Area-level Data Research Articles

Abstract C009: An exploratory ecologic analysis of the association of neighborhood obesogenic factors with prostate cancer mortality in Chicago, Illinois: Challenges and opportunities

Abstract Introduction: Obesity has been associated with aggressive prostate cancer (PCa). Environments that promote obesity, known as obesogenic environments, may also contribute to PCa progression at the population level. However, studies exploring these factors are limited. The purpose of this study was to examine the association between neighborhood obesogenic factors with PCa mortality rates in Chicago. Methods: We used an ecological study design with Chicago Community Areas (n=77) serving as the geographical area unit of analysis. All community area-level data were obtained from the Chicago Health Atlas public data portal. For each area, two obesogenic factors were examined: food environment and park access. The number of healthy and unhealthy grocery stores was used to characterize the food environment. The retail food establishments data was retrieved for the year 2013. The stores were classified into chain grocery stores, non-chain grocery stores, liquor stores, and convenience stores which included dollar tsores, corner stores and gas stations. Chain and non-chain grocery stores were grouped as healthy food access, and other (convenience store, dollar store and liquor store) as unhealthy food access. The area of public parks per 100,000 community area population was used as a measure of park access. An exclusive park area for each community was calculated by intersecting the community area and park area shape files from the year 2016. The outcome variable was community area-level PCa mortality rate age-adjusted to the community area male population. To normalize the data, all included variables were standardized to zero mean and unit standard deviation. Linear regression models analyzed the association of the standardized measures of obesogenic environment (individually and simultaneously) with age-adjusted rates using the Statistical Analysis Software 9.4. Results: The mean number of healthy (0 - 11) and unhealthy (0 - 10) food stores in each community area were 3.1 and 2.6 respectively and mean park area (14.0 - 4096.5 acres) available in each community area was 453.5 acres per 100,000 population. The mean (0 – 29.6) PCa mortality was 9.8 per 100,000 population. Healthy food access was inversely associated with PCa mortality rates, where for every for one 1-standard deviation increase in healthy food access standard deviation, there is a 0.40 unit decrease in community area PCa mortality rate standard deviation (p=0.002) while accounting for the other obesogenic factors (unhealthy food access and park access). Conclusion: The findings of this study suggest that the neighborhood access to healthy food may be a determinant of neighborhood PCa mortality rates. Ecological studies are helpful to understand the impact at the geographical level. Methodological challenges and opportunity in ecological analyses, such as addressing bias and multivariable control of confounding, will be discussed. Citation Format: Ravneet Kaur, Jiehuan Sun, Nebiyou Y. Tilahun, Vincent L. Freeman. An exploratory ecologic analysis of the association of neighborhood obesogenic factors with prostate cancer mortality in Chicago, Illinois: Challenges and opportunities [abstract]. In: Proceedings of the 17th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2024 Sep 21-24; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2024;33(9 Suppl):Abstract nr C009.

Catchment area and cancer population health research through a novel population-based statewide database: ascoping review.

Population-based linked datasets are vital to generate catchment area and population health research. The novel Cancer Information and Population Health Resource (CIPHR) links statewide cancer registry data, public and private insurance claims, and provider- and area-level data, representing more than 80% of North Carolina's large, diverse population of individuals diagnosed with cancer. This scoping review of articles that used CIPHR data characterizes the breadth of research generated and identifies further opportunities for population-based health research. Articles published between January 2012 and August 2023 were categorized by cancer site and outcomes examined across the care continuum. Statistically significant associations between patient-, provider-, system-, and policy-level factors and outcomes were summarized. Among 51 articles, 42 reported results across 23 unique cancer sites and 13 aggregated across multiple sites. The most common outcomes examined were treatment initiation and/or adherence (n = 14), mortality or survival (n = 9), and health-care resource utilization (n = 9). Few articles focused on cancer recurrence (n = 1) or distance to care (n = 1) as outcomes. Many articles discussed racial, ethnic, geographic, and socioeconomic inequities in care. These findings demonstrate the value of robust, longitudinal, linked, population-based databases to facilitate catchment area and population health research aimed at elucidating cancer risk factors, outcomes, care delivery trends, and inequities that warrant intervention and policy attention. Lessons learned from years of analytics using CIPHR highlight opportunities to explore less frequently studied cancers and outcomes, motivate equity-focused interventions, and inform development of similar resources.

