Insure Research Articles

Health care resource utilization and costs across stages of amyotrophic lateral sclerosis in the United States.

People living with ALS (plwALS) experience motor control loss, speech/swallowing difficulties, respiratory insufficiency, and early death. Advancing disease stage is typically associated with a greater burden on the health care system, and delays in diagnosis can result in substantial health care resource utilization (HCRU). To estimate HCRU and cost burden of plwALS across disease stages from a US payer perspective we assessed HCRU and costs in early-, mid-, and late-stage ALS. Using insurance claims data from the IBM MarketScan Databases between January 2013 and December 2019, we identified plwALS as having at least 2 claims at least 27 days apart with an ALS International Classification of Diseases, Ninth or Tenth Revision diagnosis code (335.20/G12.21) or at least 1 ALS diagnosis code and prescription filled for riluzole/edaravone. Eligible plwALS were aged at least 18 years and had at least 12 months of enrollment data before and at least 6 months after the index date (date diagnosis criteria met). plwALS were grouped into disease stages using an ALS severity-based staging algorithm developed using ALS symptom and staging survey data from 142 neurologists reporting on 880 plwALS. The starting date of each severity stage was defined as the first date of an ALS symptom within the early-, mid-, and late-stage categories, respectively. The ending date for a severity stage was defined as the day before the first date of an ALS symptom from a more severe category. plwALS could transition to more severe stages, with reverse transition of severity excluded. Mixed regression modeling was used to assess differences in HCRU and costs per person-year between severity stages, adjusted for age and sex. 2,273 plwALS were included in the total ALS study sample, with 1,215 early-stage, 1,511 midstage, and 1,186 late-stage plwALS. 90% of early-stage plwALS had ALS symptoms before diagnosis, and 27% of late-stage plwALS had a late-stage symptom before diagnosis. In the evaluation period, later-stage ALS groups had more overall hospital admissions (early = 0.15, middle = 0.23, and late = 0.74; P < 0.01), outpatient visits/service (early = 26.81, middle = 32.78, and late = 48.54; P < 0.01), emergency department visits (early = 0.46, middle = 0.69, and late = 1.03; P < 0.01), and total prescription count (early = 9.23, middle = 11.37, and late = 12.72; P < 0.01) over 12 months. Annualized costs increased as ALS progressed (early = $31,411, middle = $51,481, and late = $121,903; P < 0.01), which was primarily driven by higher frequency of and cost per hospital admission. HCRU and costs increased with ALS progression, with diagnosis frequently occurring even after experiencing late-stage symptoms. These findings highlight the potential value of delaying progression into a more resource-intensive stage by diagnosing and adequately treating plwALS earlier in the disease course.

Epidemiology of Acute Compartment Syndrome After Pediatric Tibial Tubercle and Tibial Shaft Fractures.

While acute compartment syndrome (ACS) is a well-reported complication after pediatric tibial shaft fractures, prior literature has suggested that pediatric patients with tibial tubercle fractures may be at increased risk of ACS due to the proximity of the tibial tubercle to the anterior tibial recurrent artery. However, this theory was largely based on a series of early case reports without substantiation in larger-scale studies. The purpose of this study is to conduct a population-level analysis of the incidence and risk factors of acute compartment syndrome following pediatric tibial tubercle and tibial shaft fractures. We hypothesize that the rate of ACS would be lower in patients with tibial tubercle fractures when compared with those with tibial shaft fractures. A retrospective cohort analysis of the PearlDiver Mariner database was performed by querying all patients diagnosed with tibial tubercle and tibial shaft fractures between January 2010 and October 2022. Matched cohorts (n=25,483) of patients with pediatric tibial tubercle and tibial shaft fractures were captured using International Classification of Diseases, Ninth Revision (ICD-9), Tenth Revision (ICD-10) billing codes, and age as inclusion/exclusion criteria. Rates of subsequent compartment syndrome were calculated by querying for insurance claims with associated CPT codes for fasciotomy and/or ICD-9/ICD-10 billing codes for compartment syndrome. Student t test and χ2 analyses were used to compare demographics between the tibial tubercle and tibial shaft cohorts. The rate of ACS was significantly lower in the tibial tubercle cohort (0.46%) than in the tibial shaft cohort (0.70%, P<0.001). Male sex and increased age were associated with an increased risk of developing ACS in both cohorts (P<0.001). In addition, polytrauma was found to be a risk factor for ACS among patients with tibial shaft fractures (P<0.001). We report a low rate of ACS following both pediatric tibial tubercle fractures and pediatric tibial shaft fractures. Among patients with tibial tubercle fractures, male sex, and increased age were noted to be risk factors for ACS, whereas male sex, older age, and polytrauma were risk factors for ACS among patients with tibial shaft fractures.

