Claims Dataset Research Articles

Occupational Injury and Suicide in Washington State, Adjusting for Pre-Injury Depression.

Occupational injuries have been associated with increased suicide mortality, but prior studies have not accounted for pre-injury depression. We linked injuries that occurred from 1994 to 2000 in the Washington State workers' compensation system with Social Security Administration data on earnings and mortality through 2018. We estimated the subdistribution hazard ratio (sHR) and 95% confidence interval using competing risks regression of suicide deaths with lost time compared with medical-only injuries separately for men and women, adjusting for age, pre-injury annual earnings, and industry. We further adjusted for pre-injury diagnosis of major depressive disorder by using a quantitative bias analysis (QBA), with the prevalence of this disorder in workers derived from an external health insurance claims data set. Elevated suicide mortality was observed following lost-time injuries compared with medical-only injuries for men (sHR = 1.49, 95% CI [1.14, 1.93]) and women (sHR = 1.30, 95% CI [1.00, 1.69]), adjusting for age, pre-injury earnings, and industry. Adjusted for pre-injury depression using a QBA, elevated suicide risk in men remained statistically significant (median sHR = 1.33, simulation interval [1.18, 1.47]) but not for women. Workplace injury requiring time off work appeared to remain influential in increasing suicide risk among men, even after controlling for pre-injury depression. The relationship between mental health before and after occupational injury is complex and studies should better integrate mental health pre-injury. Though many questions remain on the complex relationship between work, depression, injuries, and suicide, employers should work to prevent injuries and consider implementing mental health programs, which could be helpful in reducing suicide risk.

Unpaid insurance claims: Language and communication gaps

Language and communication play pivotal roles in the insurance industry, impacting claims processing, customer satisfaction, and operational efficiency. This study investigates the dynamics of errors within insurance claims across six distinct languages—African, Afrikaans, English, French, German, and Asian languages. The purpose is to identify language-specific error patterns and their implications for customer-centric methods. This study utilises a dataset of insurance claims spanning multiple languages. The analysis encompasses the probability theories approach to unravel error distributions and their associations. The study reveals that linguistic diversity significantly impacts the insurance industry, emphasising the need for language-centric approaches to error reduction This highlights the need for customised communication strategies and enlightening initiatives to address the unique challenges faced by this linguistic demographic especially the complex legal jargon that is not easily understood by all. In conclusion, this paper advises insurance companies to navigate the complex terrain of multilingual interactions. It enriches the existing body of knowledge surrounding insurance claim management, offering a roadmap to navigate the intricacies of language diversity within the industry. By addressing the linguistic-specific challenges identified in this study, insurance providers can enhance their services, reduce errors, and ultimately foster improved customer experiences.

Epidemiology, Real-World Treatment Patterns, and Patient Outcomes of Primary Advanced or Recurrent Endometrial Cancer in Germany between 2015-2021.

The aim of this study was to describe the epidemiology of primary advanced or recurrent endometrial cancer and the outcomes from real-world treatment patterns of patients affected in Germany between 2015 to 2021. In this retrospective cohort study covering the period from 1 January 2015 to 31 December 2021, data from patients with primary advanced or recurrent endometrial cancer who initiated systemic treatment for their disease were extracted from an anonymized claims dataset. Epidemiologic outcomes were cumulative incidence of endometrial cancer and point prevalence. Overall survival after the index date was assessed, with all-cause death used as an event. Endometrial cancer-related real-world treatment patterns were described for the post-index period. The incidence of primary advanced or recurrent endometrial cancer in 2021 was 4.77 cases/100 000 persons, with no substantial change over time (4.63 in 2018; 4.93 in 2019; 4.45 in 2020). The point prevalence on 1 January 2022 was 0.023%, with a slight increase in prevalence observed from 1 January 2019 onwards. Among 466 patients with confirmed endometrial cancer, the mean (standard deviation) age was 68.0 (11.6) years; the tumor material from 86 patients (18.5%) underwent immunohistochemistry or polymerase chain reaction testing. Median overall survival was estimated to be 47.5 months (95% CI 35.1 to 70.4) and the 5-year survival probability was 46.2%. The most frequent first-line systemic therapies were carboplatin (45.7%) and paclitaxel (43.1%). Second-line therapy was received by 153 patients (32.8%). The analysis of the German claims data produced contemporary epidemiologic estimates for advanced or recurrent endometrial cancer. Treatments were aligned with guideline recommendations during the study period, with tumor testing yet to enter mainstream practice.