What drives solar energy adoption in developing countries? Evidence from household surveys across countries

This study investigates household solar energy uptake in developing countries by combining household surveys for 11 countries with area-level data. We use data from World Bank surveys for countries in Africa, Asia, and Central America. Our probit regressions use up to 36,653 household observations and cover actual uptake rather than intentions. The main result shows that households further from capital cities are less likely to have solar home systems. Furthermore, there are strong links between assets and solar uptake across solar types such as solar home systems, solar lighting systems, and solar lanterns. This is an important finding given the small number of prior studies that use actual uptake data for developing countries and the mixed results from prior literature. We do not find evidence that households in sunnier areas are more likely to have solar home systems across countries. This study motivates policymakers to consider greater support for households far from capital cities, in sunnier regions, and with low levels of assets.

Analyzing the Relationship Between Socioeconomic Deprivation and Outpatient Medicare Part D Fluroquinolone Claims in Texas

Background: Only a few studies have assessed the relationship between deprivation and excessive antibiotic use. In Texas, antimicrobial prescription is particularly high compared with the rest of the US. This study analyzed the association between local area socioeconomic deprivation and providers’ fluoroquinolone claim rates among beneficiaries 65 years and older in Texas. Method: This ecological study utilized provider- and area-level data from Medicare Part D Prescribers and the Social Deprivation Index (SDI) repositories. To identify geographic patterns and autocorrelation in and between SDI and fluoroquinolone claims, spatial dependence of these two variables was assessed by bivariate Local Indicators of Spatial Association (LISA) cluster mapping along with the global and local Moran’s I analyses. Negative binomial regression models were employed to evaluate the relationship between provider- and area-level characteristics (prescriber’s gender, specialty, rural-urban community area, beneficiaries' demographics, area-level population, and normalized SDI) and fluoroquinolone claim rates per 1,000 beneficiaries. Result: A total of 11,996 providers were included. There was no spatial dependence between SDI and rates of fluoroquinolone claims in Texas (Global Moran’s I =0.01, P=0.618). Bivariant LISA maps showed 85 high-high and 38 low-low spatial clusters. Higher SDI (incidence rate ratio (IRR) 0.98, 95% confidence interval (CI) 0.97-0.99 per 1-unit increment) and male providers (IRR 0.96, 95%CI 0.94-0.99) were associated with lower claim rates. In contrast, several factors were associated with higher claim rates, including non-metropolitan areas (1.04, 95%CI 1.00-1.09), and practices with a high proportion of male patients (IRR 1.12, 95%CI 1.10-1.14), Black patients (IRR 1.05, 95%CI 1.03-1.07), or Medicaid beneficiaries (IRR 1.15, 95%CI 1.12-1.17). Effect modification was observed between SDI and rurality, with higher SDI in non-metropolitan areas associated with higher claim rates, whereas SDI in metropolitan areas was inversely related to claim rates. Conclusion: This study showed that the distribution of high and low SDI and rates of fluoroquinolone claims were more geographically clustered than expected by random chance alone. Lower fluoroquinolone claim rates among Texas Medicare providers were seen in metropolitan areas with higher SDI, indicating potential barriers to care. Conversely, higher claim rates were observed in rural areas with higher SDI, signifying a possible knowledge or attitude gap towards fluoroquinolone use. These findings provide opportunities for public health professionals to explore gaps in the knowledge and attitudes of patients and providers related to antimicrobial use, particularly in rural regions, and investigate barriers to healthcare access in metropolitan areas.