Healthcare utilization among foreign beneficiaries of the National Health Insurance program in Korea.

To explore the factors related to healthcare utilization (both inpatient and outpatient services) among foreign beneficiaries of the National Health Insurance (NHI) in Korea. The total foreign beneficiaries of the NHI in Korea 2022 were included (n = 1,386,329). For the dependent variables, utilization and either hospital stays or count of visits were used. And the independent variables were selected guided by Andersen's behavioral framework: predisposing characteristics (e.g., sex, age, nationality), enabling characteristics (e.g., insurance premium ranking, residency status, residency period, place of residency, beneficiary type), and need characteristics (e.g., disability, CCI, chronic diseases). We conducted a two-part model analysis employing Andersen's behavioral framework, using health insurance claims data from 2022. The results indicated that individual need characteristics, such as disability, CCI, and chronic diseases, had the most significant impact on total healthcare utilization and its associated costs. The next significant characteristic was the residence period; the longer the period, the higher the likelihood of healthcare utilization, but the lower the associated costs. Among the predisposing characteristics, individuals aged 65 years and older had the most significant impact on healthcare costs. Additionally, among the enabling characteristics, residency period, residency status, and beneficiary type were most significant. Foreigners residing in Korea are eligible to use the Korean healthcare system by enrolling in NHI. As the foreign population is projected to grow even faster due to Korea's aging society and low birth rate, this study can assist public health administrators in formulating evidence-based policies.

A Million Person Study Innovation: Evaluating Cognitive Impairment and other Morbidity Outcomes from Chronic Radiation Exposure Through Linkages with the Centers for Medicaid and Medicare Services Assessment and Claims Data.

The study of One Million U.S. Radiation Workers and Veterans, the Million Person Study (MPS), examines the health consequences, both cancer and non-cancer, of exposure to ionizing radiation received gradually over time. Recently the MPS has focused on mortality patterns from neurological and behavioral conditions, e.g., Parkinson's disease, Alzheimer's disease, dementia, and motor neuron disease such as amyotrophic lateral sclerosis. A fuller picture of radiation-related late effects comes from studying both mortality and the occurrence (incidence) of conditions not leading to death. Accordingly, the MPS is identifying neurocognitive diagnoses from fee-for-service insurance claims from the Centers for Medicare and Medicaid Services (CMS), among Medicare beneficiaries beginning in 1999 (the earliest date claims data are available). Linkages to date have identified ∼540,000 workers with available health information. Such linkages provide individual information on important co-factor and confounding variables such as smoking, alcohol consumption, blood pressure, obesity, diabetes and many other health and demographic characteristics. The total person-level set of time-dependent variables, outcomes, organ-specific dose measures, co-factors, and demographics will be massive and much too large to be evaluated with standard software. Thus, development of specialized open-source software designed for large datasets (Colossus) is nearly complete. The wealth of information available from CMS claims data, coupled with individual dose reconstructions, will thus greatly enhance the quality and precision of health evaluations for this new field of low-dose radiation and neurocognitive effects.

The impact of daily step count with mobile health apps on clinical outcomes among general population

Abstract Background/Introduction Numerous studies have investigated the correlation between daily step count and the reduction of all-cause mortality and cardiovascular events. However, the optimal daily step count for the general population remains controversial. Purpose This study aims to assess cardiovascular risk based on individual step counts using a large cohort that includes step count data from mHealth Apps (Kencom), health checkups databases, and Japanese insurance claims databases. Methods We analyzed data from 85,204 users registered with Kencom between January 2016 and September 2021. The mean step count during one year before and after baseline data collection was used to define each individual's step count. We divided participants into low (N=45,539) and high (N=39,665) step count groups based on their mean step counts. Additionally, we categorized users into five groups based on their step counts (&amp;lt;4000; 4000-5999; 6000-7999; 8000-9999; and ≥10000) to evaluate cardiovascular events. The primary outcome was the occurrence of cardiovascular events, including myocardial infarction, stroke, angina pectoris, heart failure, and atrial fibrillation, as classified by the International Classification of Diseases and Related Health Problems, 10th Revision (ICD-10). The cumulative incidence was estimated over 5 years. Results The high step count group was generally younger and predominantly male. No significant differences in cardiovascular risk factors such as hypertension, dyslipidemia, diabetes mellitus, and current smoking status were found between the two groups. Stratifying baseline characteristics by step count revealed significant differences among the five groups, with the exception of sex. The cumulative incidence of cardiovascular events was significantly lower in the high step count group compared to the low step count group (6.5% vs. 7.5%, log-rank P=0.005, Figure 1). After adjusting for confounding factors, the risk reduction of the high step count group remained significant (HR: 0.92, 95% CI: 0.87-0.98, P=0.005). Among the five step count groups, the incidence of primary outcome significantly differed, with only Group 5 (≥10000 steps) showing a significant reduction in the risk of cardiovascular events compared to Group 1 (&amp;lt;4000 steps) after adjustment (HR: 0.87, 95% CI: 0.77-0.99, P=0.03). Subgroup analyses based on sex indicated that the risk reduction was significant in males in Group 5 but not in females. Conclusion The risk of cardiovascular events was higher in the low step count group compared to the high step count group. After adjusting for confounders, only the highest step count group (≥10,000 steps/day) showed a significant reduction in cardiovascular event risk compared to the lowest step count group.