Model uncertainty assessment for symmetric and right-skewed distributions

In actuarial modeling, right-skewed distributions are of paramount importance. Typically, they have the property of becoming unimodal and symmetric after applying some increasing concave transformation T (e.g. log-transformation). In what follows, we refer to them as T -unimodal T -symmetric distributions. In this paper, we derive attainable bounds for the Value-at-Risk for such distributions when some partial information is available, such as information on the support, median, or certain moments. We also derive explicit upper and lower bounds for the Range Value-at-Risk under knowledge of unimodality, symmetry, mean, variance, and possibly bounded support. In passing, we provide a generalization of the Gauss inequality for symmetric distributions with known support. We show how these bounds improve on bounds available in the literature using a real-world automobile insurance claims dataset.

We Are on the Verge of Breakthrough Cures for Type 1 Diabetes, but Who Are the 2 Million Americans Who Have It?

Background: Two million Americans have type 1 diabetes (T1DM). Innovative treatments have standardized insulin delivery and improved outcomes for patients, but patients' access to such technologies depends on social determinants of health, including insurance coverage, proper diagnosis, and appropriate patient supports. Prior estimates of US prevalence, incidence, and patient characteristics have relied on data from select regions and younger ages and miss important determinants. Objectives: This study sought to use large, nationally representative healthcare claims data sets to holistically estimate the size of the current US population with T1DM and investigate geographic nuances in prevalence and incidence, patient demographics, insurance coverage, and device use. This work also aimed to project T1DM population growth over the next 10 years. Methods: We used administrative claims from 4 sources to identify prevalent and incident T1DM patients in the US, as well as various demographic and insurance characteristics of the patient population. We combined this data with information from national population growth projections and literature to construct an actuarial model to project growth of the T1DM population based on current trends and scenarios for 2024, 2029, and 2033. Results: We estimated 2.07 million T1DM patients nationally across all insurance coverages in our 2024 baseline model year: 1.79 million adults (≥20 years) and 0.28 million children. This represents a US T1DM prevalence rate of 617 per 100 000 and an incidence rate of 0.016%. By 2033, we project the US population with T1DM will grow by about 10%, reaching approximately 2.29 million patients. Discussion: Our results showed important differences in T1DM prevalence and incidence across regions, payers, and ethnic groups. This suggests studies based on more geographically concentrated data may miss important variation in prevalence and incidence across regions. It also indicates T1DM prevalence tends to vary by income, consistent with several international studies. Conclusions: Accurate projections of T1DM population growth are critical to ensure appropriate healthcare coverage and reimbursement for treatments. Our work supports future policy and research efforts with 2024, 2029, and 2033 projections of demographics and insurance coverage for people with T1DM.

We Are on the Verge of Breakthrough Cures for Type 1 Diabetes, but Who Are the 2 Million Americans Who Have It?

Background: Two million Americans have type 1 diabetes (T1DM). Innovative treatments have standardized insulin delivery and improved outcomes for patients, but patients’ access to such technologies depends on social determinants of health, including insurance coverage, proper diagnosis, and appropriate patient supports. Prior estimates of US prevalence, incidence, and patient characteristics have relied on data from select regions and younger ages and miss important determinants. Objectives: This study sought to use large, nationally representative healthcare claims data sets to holistically estimate the size of the current US population with T1DM and investigate geographic nuances in prevalence and incidence, patient demographics, insurance coverage, and device use. This work also aimed to project T1DM population growth over the next 10 years. Methods: We used administrative claims from 4 sources to identify prevalent and incident T1DM patients in the US, as well as various demographic and insurance characteristics of the patient population. We combined this data with information from national population growth projections and literature to construct an actuarial model to project growth of the T1DM population based on current trends and scenarios for 2024, 2029, and 2033. Results: We estimated 2.07 million T1DM patients nationally across all insurance coverages in our 2024 baseline model year: 1.79 million adults (≥20 years) and 0.28 million children. This represents a US T1DM prevalence rate of 617 per 100 000 and an incidence rate of 0.016%. By 2033, we project the US population with T1DM will grow by about 10%, reaching approximately 2.29 million patients. Discussion: Our results showed important differences in T1DM prevalence and incidence across regions, payers, and ethnic groups. This suggests studies based on more geographically concentrated data may miss important variation in prevalence and incidence across regions. It also indicates T1DM prevalence tends to vary by income, consistent with several international studies. Conclusions: Accurate projections of T1DM population growth are critical to ensure appropriate healthcare coverage and reimbursement for treatments. Our work supports future policy and research efforts with 2024, 2029, and 2033 projections of demographics and insurance coverage for people with T1DM.