Individual, social and area level factors associated with older people's walking: Analysis of an UK household panel study (Understanding Society)

BackgroundAmong older people, walking is a popular and prevalent activity. Walking is key to increasing physical activity levels and resulting physical and mental health. In the context of rapidly ageing populations, it is important to better understand what factors are associated with walking among older people, based on the socioecological model of health. MethodsWe used data from Understanding Society (n:6450), a national panel survey of UK adults aged 65 years and over living in Great Britain. Slope Indices of Inequality (SII) were calculated for weekly walking hours for older people according to individual, social and area characteristics. These include health, loneliness and social isolation, previous walking and sporting activity, residential self-selection, contact with neighbours, number of close friends and social activity. Spatial area-level data described local area crime, walkability, and proximity to retail, greenspace, and public transport amenities. ResultsMultivariable models indicated that poor health, particularly requiring help with walking, was the strongest predictor of weekly walking hours (SII (95% CI) comparing those needing help vs. no help: −3.58 (−4.30, −2.87)). However, both prior sporting activity (most vs. least active: 2.30 (1.75, 2.88)) and walking for pleasure (yes vs. no: 1.92 (1.32, 2.53)) were strongly associated with increased walking several years later. Similarly having close friends (most vs. fewest, 1.18 (0.72, 1.77)) and local retail destinations (any vs. none: 0.93 (0.00, 1.86)) were associated with more weekly walking. ConclusionsPast engagement in physical activity and walking for pleasure are strong predictors of walking behaviour in older people, underscoring the importance of implementing and sustaining walking interventions across the lifespan to ensure continued engagement in later years and the associated health benefits. However, poor health significantly impedes walking in this demographic, emphasising the need for interventions that offer both physical assistance and social support to promote this activity.

Changes in monthly surface area, water level, and storage of 194 lakes and reservoirs in the Yangtze River Basin during 1990–2021 using multisource remote sensing data

Long-term, high spatiotemporal resolution of surface water area, water level, and storage changes in the Yangtze River Basin (YRB) has great scientific and practical importance for improving the management of water resources. Here, three distinct area estimations were first derived using the water classification enhancement method, automated water extraction method based on random forest, and the modified normalized difference water index. The optimized area data was determined by comparing against Sentinel-2 with the minimum root mean square error. A new area data was constructed with the optimized area as the primary data, while the remaining datasets were employed to fill in gaps. The elevation-area relationship was used to derive monthly water level. Changes in water storage were calculated by applying the pyramidal frustum formula from surface water area and water level data. Finally, a new comprehensive dataset of the monthly area, level, and storage changes in the 119 lakes and 75 reservoirs across the YRB with area larger than 10 km2 from 1990 to 2021 were first reconstructed. The spatiotemporal trends of surface water area/level/storage in lakes and reservoirs over 11 sub-basins of the YRB were quantified from 1990 to 2021, as well as before (1990–2003) and after (2003−2021) the construction of the Three Gorges Dam (TGD). During 1990–2021, there was a marked decrease in surface water area/level/storage in most of the YRB sub-basins, which contain 79 % of the lakes and 30 % of the reservoirs. After TGD was constructed, the surface water in lakes decreased by 10 %, while that of reservoirs remained consistent with the pre-construction. The surface water area/level/storage in the lower sub-basins of YRB exhibited a decline to an upward trend before and after the construction of TGD. This study provides a new comprehensive dataset for understanding the dynamic changes of water resource and climate change.

