LLLan Luo

Lan Luo

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Lan Luo has a total of 30 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Intrinsic capacity in a nationally representative US population

    by Stephanie Denise M Sison, Lan Luo, Sunyoung Kim, Kevin Pritchard, Dae Hyun Kim, Sandra Shi

    Background: Intrinsic capacity (IC) is defined by the World Health Organization as the sum of an individual’s physical and mental abilities that support health and well-being. While well validated internationally, it has not been operationalized and validated in a nationally representative US cohort. Methods: Using data from 5834 community-dwelling older adults from Round 12 (2022) of the National Health and Aging Trends Study (NHATS) survey, we identified measures related to IC domains (cognition, locomotion, vitality, psychological, and sensory). We performed exploratory factor analysis (EFA), omega hierarchical reliability testing, and confirmatory factor analysis (CFA) using bi-factor and second-order structures. We also constructed the Integrated Care for Older People (ICOPE) screening tool. Descriptive statistics and regression models were used to examine the associations of IC and ICOPE scores with demographic and health characteristics. Age- and sex-specific reference centile curves were developed using empirical percentiles. Results: EFA consistently supported a six-factor structure (vision and hearing were separate). Omega hierarchical analysis supported a strong general factor (ω = 0.92; ωh = 0.72), and the bi-factor model demonstrated excellent fit. IC and ICOPE scores showed strong correlation (r = 0.71). Higher IC scores were associated with younger age, higher level of education, and income. Prevalence of comorbidities, frailty, ADL and IADL disability increased from the highest to the lowest IC and ICOPE quartiles. Reference centile curves demonstrated progressive age-related decline, with females showing lower IC scores than males. Conclusion: We demonstrate IC operationalization in a nationally representative US cohort and provide the first US-specific reference centile curves. The findings provide a foundation for future IC research in the US.

  • Performance of Chronic Conditions Data Warehouse algorithms in Medicare Advantage and fee-for-service populations

    by Chan Mi Park, Xiecheng Chen, Sandra Shi, Lan Luo, Eunji G Kim, Emilie D Duchesneau, Lily Zhong, Dae Hyun Kim, Ellen P McCarthy

    Background: The Centers for Medicare & Medicaid Services (CMS) Chronic Conditions Data Warehouse (CCW) algorithms are widely used to identify chronic conditions in Medicare fee-for-service (FFS) claims. However, their validity in Medicare Advantage (MA) encounters remains uncertain. Methods: We conducted a cross-sectional study using the National Health and Aging Trends Study (NHATS) linked to Medicare data. We analyzed NHATS rounds 9 and 11, aged ≥ 65 years, with continuous 24-month enrollment in MA or FFS. Chronic Conditions Data Warehouse algorithms were applied to identify 6 conditions: arthritis, chronic lung disease, diabetes, heart disease, hypertension, and osteoporosis. Self-reported, physician-diagnosed conditions from NHATS interviews served as the reference standard. Survey-weighted sensitivity, specificity, positive predictive value (PPV), negative predictive value, and balanced accuracy were compared between MA and FFS. Results: We included 1708 MA and 2239 FFS beneficiaries in round 9 and 1472 MA and 1566 FFS beneficiaries in round 11. Across both rounds, balanced accuracy was similar between MA and FFS for all 6 conditions. Diabetes and hypertension demonstrated high sensitivity (0.88-0.94), while arthritis and osteoporosis showed lower sensitivity but high specificity. Although statistically significant MA–FFS differences were observed for select metrics (diabetes PPV in round 9; chronic lung disease specificity in round 11), absolute differences were minor and not clinically meaningful. Conclusions: Chronic Conditions Data Warehouse algorithms perform comparably in MA and FFS Medicare data. These findings support the validity of using CCW algorithms in MA encounter data for epidemiologic research for risk adjustment, facilitating more generalizable studies of the growing MA population.

  • Performance of claims-based Alzheimer’s disease and related dementias algorithms in Medicare Advantage and fee-for-service populations

    by You Wu, Chan Mi Park, Xiecheng Chen, Lan Luo, Eunji G Kim, Sandra M Shi, Dae Hyun Kim, Ellen P McCarthy