The Issue Correlates of War (ICOW) Identity Claims Dataset, 1946-2021

The Issue Correlates of War (ICOW) Identity Claims Dataset, 1946-2021

Efficacy of a Multimodal Digital Behavior Change Intervention on Lifestyle Behavior, Cardiometabolic Biomarkers, and Medical Expenditure: Protocol for a Randomized Controlled Trial.

The US Preventive Services Task Force recommends providers offer individualized healthy behavior interventions for all adults, independent of their risk of cardiovascular disease. While strong evidence exists to support disease-specific programs designed to improve multiple lifestyle behaviors, approaches to adapting these interventions for a broader population are not well established. Digital behavior change interventions (DBCIs) hold promise as a more generalizable and scalable approach to overcome the resource and time limitations that traditional behavioral intervention programs face, especially within an occupational setting. We aimed to evaluate the efficacy of a multimodal DBCI on (1) self-reported behaviors of physical activity, nutrition, sleep, and mindfulness; (2) cardiometabolic biomarkers; and (3) chronic disease-related medical expenditure. We conducted a 2-arm randomized controlled trial for 12 months among employees of an academic health care facility in the United States. The intervention arm received a scale, a smartphone app, an activity tracker, a video library for healthy behavior recommendations, and an on-demand health coach. The control arm received standard employer-provided health and wellness benefits. The primary outcomes of the study included changes in self-reported lifestyle behaviors, cardiometabolic biomarkers, and chronic disease-related medical expenditure. We collected health behavior data via baseline and quarterly web-based surveys, biometric measures via clinic visits at baseline and 12 months, and identified relevant costs through claims datasets. A total of 603 participants were enrolled and randomized to the intervention (n=300, 49.8%) and control arms (n=303, 50.2%). The average age was 46.7 (SD 11.2) years, and the majority of participants were female (80.3%, n=484), White (85.4%, n=504), and non-Hispanic (90.7%, n=547), with no systematic differences in baseline characteristics observed between the study arms. We observed retention rates of 86.1% (n=519) for completing the final survey and 77.9% (n=490) for attending the exit visit. This study represents the largest and most comprehensive evaluation of DBCIs among participants who were not selected based on their underlying condition to assess its impact on behavior, cardiometabolic biomarkers, and medical expenditure. ClinicalTrials.gov NCT04712383; https://clinicaltrials.gov/study/NCT04712383. RR1-10.2196/50378.

Assessing the alignment of obstructive hypertrophic cardiomyopathy pharmacological treatments with ESC guidelines in Germany

Abstract Introduction The European Society of Cardiology (ESC) recently updated treatment guidelines for hypertrophic cardiomyopathy (HCM). Purpose The aim of this study was to describe the treatment of obstructive HCM in Germany and assess alignment with the 2014 and 2023 ESC guidelines. Methods Adults with HCM were identified using a nationally representative administrative claims dataset (WIG2 Benchmark database) from several German Statutory Health Insurances between 2012 and 2018. Obstructive HCM was identified as any obstructive HCM diagnosis (ICD-10: I42.1) or any HCM diagnosis (I42.2, I42.9) with septal reduction therapy. The index date was defined as the date where the patient met all eligibility criteria for HCM. Minimum follow-up time was set to one year. Patients had at least 1 year of baseline data prior to the index HCM diagnosis and were followed up for at least 1 year after the diagnosis (or until death within 1 year). Patients with phenocopies were excluded. ESC guideline-directed pharmacological treatments for HCM included beta-blockers (BB), calcium channel blockers (CCB), disopyramide, and/or mavacamten (approved by EMA June 2023). While the 2014 ESC Guidelines on Diagnosis and Management of HCM were applicable during the study, recommendations pertaining to the use of non-vasodilating BBs and non-dihydropyridine (DHP) CCBs did not change in the 2023 ESC Guidelines for the Management of Cardiomyopathies; thus, this study applies to both. Results Out of 6793 patients with HCM analysed in the period 2012-2018, 1,141 had obstructive HCM. The average follow-up was 5 years. Patients were mostly male (62%) with a mean age of 60 years. Hypertension, hyperlipidaemia, coronary artery disease, heart failure, and atrial fibrillation were among the most common comorbidities; the mean Charlson Comorbidity Index score was 2.3. At index, 46% of patients were on BB monotherapy, 23% were on CCB monotherapy, <1% were on a combination of BBs or a combination of CCBs, 9% were on a combination of BBs and CCBs, and as many as 22% received no HCM treatment. Patients experienced 1615 unique treatment combinations throughout the study across all lines of therapy (Table 1). The most common treatments were bisoprolol (24%), metoprolol (19%), amlodipine (12%), and verapamil (12%). The use of propranolol was negligible (1%). 43% of pharmacological treatments were not in line with ESC guidelines. The most common reason for not aligning was using DHP-CCBs or vasodilating BBs. Conclusion This study shows that 57% of obstructive HCM treatments received by patients during the study were aligned with recommendations from the 2023 ESC guidelines on cardiomyopathies. Research is needed to investigate barriers and enablers to implementing the guideline including awareness of recommended treatments, as well as the management of comorbidities.