The role of inequalities in managing symptoms of menstruation: harnessing shopping data to innovate female reproductive health research

Introduction & BackgroundMenstruation affects half the population, yet its patterns and management are greatly under-researched. A regular and functioning menstrual cycle is considered an important vital sign and menstrual-related issues can be strong indicators of both reproductive and wider health issues. This project explores a novel health data source - shopping data history - to study how individuals manage menstrual symptoms such as pain, intensity of flow, mental health, and other issues, and explore potential social inequalities in these management strategies. Objectives & ApproachWe present a conceptual framework of studying management of menstruation symptoms using shopping data. The core objectives of this research are to enhance our understanding of menstrual management strategies and potential inequalities in these whilst evaluating the utility and acceptability of shopping data for future female reproductive health research. Our research will focus on harnessing loyalty card data from UK supermarkets and pharmaceutical retailers to provide insights into the management of menstrual symptoms at a national level. We will study retail data to identify products and patterns of purchasing which may be relevant to menstrual management and conduct surveys for linkage with shopping data. The public will be consulted to investigate attitudes towards shopping data for health research and inform interpretations of patterns in the data. Relevance to Digital FootprintsThis project contributes to advancing of understanding of using digital data for health research on an important societal challenge. We investigate the practical applications for menstrual health and other female reproductive health issues, with scope to enact meaningful change. Conclusions & ImplicationsBy analysing shopping behaviour, combined with survey data and area-level socioeconomic data, we aim to identify regions of the UK and individual characteristics which influence the risk of experiencing menstrual symptoms and ability to manage these within a high-income context. Our research will contribute to understanding of menstrual management strategies for women and people who menstruate, and associated inequalities.

The extent and growth of prior authorization in Medicare Advantage.

To assess trends in the use of prior authorization requirements among Medicare Advantage (MA)plans. Descriptive quantitative analysis. Data were from the CMS MA benefit and enrollment files for 2009-2019, supplemented with area-level data on demographic and provider market characteristics. For each service category, we calculated the annual share of MA enrollees in plans requiring at least some prior authorization and plotted trends over time. We mapped the county-level share of MA enrollees exposed to prior authorization in 2009 vs 2019. We quantified the association between local share of MA enrollees exposed to prior authorization and characteristics of that county in the same year. Finally, we plotted the share of MA enrollees exposed to prior authorization requirements over time for the 6 largest MA carriers. From 2009 to 2019, the share of MA enrollees in plans requiring prior authorization for any service remained stable. By service category, the share of MA enrollees exposed to prior authorization ranged from 30.7% (physician specialist services) to 72.2% (durable medical equipment) in 2019, with most service categories requiring prior authorization more often over time. Several area-level demographic and provider market characteristics were associated with prior authorization requirements, but these associations weakened over time. The use of prior authorization varied widely across plans. In 2019, roughly 3 in 4 MA enrollees were in a plan requiring prior authorization. Service-level, area-level, and carrier-level patterns suggest a wide range of approaches to prior authorization requirements.

Geographic inequalities in need and provision of social prescribing link workers a retrospective study in primary care.

Long-term health conditions are major challenges for care systems. Social prescribing link workers have been introduced via primary care networks (PCNs) across England since 2019 to address the wider determinants of health by connecting individuals to activities, groups, or services within their local community. To assess whether the rollout of social prescribing link workers was in areas with the highest need. A retrospective study of social prescribing link workers in England from 2019 to 2023. Workforce, population, survey, and area-level data at the PCN-level from April 2020 to October 2023 were combined. Population need before the rollout of link workers was measured using reported lack of support from local services in the 2019 General Practice Patient Survey. To assess if rollout reflected need, linear regression was used to relate provision of link workers (measured by full-time equivalent [FTE] per 10 000 patients) in each quarter to population need for support. Populations in urban, more deprived areas and with higher proportions of people from minority ethnic groups had the highest reported lack of support. Geographically these were in the North West and London. Initially, there was no association between need and provision; then from July 2022, this became negative and significant. By October 2023, a 10-percentage point higher need for support was associated with a 0.035 (95% confidence interval = -0.634 to -0.066) lower FTE per 10 000 patients. Rollout of link workers has not been sufficiently targeted at areas with the highest need. Future deployments should be targeted at those areas.