    Background: Evidence comparing the performance of claims-based dementia algorithms across Medicare Advantage (MA) and fee-for-service (FFS) populations remains limited. We evaluated the validity of 2 claims-based algorithms across MA and FFS populations. Methods: We used nationally representative, linked National Health and Aging Trends Study (NHATS)-Medicare data to identify community-dwelling participants with continuous enrollment in MA or FFS during Round 7 (2017; n = 5018), Round 9 (2019; n = 4018), and Round 11 (2021-2022; n = 3086). Using NHATS probable dementia as reference standard, we validated the Bynum-Standard and Chronic Condition Warehouse (CCW) dementia algorithms across 3-time points. Results: Overall, both algorithms demonstrated comparable performance in MA and FFS populations, with high specificity (>97.9%) and negative predictive value (>92.9%). The Bynum algorithm showed similar sensitivity between MA and FFS in Rounds 7 (30.2% vs 29.1%) and 11 (25.6% vs 23.9%), but lower sensitivity in MA for Round 9 (25.0% vs 36.4%). Positive predictive value (PPV) was similar between MA (62.7%-70.6%) and FFS (62.1%-70.4%) across all rounds. The CCW algorithm demonstrated greater MA-FFS differences, with lower sensitivity in MA for Rounds 9 (30.8% vs 46.9%) and 11 (35.0% vs 49.3%) and similar sensitivity for Round 7 (40.4% vs 42.2%). PPV varied across rounds in both MA and FFS ([Round 7: 76.3% vs 71.2%] to [Round 11: 68.5% vs 76.4%]). Both algorithms identified participants with more severe dementia. Conclusion: Claims-based dementia algorithms demonstrate broadly consistent performance in MA and FFS populations, with high specificity and moderate PPV but comparable or lower sensitivity in MA, informing case ascertainment across Medicare plan type.

  • Performance of claims-based frailty indices in Medicare Advantage and fee-for-service populations

    by Lan Luo, Kevin Pritchard, Chan Mi Park, Xiecheng Chen, Eunji G Kim, Ellen P McCarthy, Dae Hyun Kim, Sandra M Shi

    Background: Despite growing enrollment in Medicare Advantage (MA) plans, the performance of claims-based frailty algorithms, which were developed for Medicare fee-for-service (FFS), remains unclear in MA. Methods: We calculated the Kim claims-based frailty index (CFI), Veterans Affairs-Frailty Index (VA-FI), and Hospital Frailty Risk Score (HFRS) using the National Health and Aging Trends Study linked to Medicare FFS and MA data. We evaluated their performance against the comprehensive geriatric assessment-based frailty index (CGA-FI) and the frailty phenotype (FP) in Round 11 and Round 9. Results: Among 1527 FFS and 1457 MA beneficiaries in Round 11, frailty prevalence was 34.2% and 35.5%, respectively, by CGA-FI ≥0.25, and 13.3% and 16.2%, respectively, by FP. Across FFS and MA populations, the CFI, which prioritizes specificity and positive predictive value, classified 10%-12% of the overall population and 30%-36% of those with prior hospitalization as frail, whereas the VA-FI and HFRS, which emphasize sensitivity, classified 53%-71% of the overall population and 56%-87% of those with prior hospitalization as frail. Although claims-based frailty measures performed similarly across FFS and MA populations, their discriminatory ability was higher for CGA-FI ≥0.25 (C-statistics: CFI, 0.79-0.82; VA-FI, 0.77-0.82; HFRS, 0.63-0.77) than for FP (CFI, 0.71-0.77; VA-FI, 0.67-0.75; HFRS, 0.60-0.74). Performance was consistent by race and survey rounds, with modestly better discrimination in women than men. Conclusions: Claims-based frailty measures are scalable tools for contemporary FFS and MA populations. However, applying published thresholds led to differences in frailty prevalence and model performance. Thresholds should therefore be tailored to the intended.

  • Open Access

    Interpretable Predictive Analytics for Online Learning

    by Chaewon Lee, Lan Luo, Shelbi L. Kuhlmann, Robert D. Plumley, Abigail T. Panter, Matthew L. Bernacki, Jeffrey A. Greene, Kathleen M. Gates

    The increasing use of learning management systems (LMSs) generates vast amounts of clickstream data, opening new avenues for predicting learner performance. Traditionally, LMS predictive analytics have relied on either supervised machine learning or Markov models to classify learners based on predicted learning outcomes. Machine learning excels at pattern recognition but often overlooks temporal learning dynamics and obscures the reasoning behind predictions due to the black-box nature of many algorithms. Alternatively, Markov models provide an effective solution by capturing temporal learning dynamics for prediction, uncovering distinctive learning patterns between high and low performers. Despite these advantages, Markov model classification struggles with the heterogeneity of learning sequences, limiting its broad applicability. To address these limitations and bridge the gap between the two dominant approaches, we propose a hybrid framework: sequence-based Markov machine learning classification (seqMAC). Leveraging early-stage clickstream data, seqMAC provides an interpretable sequence classification method that captures critical behavioural transitions and identifies distinct learning patterns across performance groups. Tested on six LMS samples, seqMAC effectively identified at-risk students despite sequence heterogeneity, uncovering key predictive learning dynamics that differentiate performance groups. It also demonstrated promising generalizability, accurately identifying future at-risk students based on historical clickstream data.