Nationwide prevalence and trends in high-risk drinking among adult men the disabled and non-disabled

Abstract Background Excessive drinking is an important public issue that causes various levels of physical, psychological, behavioral problems. Especially for the disabled, it matters in that it could lead to poor health and secondary health conditions. Methods This study analyzed high-risk drinking (HRD) behaviors over a 9-year period in adult South Korean men according to disability presence, severity, type. We linked national disability registration data with the National Health Insurance claims data. Age-standardized prevalence of HRD were analyzed for each year and the odds of current HRD were determined by multivariate logistic regression after adjusting for income level, residence, insurance type, Charlson comorbidity index. Results The age-standardized HRD prevalence of the mildly disabled was higher than that among the non-disabled. Among the mildly disabled and physically disabled, younger aged group showed higher prevalence of HRD than that of non-disabled. In multivariate logistic regression analysis, the mildly (aOR=1.26, 95% CI = 1.24-1.29), physically (aOR=1.31, 95% CI = 1.28-1.33), communication disabled (aOR=1.12, 95% CI = 1.07-1.17) had higher prevalence of HRD than the non-disabled. Conclusions The prevalence of HRD was higher among the young mildly disabled than the non-disabled, this was more evident in young men physically disabled, communication disabled. This may be a result of the physical and social limitations of severely disabled and should be interpreted with caution. Social activities should be supported, and health policies on alcohol consumption should be implemented to prevent secondary conditions and reduce health disparities. Funding: This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Ministry of Education (No. 2022R1I1A3070074) and by a grant of Patient-Centered Clinical Research Coordinating Center (PACEN) funded by the Ministry of Health &amp; Welfare, Republic of Korea (grant number: HC23C0165). Key messages • The prevalence of high-risk drinking was highest among the mild disabled, especially the mild, physical, communication disabled in younger aged men. • Health policies and efforts to improve health disparities aimed at reducing alcohol consumption among people with disabilities are needed.

The cost-effectiveness of a digitally enabled telerehabilitation program delivered nationally to patients in Australia

Abstract Introduction Traditionally delivered cardiac rehabilitation (CR) programs have considerable benefits, yet access and participation remains low. Emerging digital health models offer promise in healthcare delivery, but their benefits and impact on healthcare utilization and cost-effectiveness are uncertain. We conducted a cost-effectiveness analysis of a telehealth program with a CR mobile app for post-hospitalization coronary heart disease (CHD) patients. Purpose To assess whether a digitally enabled CR program is cost-effective compared with a propensity matched usual care control group in the Australian health system. Methods Patients were recruited nationally, through a health insurer, following hospitalization for CHD. Participants received weekly telehealth consultations for up to 8 weeks, supplemented by use of an mHealth app. Using administrative private health insurance claims data, we assessed differences in hospital costs, mortality rates and bed days at 12-months post discharge. Incremental cost effectiveness ratios (ICERS) were calculated to determine the cost per deaths prevented and cost per bed days saved for the digital-CR group (n=133) compared with usual care (n=266). Results Among the cohort, 36% lived in regional/rural areas, 70% were male and the mean age was 68 years. The mean total hospital all cause readmission costs at 12-months were $1,488 AUD less in the digital-CR group ($16,411) compared to the usual care group ($17,899). Digital-CR had fewer total hospital days relative to the usual care group (median of 2 vs 3 days) and patients in the digital-CR group had a significantly lower mortality rate at 12-months post discharge. The ICERs were -$337 per 100 patient deaths prevented and -$3,033 per median bed days saved. Conclusion The implementation of a digital-CR program demonstrates cost-effectiveness for patients with CHD when compared to usual care. Thus, advocating for the widespread availability of digital-CR programs as an option for CHD patients following hospital discharge is warranted.