Analyzing missingness patterns in real-world data using the SMDI toolkit: application to a linked EHR-claims pharmacoepidemiology study

BackgroundMissing data in confounding variables present a frequent challenge in generating evidence using real-world data, including electronic health records (EHR). Our objective was to apply a recently published toolkit for characterizing missing data patterns and based on the toolkit results about likely missingness mechanisms, illustrate the decision-making process for analyses in an empirical case example.MethodsWe utilized the Structural Missing Data Investigations (SMDI) toolkit to characterize missing data patterns in the context of a pharmacoepidemiology study comparing cardiovascular outcomes of initiating sodium-glucose-cotransporter-2 inhibitors (SGLT2i) and dipeptidyl peptidase‐4 inhibitors (DPP‐4i) among older adults. The study used a linked EHR-Medicare claims dataset from Duke Health patients (2015–2017), focusing on partially observed confounders from EHR data (HbA1c lab and body mass index [BMI] values). Our analysis incorporated SMDI's descriptive functions and diagnostic tests to explore missingness patterns and determine missingness mitigation approaches. We used findings from these investigations to inform estimation of adjusted hazard ratios comparing the two classes of medications.ResultsHigh levels of missingness were noted for important confounding variables including HbA1c (63.6%) and BMI (16.5%). Diagnostic tests resulted in output that described: 1) the distributions of patient characteristics, exposure, and outcome between patients with or without an observed value of the partially observed covariate, 2) the ability to predict missingness based on observed covariates, and 3) estimate if the missingness of a partially observed covariate is differential with respect to the outcome. There was evidence that missingness could be sufficiently described using observed data, which allowed multiple imputation by chained equations using random forests to address missing confounder data in estimating treatment effects. Multiple imputation resulted in improved alignment of effect estimates with previous studies.ConclusionsWe were able to demonstrate the practical application of the SMDI toolkit in a real-world setting. Application of the SMDI toolkit and the resulting insights of potential missingness patterns can inform the choice of appropriate analytic methods and increase transparency of research methods in handling missing data. This type of approach can inform analytic decision making and may increase our ability to generate evidence from real-world data.