Analyzing the relationship between socioeconomic deprivation and outpatient Medicare Part D fluoroquinolone claim rates in Texas.

Only a few studies have assessed the relationship between deprivation and excessive antibiotic use. In Texas, antimicrobial prescription rates are particularly high compared with the rest of the US. This study analyzed the association between local area socioeconomic deprivation and providers' fluoroquinolone claim rates among beneficiaries 65 years and older in Texas. This ecological study utilized provider- and area-level data from Medicare Part D Prescribers and the Social Deprivation Index (SDI) repositories. Negative binomial regression models were employed to evaluate the relationship between provider- and area-level characteristics (prescriber's gender, specialty, rural-urban community area, beneficiaries' demographics, area-level population, and SDI) and fluoroquinolone claim rates per 1,000 beneficiaries. A total of 11,996 providers were included. SDI (IRR 0.98, 95% CI 0.97-0.99) and male providers (IRR 0.96, 95% CI 0.94-0.99) were inversely associated with claim rates. In contrast, several factors were associated with higher claim rates, including non-metropolitan areas (1.04, 95% CI 1.00-1.09), and practices with a high proportion of male (IRR 1.12, 95% CI 1.10-1.14), Black (IRR 1.05, 95% CI 1.03-1.07), or Medicaid beneficiaries (IRR 1.15, 95% CI 1.12-1.17). Effect modification was observed between SDI and rurality, with higher SDI in non-metropolitan areas associated with higher claim rates, whereas SDI in metropolitan areas was inversely related to claim rates. Lower fluoroquinolone claim rates were observed among Texas Medicare providers in metropolitan areas with higher SDI. Conversely, higher rates were observed in rural areas with higher SDI. More studies are needed to understand the underlying causes of this variation and develop effective stewardship interventions.

Relationship between area-level socioeconomic status and health-related quality of life among cancer survivors.

Area-level socioeconomic status (SES) impacts cancer outcomes, such as stage at diagnosis, treatments received, and mortality. However, less is known about the relationship between area-level SES and health-related quality of life (HRQOL) for cancer survivors. To assess the additive value of area-level SES data and the relative contribution of area- and individual-level SES for estimating cancer survivors' HRQOL, we conducted a secondary analysis of data from a population-based survey study of cancer survivors (the Measuring Your Health [MY-Health] Study). Multilevel multinomial logistic regression models were used to examine the relationships between individual-level SES, area-level SES as measured by the Centers for Disease Control and Prevention's Social Vulnerability Index, and HRQOL group membership (high, average, low, or very low HRQOL). Area-level SES did not significantly increase model estimation accuracy compared to models using only individual-level SES. However, area-level SES could be an appropriate proxy when the individual-level SES is missing.

AGING WELL: OLDER ADULTS AND DATA MAPPING

Abstract This presentation will feature how the National Council on Aging (NCOA) has advanced a program of work on data mapping and visualization. It will explain how this work is part of NCOA’s mission to improve the lives of millions of older adults, especially those who are struggling. Guided by the idea that how we age should not be determined by gender, color, sexuality, income, or zip code, the mapping tools display area-level data to highlight different aspects of the lives of low-income older adults. This presentation will describe two of NCOA’s mapping tools and how they can influence policy and advocacy efforts toward aging equity at the local, regional, state, and national levels. The first tool, released in 2021, displays, in part, area-level access to smartphones and high-speed internet among low-income older adults. This map makes visible the invisible by demonstrating low rates of access even in geo-spatial areas with overall high levels of connectivity. The second tool, to be released in the third quarter of 2023, presents sub-state level data to identify gaps between area-level eligibility for each of three essential benefit programs [Supplemental Nutrition Assistance Program (SNAP)], Supplemental Security Income (SSI), and Medicare Savings Program (MSP)] and area-level enrollment rates. This map demonstrates how socio-spatial positions influence access to these national benefits, and how maps can be used to identify regions and communities to prioritize as part of strategic efforts to improve benefit access and promote aging equity.