Nationwide trends in the suicide mortality among the disabled in Korea from 2006 to 2017

Abstract Aims The disabled are one of the priority populations at risk of suicide, but there has been no study of the factors influencing suicide mortality among the disabled in South Korea. Methods To analyze age-standardized suicide mortality (ICD-10 Codes, X60-84) for each year according to the presence, severity and type of disability, we linked three population datasets (National Health Insurance Claims Database, Disability Registry, and Deaths by Cause). Our dataset covers the entire population of South Korea, extended over a 12-year period from 2006 to 2017. The analyses were stratified by sex, and hazard ratios for suicide were determined by multivariate logistic regression after adjustment for socio-demographic and clinical variables. Results The suicide rate among the disabled was 2.24 times higher than among the non-disabled in 2017. The highest suicide rates were observed among intellectual or mental disabled, and severe disabled. In the disabled group, suicide rates were higher among those with medical benefits, older age (over 70 years), and depressive disorders in both male and female. Conclusions Although the disabled people’s suicide is often seen as an individual problem, it actually affects their families, communities and society. Therefore, government must strive to emphasize the broad view of the disabled population - not just the health of individuals, and track suicide trends and identifying risk and protective factors for suicidal behavior. From this information, suicide prevention strategies for the disabled people should be evaluated to identify the most effective interventions. Financial support: This research was supported by the Basic Science Research Program through the National Research Foundation of Korea funded by the Ministry of Education (No. 2022R1I1A3070074, and 2022R1I1A1A01068449), and a grant of Patient-Centered Clinical Research Coordinating Center funded by the Ministry of Health &amp; Welfare, Republic of Korea (No. HC23C016500). Key messages • The suicide rate among the disabled was 2.24 times higher than among the non-disabled in 2017. • In the disabled group, suicide rates were higher among those with medical benefits, older age (over 70 years), and depressive disorders in both male and female.

Disparities in Mental Health and the Social Determinants of Health Among Privately Insured Transgender Patients in the United States

Abstract Objectives The purpose of this study was to examine disparities in mental health and the social determinants of health (SDoH) among privately insured transgender and gender minority (TGM) patients in the United States. Methods Data came from FAIR Health, a repository of private health insurance claims data covering over 150 million individuals in the United States. TGM patients (n = 274,479) were identified with medical claims that included a diagnosis for gender dysphoria or gender identity disorder. A random 5% selection of cisgender patients (n = 7,456,823) was used as a comparison group. We estimated descriptive statistics and used multivariable logistic regression models to compare health outcomes between TGM and cisgender patients. Results TGM patients were more likely to have anxiety (odds ratio [OR]=7.02; 95% confidence interval [CI]=6.94-7.09), depression (OR = 9.27; 95% CI = 9.16-9.37), post-traumatic stress disorder (OR = 9.03; 95% CI = 8.86-9.21), and substance use disorder (OR = 2.93; 95% CI = 2.86-3.00) compared to their cisgender peers. TGM patients were also more likely to have problems with education and literacy (OR = 5.95; 95% CI = 5.64-6.28), employment (OR = 3.67; 95% CI = 3.44-3.92), economic circumstances (OR = 5.74; 95% CI = 5.39-6.12), and social exclusion (OR = 7.35; 95% CI = 6.91-7.81). Compared to TGM patients living in states with transphobic policy environments, TGM patients in states with comprehensive protections from discrimination fared better. Conclusions We found wide disparities in mental health and the SDoH among privately insured TGM. More research is critically needed to better understand the health and social needs of TGM populations. Meanwhile, policymakers should consider enacting policies that would prohibit discrimination against TGM people throughout the SDoH -- or the conditions where we live, learn, work, play, and age. Health care providers should also receive education on the unique health needs of TGM patients to advance health equity.

Development of the Unbiased Disease Network Using Health Insurance Claims Big Data in South Korea