Patterns of Migration Following Dementia Diagnosis

Diseases have historically prompted individual relocations to mitigate the risk of disease acquisition or improve access to care. As dementia prevalence increases, comprehending the migration patterns of affected individuals is vital for public policy. To quantify the association of dementia diagnosis with migration patterns by examining the proportion of individuals with dementia who relocate, the timing of their moves relative to diagnosis, and the nature of their new living arrangements, whether in institutional settings or different households. This cohort study leveraged a comprehensive dataset of national Medicare claims and assessments spanning from 2012 to 2020, including Medicare Beneficiary Summary File and nursing home administrative datasets. The study focused on beneficiaries who received diagnoses in 2016 of dementia, myocardial infarction, chronic obstructive pulmonary disease, or colon cancer. Analyses were performed from March 2023 to August 2024. The primary outcome was migration, defined as change in county or state. The analysis distinguished between migrations with and without a nursing home stay. By tracking patients' residential county for 4 years before and after diagnosis, a difference-in-differences approach was used to contrast migration tendencies associated with dementia against the other 3 conditions. The sample included 1 626 127 Medicare beneficiaries (mean [SD] age, 80.1 [8.0] years; 922 194 women [56.7%]) who received diagnoses of the 4 conditions in 2016. In total, 818 862 had a new dementia diagnosis (age, 82.0 [7.8] years; 492 146 women [60.1%]). Comparing between the prediagnosis and postdiagnosis months, the proportion migrating to a different county increased by 8.5 percentage points (95% CI, 7.6-9.4 percentage points) for individuals with dementia and between 4.2 to 5.8 percentage points among those with myocardial infarction, chronic obstructive pulmonary disease, or colon cancer. The difference-in-difference estimates indicated a 3.9 percentage point (95% CI, 3.7-4.0 percentage points) increase in intercounty migration and a 1.9 percentage point (95% CI, 1.8-2.0 percentage points) increase in interstate migration for patients with dementia, effectively doubling the likelihood of migration compared with the other conditions. Of the excess migrations resulting from dementia diagnosis, 55% occurred in community settings, and 45% occurred in institutional settings. In this retrospective cohort study of Medicare fee-for-service beneficiaries, dementia was associated with a marked increase in migration rates over other major illnesses. This finding underscores the need to understand the factors associated with these distinct migration behaviors.

Impact of continuous glucose monitoring on emergency department visits and all-cause hospitalization rates among Medicaid beneficiaries with type 2 diabetes treated with multiple daily insulin or basal insulin therapy.

The increasing prevalence of diabetes in the United States continues to drive a steady rise in health care resource utilization, especially emergency department visits and all-cause hospitalizations, and the associated costs. To investigate the impact of continuous glucose monitoring (CGM) on emergency department visits and all-cause hospitalizations among Medicaid beneficiaries with type 2 diabetes (T2D) treated with multiple daily insulin injections (MDIs) or basal insulin therapy (BIT) in a real-world setting. In this retrospective, 12-month analysis, we used the Inovalon Insights claims dataset to evaluate the effects of CGM acquisition on emergency department visits and all-cause hospitalizations in the Managed Medicaid population. The analysis included 44,941 beneficiaries with T2D who were treated with MDIs (n = 35,367) or BIT (n = 9,574). Primary outcomes were changes in the number of emergency department visits and all-cause hospitalizations following 6 months after acquisition of CGM (post-index period) compared with 6 month prior to CGM acquisition (pre-index period). The first claim for CGM was the index date. Inclusion criteria were as follows: aged younger than 65 years, diagnosis of T2D, claims for short- or rapid-acting insulin (MDI group) or basal insulin (not rapid-acting) (BIT group), acquisition of a CGM device between January 1, 2017, and September 30, 2022, and continuous enrollment in their health plan throughout the pre-index and post-index periods. In the MDI group, all-cause inpatient hospitalization rates decreased from 3.25 to 2.29 events/patient-year (hazard ratio = 0.12; 95% CI = 0.11-0.13; P < 0.001) and emergency department visit rates decreased from 2.15 to 1.86 events/patient-year (hazard ratio = 0.52; 95% CI = 0.50-0.53; P < 0.001). In the BIT group, all-cause inpatient hospitalization rates decreased from 1.63 to 1.39 events/patient-year (hazard ratio = 0.11; 95% CI = 0.09-0.12; P < 0.001) and emergency department visit rates decreased from 1.60 to 1.43 events/patient-year (hazard ratio = 0.47; 95% CI = 0.44-0.50; P < 0.001). Acquisition of CGM is associated with significant reductions in emergency department visits and all-cause hospitalizations among people with T2D treated with MDIs or BIT.

Postpartum Emergency Care Visits Among North Carolina Medicaid Beneficiaries, 2013-2019.

