levels also showed no association between XOI and cardiovascular risk.

Prevalence and characteristics of patients with low levels of low-density lipoprotein cholesterol in northern Denmark: a descriptive study

BACKGROUND

With the emergence of new lipid-lowering therapies, more patients are expected to achieve substantial lowering of low-density lipoprotein cholesterol (LDL-C). However, there are limited data examining the clinical experience of patients with low (<1.3 mmol/L) or very low (<0.65 mmol/L) levels of LDL-C. To provide information on patients with low LDL-C, we identified and characterized persons with low LDL-C using data from Danish medical databases.

METHODS

Using a population-based clinical laboratory database, we identified adults with at least one LDL-C measurement in northern Denmark between 1998 and 2011 (population approximately 1.5 million persons). Based on the lowest measurement during the study period, we divided patients into groups with low (<1.3 mmol/L), moderate (1.3-3.3 mmol/L), or high (>3.3 mmol/L) LDL-C. We described their demographic characteristics, entire comorbidity history, and 90-day prescription history prior to the lowest LDL-C value measured. Finally, we further restricted the analysis to individuals with very low LDL-C (<0.65 mmol/L).

RESULTS

Among 765,503 persons with an LDL-C measurement, 23% had high LDL-C, 73% had moderate LDL-C, and 4.8% had low LDL-C. In the latter group, 9.6% (0.46% of total) had very low LDL-C. Compared with the moderate and high LDL-C categories, the low LDL-C group included more males and older persons with a higher prevalence of cardiovascular disease, diabetes, chronic pulmonary disease, ulcer disease, and obesity, as measured by hospital diagnoses or relevant prescription drugs for these diseases. Cancer and use of psychotropic drugs were also more prevalent. These patterns of distribution became even more pronounced when restricting to individuals with very low LDL-C.

CONCLUSION

Using Danish medical databases, we identified a cohort of patients with low LDL-C and found that cohort members differed from patients with higher LDL-C levels. These differences may be explained by various factors, including prescribing patterns of lipid-lowering therapies.

Measuring frailty using claims data for pharmacoepidemiologic studies of mortality in older adults: evidence and recommendations

PURPOSE

Geriatric frailty is a common syndrome of older adults that is characterized by increased vulnerability to adverse health outcomes and influences treatment choice. Pharmacoepidemiologic studies that rely on administrative claims data in older adults are limited by confounding due to unmeasured frailty. A claims-based frailty score may be useful to minimize confounding by frailty in such databases. We provide an overview of definitions and measurement of frailty, evaluated the claims-based models of frailty in literature, and recommend ways to improve frailty adjustment in claims analysis.

METHODS

We searched MEDLINE and EMBASE from inception to April 2014, without language restriction, to identify claims-based multivariable models that predicted frailty or its related outcome, disability. We critically appraised their approach, including population, predictor selection, outcome definition, and model performance.

RESULTS

Of 152 reports, three models were identified. One model that predicted poor functional status using healthcare service claims in a representative sample of community-dwelling and institutionalized older adults showed an excellent discrimination (C statistic, 0.92). The other two models that predicted disability using either diagnosis codes or prescription claims alone in institutionalized or frail adults had limited generalizability and modest model performance. None of the models have been applied to reduce confounding bias in pharmacoepidemiologic studies of drug therapy.

CONCLUSIONS

We found little research conducted on development and application of a claims-based frailty index for confounding adjustment in pharmacoepidemiologic studies in older adults. More research is needed to advance this innovative, potentially useful approach by incorporating the expertise from aging research.

Claims-based algorithms for identifying Medicare beneficiaries at high estimated risk for coronary heart disease events: a cross-sectional study

BACKGROUND

Databases of medical claims can be valuable resources for cardiovascular research, such as comparative effectiveness and pharmacovigilance studies of cardiovascular medications. However, claims data do not include all of the factors used for risk stratification in clinical care. We sought to develop claims-based algorithms to identify individuals at high estimated risk for coronary heart disease (CHD) events, and to identify uncontrolled low-density lipoprotein (LDL) cholesterol among statin users at high risk for CHD events.

METHODS

We conducted a cross-sectional analysis of 6,615 participants ≥66 years old using data from the REasons for Geographic And Racial Differences in Stroke (REGARDS) study baseline visit in 2003-2007 linked to Medicare claims data. Using REGARDS data we defined high risk for CHD events as having a history of CHD, at least 1 risk equivalent, or Framingham CHD risk score >20%. Among statin users at high risk for CHD events we defined uncontrolled LDL cholesterol as LDL cholesterol ≥100 mg/dL. Using Medicare claims-based variables for diagnoses, procedures, and healthcare utilization, we developed algorithms for high CHD event risk and uncontrolled LDL cholesterol.

RESULTS

REGARDS data indicated that 49% of participants were at high risk for CHD events. A claims-based algorithm identified high risk for CHD events with a positive predictive value of 87% (95% CI: 85%, 88%), sensitivity of 69% (95% CI: 67%, 70%), and specificity of 90% (95% CI: 89%, 91%). Among statin users at high risk for CHD events, 30% had LDL cholesterol ≥100 mg/dL. A claims-based algorithm identified LDL cholesterol ≥100 mg/dL with a positive predictive value of 43% (95% CI: 38%, 49%), sensitivity of 19% (95% CI: 15%, 22%), and specificity of 89% (95% CI: 86%, 90%).

CONCLUSIONS

Although the sensitivity was low, the high positive predictive value of our algorithm for high risk for CHD events supports the use of claims to identify Medicare beneficiaries at high risk for CHD events.

Electronic clinical laboratory test results data tables: lessons from Mini-Sentinel

PURPOSE

Developing electronic clinical data into a common data model posed substantial challenges unique from those encountered with administrative data. We present here the design, implementation, and use of the Mini-Sentinel Distributed Database laboratory results table (LRT).

METHODS

We developed the LRT and guided Mini-Sentinel data partners (DPs) in populating it from their source data. Data sources included electronic health records and internal and contracted clinical laboratory systems databases. We employed the Logical Observation Identifiers, Names, and Codes (LOINC®) results reporting standards. We evaluated transformed results data using data checks and an iterative, ongoing characterization and harmonization process.

RESULTS

Key LRT variables included test name, subcategory, specimen source, LOINC, patient location, specimen date and time, result unit, and unique person identifier. Selected blood and urine chemistry, hematology, coagulation, and influenza tests were included. Twelve DPs with outpatient test results participated; four also contributed inpatient test results. As of September 2013, the LRT included 385,516,239 laboratory test results; data are refreshed at least quarterly. LOINC availability and use varied across DP. Multiple data quality and content issues were identified and addressed.

CONCLUSION

Developing the LRT brought together disparate data sources with no common coding structure. Clinical laboratory test results obtained during routine healthcare delivery are neither uniformly coded nor documented in a standardized manner. Applying a systematic approach with data harmonization efforts and ongoing oversight and management is necessary for a clinical laboratory results data table to remain valid and useful.