Abstract Background Research on the overall relationship between disease is needed. The disease network model can comprehensively show the complex relationships between diseases and progression patterns. Existing disease network studies were estimated by univariate relative risks. We aim to develop the unbiased disease network. Methods We used NHIS national data of 50 million patients, collected from 2008 to 2022. We used only the primary diagnosis codes (three-digit codes) and defined 603 disease categories based on the special tabulation list for morbidity in the detailed subcategories of ICD-10, prior research, and the frequency of diseases in the dataset. To develop a more unbiased disease network, we first adjusted for age, sex, period, season, and encounter type, then added the confounding disease exposure according to the Directed Acyclic Graph (DAG) and performed Poisson regression to estimate relative risk (RR) applying Inverse Probability Weighting (IPW). We identified statistically significant disease pairs (RR &amp;gt; 1.10, p &amp;lt; 0.001) and repeated the IPW analysis up to the third round because the network would be updated every round. The significant disease pairs identified in the last round were connected to develop the overall network and subnetworks by age and sex. Results We repeated the process up to the third round for all 346,948 disease pairs and fitted 9,024 pairs as significant. We identified general population disease network and subnetwork per age and sex subgroups. Conclusions These disease networks represented the overall progression relationship between diagnosis codes. Key messages • We developed the unbiased disease network that shows progression patterns by adjusting for confounders and using IPW. • The disease network can be used to extract features from various diagnosis codes in EHR/Claims data, analyze causal effects, and predict disease or healthcare usage using RR or graph embedding.

Mother-to-child Transmission using the National HBV Perinatal Prevention Data in Korea, 2002-2021

Abstract Aims To define risk factors associated with mother-to-child transmission (MTCT) and to create supporting evidence for the National Hepatitis B Perinatal Prevention Program (HBPPP) in Korea. Methods We analyzed HBPPP mother-infant pair data enrolled from 2002 to 2021. The major outcome was the HBsAg status of the infant resulting from post-vaccination serological testing. Additionally, we used prescription records of antivirals against the hepatitis B virus (telbivudine, tenofovir disoproxil fumarate) from 2010 to 2021 by data linkage with the National Health Insurance claims data. To select significant risk factors by maternal HBeAg status based on the data from 34 to 42 weeks of pregnancy, multivariate logistic regression analysis was used. Results Data from 232,242 mother-infant pairs were analyzed. Among 154,331 (66.5%) infants with post-vaccination serological testing results, 3,620 (2.3%) were HBsAg-positive. The incidence of MTCT has shown a decreasing trend, from 4.1% in 2002 to 1.0% in 2021. However, the rate was high at 5.5% (1822/32986) for infants of HBeAg-positive mothers. Maternal antiviral prescription during the pregnancy period, old age, HBeAg negative, delivery type (C-section), and breastfeeding were significant factors in decreasing the risk of MTCT. Among the HBeAg-positive, there was a significant difference in the antiviral prescription 0.7% (11/1692) compared to the non-prescription 4.9% (789/16264) [aOR, 7.71; 95% CI, 4.24-14.02]. The result was similar for the HBeAg-negative group, 0.7% (302/44888). Conclusions To eradicate the hepatitis B virus on a national level, MTCT must be further reduced. To achieve this, it is necessary to introduce strong additional preventive measures, including quantitative HBV DNA testing for mothers with chronic hepatitis B and administering antiviral prophylaxis at the third trimester of pregnancy with HBV-DNA ≥200,000 IU/mL into the HBPPP. Key messages • The incidence of mother-to-child transmission has shown a decreasing trend, from 4.1% in 2002 to 1.0% in 2021. • Maternal antiviral prescription, old age, HBeAg negative, delivery type (C-section), and breastfeeding were significantly decrease the risk of mother-to-child transmission.

Prognostic impact of cardiac events in asymptomatic mitral regurgitation

Abstract Background Mitral regurgitation (MR) is the most common valvular heart disease in adults and is associated with increased morbidity and mortality. Treatment recommendations of degenerative MR have been liberalized, however, the role of asymptomatic patients is still insufficiently defined. Purpose The purpose of this study were to evaluate the prevalence of cardiac events in a large asymptomatic MR cohort, based on health insurance data analysis, in comparison with healthy propensity score-matched general population and to define the prognostic impact of such cardiac events on ten-year survival. Methods We retrospectively analyzed health insurance claims data from Germany’s second largest health insurance fund, which maintains longitudinal data on outpatient and inpatient medical services for 8.7 million German residents. We included all patients with an outpatient diagnosis of MR (I34.0) in a minimum of two quarters during a calendar year and first recorded diagnosis between 2008 and 2011. To identify an asymptomatic MR cohort (study group), all patients with at least one complication attributable to MR / mitral valve intervention during the incidence quarter of MR were excluded. Ten-year outcomes were compared between study group vs. age- and sex-matched healthy control group. Cardiac events of interest were new congestive heart failure, new-onset atrial fibrillation, pulmonary hypertension, or cardiac decompensation. Results A total of 56.577 (62.7%) asymptomatic MR patients were identified from the whole cohort of 90.245 individuals with a newly diagnosed MR in the BARMER database. During a ten-year follow-up, 26.282 (46.5%) patients in the study group experienced at least one cardiac event vs. 11.763 (20.8%) individuals in the control group (OR: 3.31, 95%CI: 0.251-0.262, P&amp;lt;0.0001). The occurrence of at least one cardiac event (i.e., congestive heart failure, new-onset atrial fibrillation, pulmonary hypertension or cardiac decompensation) in the study group was associated with a significantly worse ten-year survival as compared to the control group. The risk of cardiac events was age- and sex-dependent, while female patients with an asymptomatic MR had a lower risk as compared to males in all age groups (OR: 0.38, 95%CI: -1.193- -0.754, P&amp;lt;0.0001). Only 1.530 (2.7%) patients underwent mitral valve (MV) intervention in the study group during ten-year follow-up. Conclusions Our study reveals high incidence of asymptomatic MR in a representative cohort of health insurance claims data in Germany. Every second patient with an asymptomatic MR experiences at least one cardiac event during a ten-year follow-up period that is associated with a significantly worse survival. The definition of asymptomatic MR cohort at risk of developing cardiac events may potentially improve early treatment strategies and lead to better outcome in the future.Abstract picture