Objective: To describe the rate, timing, and primary diagnosis codes for emergency care visits up to 8 weeks (56 days) after live birth among Medicaid beneficiaries in North Carolina (NC). Materials and Methods: Using a linked dataset of Medicaid hospital claims and certificates of live birth, which included Medicaid beneficiaries who had a live-born infant in NC between January 1, 2013, and November 4, 2019, and met inclusion criteria (n = 321,879), we estimated week-specific visit rates for emergency care visits that did not result in hospital admission (outpatient) and those that did (inpatient). We assessed the 10 leading diagnosis code categories for emergency care visits and described the characteristics of people with 0, 1, or ≥2 outpatient emergency care visits. Results: One in eight (12.4%) Medicaid beneficiaries had an emergency care visit that did not result in inpatient hospital admission during the first 8 weeks postpartum. Visit rates peaked in postpartum week 2. Diagnosis codes for nonspecific symptoms and substance use were the two leading diagnosis code categories for outpatient emergency care visits. Respiratory concerns and gastrointestinal concerns were the two leading diagnosis code categories for inpatient emergency care visits. Compared with those with zero outpatient emergency care visits, a greater proportion of people with ≥2 visits had less than a high school education, used tobacco during pregnancy, had Medicaid insurance outside of pregnancy, had mental health as a medical comorbidity, and/or had ≥2 medical comorbidities. Conclusions: These findings support scheduling health care visits early in the postpartum period, when emergency care visits are most frequent, and point to unmet needs for substance use support.

Does air pollution cause more car accidents? Evidence from auto insurance claims

Using a proprietary data set of auto insurance claims from May 2014 to December 2016, this paper examines the influence of air pollution on the number and severity of traffic accidents in China. Combining an instrumental variable strategy with high-dimensional fixed effects, we find that air pollution significantly increases the occurrence of traffic accidents, with each 1 μg/m3 increase in the particulate matter 2.5 (PM2.5) resulting in a 0.12 % increase in traffic accident probability and a 0.40 % increase in traffic accident number within one day. A different pattern is revealed in our analysis of accident severity, evidenced by a decrease of 1.20 % in the average claim ratio compared to its mean value and a reduction of 26 yuan in the average claim amount made with an increase of 1 μg/m3 in PM2.5. Combining the effect on the number and severity of traffic accidents, for each 1 μg/m3 increase in daily PM2.5, the district daily claim amount decreases by approximately 34 yuan. Further analysis indicates that this may be related to cautious driving behavior resulting from the driver's increased risk aversion. By exercising caution and care on the road, drivers can reduce the negative influence of air pollution on road safety and avoid non-subjective behavioral biases.

Uptake of telehealth services by people recovering from road trauma in Victoria, Australia

Introduction In recent years, injury insurers in Australia have introduced compensation cover for services delivered via telehealth. The aim of this study was to investigate the uptake of telehealth services by people recovering from road trauma in Victoria, Australia, and the characteristics of participants accessing these services. Methods All Victorian State Trauma Registry and Victorian Orthopaedic Trauma Outcomes Registry cases, aged ≥16 years, who sustained transport-related injury from March to August 2020, and were linked with the Transport Accident Commission claims dataset, were included in this registry-based cohort study. The primary outcome was the uptake of services accessed in-person and via telehealth. Results Within 12 months of injury, 413 participants accessed 93,035 services, with 1432 of these via telehealth (1.5%). Seventy-five percent of family counselling, 26% of psychology and 7% of medical consultations were accessed via telehealth. For telehealth users, a greater proportion had worked/studied prior to injury, met major trauma criteria, and were discharged to inpatient rehabilitation, compared to non-users. For participants with no telehealth use, a greater proportion were discharged home, and were classified as ‘rapid recovery’ clients. Conclusions While telehealth uptake was low during the study period, it was accessed by people with more severe injuries. There was also greater telehealth uptake during some of Victoria's strictest COVID-19 lockdowns. Telehealth may continue to be an appealing option for medical and psychological services into the future, and may assist with return to work following transport injury.

Does environmental regulation matter for healthcare utilisation in China? An interrupted time series study.

In recent years, China has implemented a series of environmental regulation policies to improve air quality, but the health effects of these policy changes need to be accurately clarified and quantified. The objective of this study was to empirically examine the effects of levying construction dust pollution fees on healthcare utilisation in a southern city of China. The study used a unique administrative insurance claim dataset from the city's Urban Employee Basic Medical Insurance scheme between 2013 and 2015. The sample included 69,961 enrolees. An interrupted time series design was employed to investigate whether and how the healthcare utilisation of enrolees changed after the policy change. The results showed that this environmental regulation policy did not affect the inpatient utilisation of enrolees but did negatively impact outpatient utilisation. In addition, outpatient utilisation of chronic disease coverage decreased for patients with chronic diseases, including diabetes and hypertension. This study provides evidence that enhancing environmental regulations helps reduce medical costs, which can benefit China and other developing countries to improve environmental quality and promote public health.