  • Selecting scaling indicators in structural equation models (sems).

    by Kenneth A. Bollen, Adam G. Lilly, Lan Luo

    It is common practice for psychologists to specify models with latent variables to represent concepts that are difficult to directly measure. Each latent variable needs a scale, and the most popular method of scaling as well as the default in most structural equation modeling (SEM) software uses a scaling or reference indicator. Much of the time, the choice of which indicator to use for this purpose receives little attention, and many analysts use the first indicator without considering whether there are better choices. When all indicators of the latent variable have essentially the same properties, then the choice matters less. But when this is not true, we could benefit from scaling indicator guidelines. Our article first demonstrates why latent variables need a scale. We then propose a set of criteria and accompanying diagnostic tools that can assist researchers in making informed decisions about scaling indicators. The criteria for a good scaling indicator include high face validity, high correlation with the latent variable, factor complexity of one, no correlated errors, no direct effects with other indicators, a minimal number of significant overidentification equation tests and modification indices, and invariance across groups and time. We demonstrate these criteria and diagnostics using two empirical examples and provide guidance on navigating conflicting results among criteria. (PsycInfo Database Record (c) 2024 APA, all rights reserved).


  • Estimating both directed and undirected contemporaneous relations in time series data using hybrid-group iterative multiple model estimation.

    by Lan Luo, Zachary F. Fisher, Cara Arizmendi, Peter C. M. Molenaar, Adriene Beltz, Kathleen M. Gates

    Researchers across varied fields increasingly are collecting and analyzing intensive longitudinal data (ILD) to examine processes across time at the individual level. Two types of relations are typically examined: lagged and contemporaneous. Lagged relations capture how variables at a prior time point can be used to explain variance in variables at a later time point. These are always modeled using auto- and cross-regressions by means of vector autoregression (VAR). By contrast, there are two types of relations commonly used to model the contemporaneous relations, which model how variables relate instantaneously. Until now, researchers must opt to either model contemporaneous relations as undirected relations among residuals (e.g., partial or full correlations) or as directed relations among the variables (e.g., paths or regressions). The choice for how to model contemporaneous relations has implications for inferences as well as the potential to introduce bias in the VAR lagged relations if the wrong type of relation is used. This article introduces a novel data-driven method, hybrid-group iterative multiple model estimation (GIMME), that provides a solution to the problem of having to choose one or the other type of contemporaneous relation to model. The modeling framework utilized in hybrid-GIMME allows for both types of contemporaneous relations in addition to the standard VAR relations. Both simulated and empirical data were used to test the performance of hybrid-GIMME. Results suggest this is a robust method for recovering contemporaneous relations in an exploratory manner, particularly with an ample number of time points per person. (PsycInfo Database Record (c) 2023 APA, all rights reserved).


  • An introduction to model implied instrumental variables using two stage least squares (MIIV-2SLS) in structural equation models (SEMs).

    by Kenneth A. Bollen, Zachary F. Fisher, Michael L. Giordano, Adam G. Lilly, Lan Luo, Ai Ye

    Structural equation models (SEMs) are widely used to handle multiequation systems that involve latent variables, multiple indicators, and measurement error. Maximum likelihood (ML) and diagonally weighted least squares (DWLS) dominate the estimation of SEMs with continuous or categorical endogenous variables, respectively. When a model is correctly specified, ML and DWLS function well. But, in the face of incorrect structures or nonconvergence, their performance can seriously deteriorate. Model implied instrumental variable, two stage least squares (MIIV-2SLS) estimates and tests individual equations, is more robust to misspecifications, and is noniterative, thus avoiding nonconvergence. This article is an overview and tutorial on MIIV-2SLS. It reviews the six major steps in using MIIV-2SLS: (a) model specification; (b) model identification; (c) latent to observed (L2O) variable transformation; (d) finding MIIVs; (e) using 2SLS; and (f) tests of overidentified equations. Each step is illustrated using a running empirical example from Reisenzein's (1986) randomized experiment on helping behavior. We also explain and illustrate the analytic conditions under which an equation estimated with MIIV-2SLS is robust to structural misspecifications. We include additional sections on MIIV approaches using a covariance matrix and mean vector as data input, conducting multilevel SEM, analyzing categorical endogenous variables, causal inference, and extensions and applications. Online supplemental material illustrates input code for all examples and simulations using the R package MIIVsem. (PsycInfo Database Record (c) 2022 APA, all rights reserved).


  • Exploratory Factor Analysis (EFA) Programs in R

    by Lan Luo, Cara Arizmendi, Kathleen M. Gates

    We provide a brief overview of two R packages that can conduct exploratory factor analysis (EFA): psych and EFAutilities. After introducing EFA and the exemplar data used in this paper we discuss best practices for EFA. Next, we describe the approaches used in the two packages for EFA. During this explanation, we provide sample code and discuss the usage and results of two empirical datasets. Finally, we highlight the similarities and distinctions of each package on modeling EFA.