Contemporary use of beta-blockers in the outpatient treatment of heart failure with atrial fibrillation

Abstract Background Evidence of the effectiveness of β-blockers in heart failure (HF) and atrial fibrillation (AF) in a contemporary cohort is controversial. This study investigated the association between the use of β-blockers and prognosis in outpatients across a broad spectrum of HF with AF. Methods Patients hospitalized with the first episode of acute HF were identified from the National Database of Health Insurance Claims and Specific Health Checkups of Japan between 2014 and 2021. Associations of β-blocker use at discharge and prognosis were compared by propensity score matching among the AF or non-AF group. A multilevel mixed-effects survival model was used with hazard ratio (HR) and 95% confidence interval (CI). Results Among 428,650 patients discharged with HF in 4,433 hospitals, 175,174 (40.9%) were ≥85 years old, 151,873 (35.4%) had complicated AF, and 236,457 (55.2%) were β-blocker users. In a matched AF group, β-blocker use was associated with a lower composite outcome of all-cause mortality or HF rehospitalization (HR [95% CI], 0.95 [0.93–0.97]). A similar result was obtained in a matched non-AF group (0.95 [0.94–0.96]). In addition, the HRs in patients aged ≥85 years and female patients were 1.00 [0.98–1.02] and 1.01 [0.98–1.03] in the AF group and 1.03 [1.01–1.05] and 0.98 [0.97–1.00] in the non-AF group, respectively (Figure). Conclusions The favorable prognostic associations of β-blocker use were observed regardless of AF in outpatients across a broad spectrum of HF in a superaged society. However, β-blocker use may not be effective in older or female patients.

Detecting Unrecognized Dementia Using Deep Learning Methods with Korean National Claims Data

Abstract Background Dementia is one of the leading causes of death in elderly people. In Korea, the government supports dementia screening for people aged 60 and older, but it is not a population-based organized screening program. It is difficult to achieve early diagnosis and early treatment. The objective of this study is to detect unrecognized dementia using Korean National Claims Data for early diagnosis. Methods This study used the National Health Insurance Claims data in Korea. The case group is the new dementia visit as an unrecognized group, while the control group was randomly selected from the general population without dementia. The predictors included health utilization, socioeconomic and demographic, procedure codes, diagnostic codes and health screenings data. The health utilization variables included length of stay, the number of outpatient visits, health expenditures and other encounter information. All medical utilization data was aggregated monthly. In the case group, 12 months of data from 6-month or 2-years before the onset of dementia were used. In the control group, 12 months of data from randomly selected time point were included. Transformer, embedding methods, and Time2Vec were used as the deep learning methods. This study included 453,306 incidents of dementia, representing almost all cases reported from 2010 to 2019 in Korea. The control group consisted of 669,873. Results After training the deep learning prediction model, the AUC for 6-month dementia prediction on the test dataset was 0.87, and for 2-year dementia prediction, it was 0.76. The most important variables were health utilization variables, especially health expenditures. The health utilization trajectories of the case group were significantly different from those of the control group. Conclusions Based on the claims data, we could detect unrecognized dementia case. We can extend the model to other chronic disease such as cancer and stroke, which are significant burden of disease. Key messages • Screening in healthcare facilities or using lab test results would be more accurate, but it would be difficult to apply to a large population due to the cost-effectiveness and timeliness constraints. • On the other hand, the detection model using the existing claims data would be helpful to to develop cost-effective mass screening program. It can cover the weakness of current screening program.