RACIAL DIFFERENCES IN HEARING LOSS AMONG MIDDLE-AGED AND OLDER ADULTS

Abstract Prior studies have documented how social and systemic factors contribute to racial differences in health; however, less is known about the life course and social determinants of racial differences in hearing loss (HL) among middle-aged and older adults. Evidence from U.S. population-based cohorts indicates that Black adults have better pure-tone thresholds and a 40-70% lower prevalence of HL compared to White adults. However, beyond documenting racial differences in general hearing sensitivity (i.e., degree of HL), it is unknown how the characteristics of HL (e.g., type, laterality, etc.) vary by race. Using ICD codes and the AMCLASS classification system, we will evaluate types of HL (e.g., sensorineural, conductive, and mixed HL); symmetry categories (e.g., symmetrical or asymmetrical HL); and severity categories (by classifying pure-tone average according to standard metrics: normal [-10-15 dB HL], slight [16–25 dB HL], mild [26–40 dB HL], moderate [41–55 dB HL], moderately severe [56–70 dB HL], severe [71–90 dB HL], and profound [>90 dB HL]). We will examine audiological data on 22,905 adults aged 50 and older (17.9% Black, 82.1% White) using electronic health records (EHR) and linked area-level data from a large healthcare system. We will also investigate how socioeconomic (e.g., education, income, health insurance), psychosocial (e.g., mental health), behavioral (e.g., smoking), and clinical factors (e.g., disease comorbidities) are associated with HL characteristics in White and Black adults. Findings will provide important and actionable insights into racial differences in HL.

Mapping the prevalence of cancer risk factors at the small area level in Australia

BackgroundCancer is a significant health issue globally and it is well known that cancer risk varies geographically. However in many countries there are no small area-level data on cancer risk factors with high resolution and complete reach, which hinders the development of targeted prevention strategies.MethodsUsing Australia as a case study, the 2017–2018 National Health Survey was used to generate prevalence estimates for 2221 small areas across Australia for eight cancer risk factor measures covering smoking, alcohol, physical activity, diet and weight. Utilising a recently developed Bayesian two-stage small area estimation methodology, the model incorporated survey-only covariates, spatial smoothing and hierarchical modelling techniques, along with a vast array of small area-level auxiliary data, including census, remoteness, and socioeconomic data. The models borrowed strength from previously published cancer risk estimates provided by the Social Health Atlases of Australia. Estimates were internally and externally validated.ResultsWe illustrated that in 2017–2018 health behaviours across Australia exhibited more spatial disparities than previously realised by improving the reach and resolution of formerly published cancer risk factors. The derived estimates revealed higher prevalence of unhealthy behaviours in more remote areas, and areas of lower socioeconomic status; a trend that aligned well with previous work.ConclusionsOur study addresses the gaps in small area level cancer risk factor estimates in Australia. The new estimates provide improved spatial resolution and reach and will enable more targeted cancer prevention strategies at the small area level. Furthermore, by including the results in the next release of the Australian Cancer Atlas, which currently provides small area level estimates of cancer incidence and relative survival, this work will help to provide a more comprehensive picture of cancer in Australia by supporting policy makers, researchers, and the general public in understanding the spatial distribution of cancer risk factors. The methodology applied in this work is generalisable to other small area estimation applications and has been shown to perform well when the survey data are sparse.

Machine learning and data augmentation in the proxy means test for poverty targeting

Recent years have seen increased interest in the use of alternative data sources in the definition and production of official statistics and indicators for the UN Sustainable Development Goals. In this paper, we consider the application of data science to the production of official statistics, illustrating our perspective through the use of poverty targeting as an application. We show that machine learning can play a central role in the generation of official statistics, combining a variety of types of data (survey, administrative and alternative). We focus on the problem of poverty targeting using the Proxy Means Test in Indonesia, comparing a number of existing statistical and machine learning methods, then introducing new approaches in the spirit of small area estimation that utilize area-level features and data augmentation at the subdistrict level to develop more refined models at the district level, evaluating the methods on three districts in Indonesia on the problem of estimating 2020 per capita household expenditure using data from 2016–2019. The best performing method, XGBoost, is able to reduce inclusion/exclusion errors on the problem of identifying the poorest 40% of the population in comparison to the commonly used Ridge Regression method by between 4.5% and 13.9% in the districts studied.