Increase in major osteoporotic fractures after therapy with immune checkpoint inhibitors

BackgroundImmune checkpoint inhibitors (ICIs) can cause severe and sometimes long-standing immune-related adverse events (irAEs). Enhanced immune activation from ICI can theoretically result in osteoclast activation, bone loss and fracture. The...

Benchmarking commercial healthcare claims data.

Commercial healthcare claims datasets represent a sample of the US population that is biased along socioeconomic/demographic lines; depending on the target population of interest, results derived from these datasets may not generalize. Rigorous comparisons of claims-derived results to ground-truth data that quantify this bias are lacking. (1) To quantify the extent and variation of the bias associated with commercial healthcare claims data with respect to different target populations; (2) To evaluate how socioeconomic/demographic factors may explain the magnitude of the bias. This is a retrospective observational study. Healthcare claims data come from the Merative™ MarketScan® Commercial Database; reference data for comparison come from the State Inpatient Databases (SID) and the US Census. We considered three target populations, aged 18-64 years: (1) all Americans; (2) Americans with health insurance; (3) Americans with commercial health insurance. We analyzed inpatient discharge records of patients aged 18-64 years, occurring between 01/01/2019 to 12/31/2019 in five states: California, Iowa, Maryland, Massachusetts, and New Jersey. We estimated rates of the 250 most common inpatient procedures, using claims data and using reference data for each target population, and we compared the two estimates. The average rate of inpatient discharges per 100 person-years was 5.39 in the claims data (95% CI: [5.37, 5.40]) and 7.003 (95% CI: [7.002, 7.004]) in the reference data for all Americans, corresponding to a 23.1% underestimate from claims. We found large variation in the extent of relative bias across inpatient procedures, including 22.8% of procedures that were underestimated by more than a factor of 2. There was a significant relationship between socioeconomic/demographic factors and the magnitude of bias: procedures that disproportionately occur in disadvantaged neighborhoods were more underestimated in claims data (R 2 51.6%, p < 0.001). When the target population was restricted to commercially insured Americans, the bias decreased substantially (3.2% of procedures were biased by more than factor of 2), but some variation across procedures remained. Naïve use of healthcare claims data to derive estimates for the underlying US population can be severely biased. The extent of bias is at least partially explained by neighborhood-level socioeconomic factors.

High-Dimensional Propensity Score and Its Machine Learning Extensions in Residual Confounding Control

“The use of health care claims datasets often encounters criticism due to the pervasive issues of omitted variables and inaccuracies or mis-measurements in available confounders. Ultimately, the treatment effects estimated using such data sources may be subject to residual confounding. Digital electronic administrative records routinely collect a large volume of health-related information; and many of which are usually not considered in conventional pharmacoepidemiological studies. A high-dimensional propensity score (hdPS) algorithm was proposed that uses such information as surrogates or proxies for mismeasured and unobserved confounders in an effort to reduce residual confounding bias. Since then, many machine learning and semi-parametric extensions of this algorithm have been proposed to better exploit the wealth of high-dimensional proxy information. In this tutorial, we will (i) demonstrate logic, steps and implementation guidelines of hdPS using an open data source as an example (using reproducible R codes), (ii) familiarize readers with the key difference between propensity score versus hdPS, as well as the requisite sensitivity analyses, (iii) explain the rationale for using the machine learning and double robust extensions of hdPS, and (iv) discuss advantages, controversies, and hdPS reporting guidelines while writing amanuscript.

AUTOMOBILE INSURANCE FRAUD DETECTION USING ENSEMBLE LEARNING MODELS

Automobile insurance fraud is a universal problem that has negative effects on both insurance companies and policyholders. This research proposes a novel ensemble learning model to accurately detect potential fraudulent vehicle insurance claims. By leveraging advanced machine learning techniques and addressing the challenge of imbalanced data, our model aims to enhance fraud detection efficiency and reduce financial losses. Our approach combines a stacking ensemble learner with carefully selected base classifiers, meta-classifier and data pre-processing techniques. We evaluate the model's performance on a real-world dataset of insurance claims, demonstrating superior results compared to existing methods. Notably, our model achieves high accuracy, recall, precision and area under curves, ensuring comprehensive detection and minimizing false positives.