Prevalence of and factors associated with lipid screening in young people aged 16– to 21 years in the United States: analysis of nationwide cross-sectional data

Background & AimsThere is no consensus on whether it is worthwhile for young people to undergo lipid screening. An estimate of the national prevalence of lipid screening in this population can serve as a basic parameter in analyses of the utilization and cost-effectiveness of health care services. Previous studies were mostly based on electronic health records or insurance claims data or were restricted to adolescents at younger ages. This study aimed to investigate the prevalence of lipid screening and the factors associated with it in young people aged 16– to 21 years in the United States based on a nationally representative sample.MethodsA cross-sectional study of 16- to 21-year-old participants from the National Health and Nutrition Examination Survey (NHANES), 2013–March 2020 (four cycles), was conducted. The prevalence of lipid screening uptake was calculated for each cycle as well as for all four cycles combined, and the associations of lipid screening uptake with the covariates were analysed by logistic regression. An appropriate sample weight was incorporated into the analysis as recommended.ResultsA total of 3,600 participants were included in the analysis, and 32.02% of them had ever received lipid screening at the time of the survey, with no statistically significant secular changes observed over the four cycles. Among the participants, 28.83% received lipid screening within the past 5 years. In multivariable regression analysis adjusted for all covariates, non-Hispanic black ethnicity (adjusted odds ratio [AOR]: 1.43), having health insurance (AOR: 1.54), obesity (AOR: 1.65) and sufficient physical activity (AOR: 1.36) were significantly associated with a greater chance of ever receiving lipid screening.ConclusionsNearly one-third of young people aged 16– to 21 years in the United States have received lipid screening. Ethnicity, health insurance, weight problems, and physical activity level were associated with the likelihood of lipid screening uptake. These findings could help people understand the current situation of lipid screening in young people in the United States and may serve as fundamental parameters in the assessment of the utilization and cost-effectiveness of related health care services.

Estimated cost of VEGFR TKI associated adverse events in metastatic renal cell carcinoma patients

IntroductionThe majority of metastatic renal cell carcinoma (mRCC) patients receive one or more VEGFR TKI agents, alone or in combination with an immune-oncology (IO) agent or an mTOR inhibitor. To date, the cost of adverse events (AEs) common to VEGFR TKIs has not been quantified. This study estimated the potential impact of differences in VEGFR TKI AE profiles on treatment cost efficiency in the relapsed/refractory (R/R) setting.MethodsPatients with documented mRCC who were treated with VEGFR TKI therapies between Jan 2015 and Mar 2021 were identified using EMR. ICD-10 diagnosis codes were used to identify the first occurrence of each class effect AE. Patients were matched to 3rd party insurance claims, and costs associated to TKI AEs within 90 days of index event were captured. Average per patient AE cost data was calculated and applied to published incidence data to estimate regimen-specific AE total cost burden within a hypothetical commercial plan for mRCC patients undergoing treatment in the R/R setting.ResultsThe highest total cost for AE management was attributed to lenvatinib and everolimus use at $13,303, followed closely by sunitinib at $13,092. Tivozanib treatment was associated with the lowest total cost of AE management at $7,523, driven by the relatively lower incidence of certain high-cost AEs.ConclusionsThe estimated costs of managing VEGFR TKI class-effect AEs were lowest with tivozanib, and highest with lenvatinib and everolimus, indicating potentially differential healthcare resource burden by TKI regimen. The use of tivozanib in the 3 L + mRCC setting suggests potential costs offsets when compared to other TKI regimens.

Social determinants of health and surgical outcomes of minimally invasive radical prostatectomy: a national population-based study.

Socioeconomic determinants of health (SDOH) are often unvalued during surgery risk stratification; hence, they might be a major source of disparity that can jeopardize outcomes related to urological surgery. The aim of our study is to evaluate the impact of SDOH on postoperative outcomes following minimally invasive radical prostatectomy (MIRP). Patients who underwent MIRP between 2011 and 2021 were retrospectively analyzed by using PearlDiver-Mariner, an all-payer insurance claims database. International Classification of Diseases diagnosis and procedure codes were used to identify patient's characteristics, postoperative complications and SDOH. Outcomes were compared using multivariable regression models. Overall, 100,035 patients (mean age = 63.24 ± 7.07) underwent MIRP. The 60-day postoperative complication rate was 18%. Approximately 6% of patients reported at least one SDOH at baseline. SDOH were associated with higher odds of 60-day postoperative complications (OR:1.24, 95% CI:1.15-1.34), including urinary tract infection (OR:1.32, 95% CI:1.20-1.45) and acute kidney injury (OR:1.31, 95% CI:1.00-1.39). Postoperative urethral stricture (OR:1.37, 95% CI:0.92-1.98) did not reach statistical significance at multivariable analysis. Patients with SDOH have a significantly higher risk of postoperative complications following MIRP, especially urinary infection and acute kidney injury. These findings are multifactorial and should prompt identifying measures that might help prevent this large-scale phenomenon.