Do Attitudes Towards Immigrants Matter? The Subjective Wellbeing of Immigrants in England and Wales and Their Exposure to Non-migrants

The wellbeing of immigrants is affected by those around them and the context in which they live. Yet we still know relatively little about the impact that attitudes towards immigrants (ATI) have on immigrants’ life satisfaction, nor do we know the routes by which it manifests. By matching individual data from the UK Understanding Society study to area-level data on ATI for England and Wales from the 2018 European Values Study, I examine whether subnational ATI are associated with immigrants’ life satisfaction. If so, I aim to determine the geographical level at which it is prominent and identify the channels through which this association operates. By exploiting the different geographical scales at which ATI are aggregated, I show within-country variation in ATI. Controlling for contextual- and individual-level characteristics, I find that immigrants’ wellbeing is sensitive to exposure to the negative ATI of non-migrants at the regional level but not at the municipal level. Theoretically identified channels (local social cohesion and ethnic composition) are not drivers of this association, but it is moderated by (interethnic) friendships. Further, I show that ATI are a measure of environment rather than a function of intergroup contact or exposure and that the entire composition of the ATI in an area is more important than the most negative attitudes. I discuss the implications of these findings.

Building Data Infrastructure for Disease-Focused Health Economics Research.

Data infrastructure for cancer research is centered on registries that are often augmented with payer or hospital discharge databases, but these linkages are limited. A recent alternative in some states is to augment registry data with All-Payer Claims Databases (APCDs). These linkages capture patient-centered economic outcomes, including those driven by insurance and influence health equity, and can serve as a prototype for health economics research. To describe and assess the utility of a linkage between the Colorado APCD and Colorado Central Cancer Registry (CCCR) data for 2012-2017. This cohort study of 91,883 insured patients evaluated the Colorado APCD-CCCR linkage on its suitability to assess demographics, area-level data, insurance, and out-of-pocket expenses 3 and 6 months after cancer diagnosis. The linkage had high validity, with over 90% of patients in the CCCR linked to the APCD, but gaps in APCD health plans limited available claims at diagnosis. We highlight the advantages of the CCCR-APCD, such as granular race and ethnicity classification, area-level data, the ability to capture supplemental plans, medical and pharmacy out-of-pocket expenses, and transitions in insurance plans. Linked data between registries and APCDs can be a cornerstone of a robust data infrastructure and spur innovations in health economics research on cost, quality, and outcomes. A larger infrastructure could comprise a network of state APCDs that maintain linkages for research and surveillance.

Enabling real-world cancer care quality surveillance and analytics through a novel population-based statewide linked database.

545 Background: Population-based linked datasets are vital tools to generate real-world evidence (RWE) that can inform interventions and policies to improve cancer care quality and outcomes. The novel University of North Cancer Information & Population Health Resource (CIPHR) links statewide cancer registry data, public and private health insurance claims, and provider- and area-level data since 2003, representing more than 80% of North Carolina’s large, diverse cancer patient population. This scoping review included all articles that have used CIPHR data (N=44) to characterize the breadth of RWE generated and to identify further opportunities for population-based care quality surveillance. Methods: Articles published between 2012-2022 were categorized by cancer site(s) of focus, exposure variables, and outcomes examined across the cancer care continuum. For each article, notable results and conclusions were extracted and summarized, including statistically significant associations between patient-level factors (e.g., demographics, clinical variables, insurance type, distance to providers), provider-level factors (e.g., specialties, facility type, networks), system/policy-level factors and outcomes. Results: Of the 44 articles included, findings were reported across 23 unique cancer sites; 9 articles reported results aggregated across multiple cancer sites. Across the cancer care continuum, the majority of articles’ outcomes were categorized as focused on treatment and survivorship (n=25), with most examining treatment initiation and/or adherence (n=13). Other common outcomes included costs of care (n=8) and mortality/survival (n=8). Fewer articles focused on cancer screening (n=2), recurrence (n=1), long-term care use (n=1), or incidence (n=1). The cancer site most frequently examined was breast cancer (n=19), followed by colorectal (n=6), cervical (n=5), and uterine cancers (n=5). The majority of articles focused on racial/ethnic, rural/geographic, and socioeconomic inequities in care and outcomes, as well as multilevel predictors of observed trends. Conclusions: These findings demonstrate the value of creating and sustaining robust, longitudinal, population-based databases to facilitate the generation of RWE that can be used to monitor cancer care delivery trends, including cancer inequities that warrant intervention and policy attention. Lessons learned from more than a decade of analytics utilizing this unique data resource highlight opportunities to explore less frequently studied cancers and outcomes; motivate targeted implementation of equity-focused interventions and policies responding to trends; and inform development of similar resources in other states.

Association between area-level walkability and glycated haemoglobin: a Portuguese population-based study

Abstract Diabetes is responsible for a high burden of disease, yet there is a great potential for prevention. Physical inactivity is one of the major risk factors for type 2 diabetes which can be tackled by increasing area-level walkability. That said, there are still few population-based studies exploring the association between area-level walkability and objective measures of diabetes. The aim of this study is therefore to estimate the association between area-level walkability and individual levels of glycated haemoglobin in the Portuguese adult population. Area-level data required to estimate walkability was obtained from the 2011 census and an updated street map. The walkability index was constructed using measure of residential density, land-use mix and street connectivity. Individual-level health data was obtained from The National Health Examination Survey (INSEF) 2015, a population-based survey representative of Portuguese adult population. We used gamma regression to estimate associations. The regression coefficients showed that living in medium walkable areas reduced the average glycated haemoglobin (Exp(β)=0.906; 95%CI: 0.821, 0.999) in a statistically significant way when compared with least walkable areas. The association was smaller, and not significant, for the third tercile of walkability (Exp(β)=0.919; 95%CI: 0.822, 1.028). Our findings suggest a non-linear protective association of walkability on glycated haemoglobin. These findings could have important policy implications for urban planning, with a goal of preventing diabetes and promoting health by increasing walkability.

A Primer on Clustering of Forest Management Units for Reliable Design-Based Direct Estimates and Model-Based Small Area Estimation

This study employs clustering analysis to group forest management units using auxiliary, satellite imagery-derived height metrics and past wall-to-wall tree census data from a natural, uneven-aged forest. Initially, we conducted an exhaustive exploration to determine the optimal number of clusters k, considering a wide range of clustering schemes, indices, and two specific k ranges. The optimal k is influenced by various factors, including the minimum k considered, the selected clustering algorithm, the clustering indices used, and the auxiliary variables. Specifically, the minimum k, the Euclidean distance metric, and the clustering index were instrumental in determining the optimal cluster numbers, with algorithms exerting minimal influence. Unlike traditional validation indices, we assessed the performance of these optimally defined clusters based on direct estimates and additional criteria. Subsequently, our research introduces a twofold methodology for Small Area Estimation (SAE). The first approach focuses on aggregating forest management units at the cluster level to increase the sample size, thereby yielding reliable design-based direct estimates for key forest attributes, including growing stock volume, basal area, tree density, and mean tree height. The second approach prepares area-level data for the future application of model-based estimators, contingent on establishing a strong correlation between target and auxiliary variables. Our methodology has the potential to enhance forest inventory practices across a wide range of forests where area-level auxiliary covariates are available.