National Health Insurance Claim of Pediatric Appendectomy With Mandatory Diagnosis-Related Group Payment System in Korea

IntroductionEnforcement of diagnosis-related group system for appendectomy under the National Health Insurance Service has gradually widened and since July 2013 it has been enforced in all medical institutions in Korea. We have analyzed Health Insurance Review & Assessment Service data to observe changes in claim patterns of pediatric appendectomy during this period. MethodsAll claims data for appendectomy of patients younger than 18 y from 2011 to 2015 were collected. We analyzed the following factors of all cases: age, gender, length of hospital stay, medical cost, method of insurance claim and region. ResultsA total of 112,143 claims were made for appendectomy during the study period. In that, 66,510 (59.3%) were boys and 45,633 (40.7%) were girls. Median length of stay was 5 d and median sum of reimbursement of each claim was 2,198,630 won. Annual number of claims for appendectomy showed a steady decline from 24,888 in 2011 to 19,070 in 2015. Median sum of reimbursement of each claim was 1,862,615, 1,975,500, 2,233,360, 2,376,700, and 2,468,000 won, respectively from 2011 to 2015. Reimbursement for complicated appendectomy increased from 4400 out of 24,888 cases (17.7%) in 2011 to 3865 out of 19,070 cases (20.3%) in 2015. Overall medical cost of all reimbursement for pediatric appendectomy increased from 46,113,202,580 to 47,572,253,300 won. ConclusionsFollowing the universal enforcement of the diagnosis-related group claim system for appendectomy by the National Health Insurance Service, we observed an increase in the median sum of reimbursement per claim and a rise in the rate of claims for complicated appendectomies. These changes were associated with an overall increase in national medical costs.

A Systematic Literature Review of Insurance Claims Risk Measurement Using the Hidden Markov Model

In the rapidly evolving field of insurance, accurate risk measurement is crucial for effective claims management and financial stability. Therefore, this research presented a systematic literature review (SLR) on insurance claims risk measurement using the Hidden Markov Model (HMM). Bibliometric analysis was conducted using VOSviewer 1.6.20 and ResearchRabbit software to map research trends and collaboration networks in this topic. This review explored the implementation of the HMM in predicting the frequency and severity of insurance claims, with a focus on the statistical distribution methods used. In addition, the research emphasized the influence of the number of hidden states in the HMM on claims behavior, both in terms of frequency and magnitude, and provided interpretations of these hidden dynamics. Data sources for this review comprised three databases, namely, Scopus, ScienceDirect, and Dimensions, and additional papers from a website. The article selection process followed updated PRISMA 2020 guidelines, resulting in twelve key papers relevant to the topic. The results offered insights into the application of the HMM for forecasting the frequency and severity of insurance claims and opened avenues for further investigation on distribution models and hidden state modeling.

Increased Spending on Low-Value Care During the COVID-19 Pandemic in Virginia

Abstract Characterizing the value and equity of care delivered during the COVID-19 pandemic is crucial to uncovering health system vulnerabilities and informing post-pandemic recovery. We used insurance claims to evaluate low-value (no clinical benefit, potentially harmful) and clinically-indicated utilization of a subset of 11 ambulatory services within a cohort of ∼2 million Virginia adults during the first two years of the pandemic (3/1/2020-12/31/21). In 2020, low-value and clinically-indicated utilization decreased similarly, while in 2021, low-value and clinically-indicated utilization were 7% higher and 4% lower than pre-pandemic rates, respectively. Extrapolated to Virginia’s population of insured adults, ∼$1.3 billion in spending was associated with low-value utilization of a subset of ambulatory services, reflecting 6% higher spending than pre-pandemic rates in 2021. Compared with pre-pandemic rates, low-value and clinically-indicated utilization of the 11 services were 15% and 16% lower, respectively, during 3/1/2020-12/31/21 among patients with the greatest socioeconomic deprivation. Meanwhile, low-value and clinically-indicated utilization remained closer to pre-pandemic rates among patients with the least socioeconomic deprivation during the same timeframe. These results highlight widening healthcare disparities and underscore the need for policy-level efforts to address the complex drivers of low-value care and equitably redistribute expenditures to services that enhance health.