Pith. sign in

REVIEW 4 major objections 5 minor 54 references

A Bayesian Spatio-Temporal Top-Down Framework for Estimating Opioid Use Disorder Risk Under Data Sparsity

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper claims that county-level opioid use disorder risk can be inferred from state-level surveillance data alone through a two-stage Bayesian top-down model that disaggregates state totals to counties with full uncertainty…

desk verdict Useful two-stage downscaling framework for sparse-data disease mapping, but the uncertainty quantification claim and internal simulation need more support before I would trust the county-level maps. read the letter →

arxiv 2506.02303 v1 pith:U7CFKBCR submitted 2025-06-02 stat.AP

Emily N Peterson , Alex Edwards , Martha Wetzel , Lance A Waller , Hannah Cooper , Courtney Yarbrough This is my paper
classification stat.AP MSC 62F1562M3062P10
keywords Bayesianhierarchicalmodelsmallareaestimationopioidusedisorderspatio-temporaltop-downdownscalingdatasparsityuncertaintyquantificationpublichealthsurveillance
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that county-level opioid use disorder (OUD) risk can be reliably inferred from observed state-level case counts even when no county-level counts are directly observed. The method, called B-Step, first estimates state-level risk and then disaggregates the estimated state totals down to counties using covariates, population weighting, and spatio-temporal random effects. Simulation experiments with 100 synthetic datasets show low bias and near-nominal coverage of 95% credible intervals at both spatial scales. If correct, the approach supplies policymakers with uncertainty-aware county-level risk estimates for all 3,143 U.S. counties from 2010 to 2025.

What carries the argument

The load-bearing mechanism is the two-stage top-down disaggregation with a population-weighted softmax allocation. In Stage I a Poisson likelihood links observed state counts to latent risk through an adjustment factor $\Delta_{s,t} \sim \mathrm{Uniform}(r_{s,t},1)$ that rescales post-2020 counts; in Stage II the estimated state total is split across counties by Poisson means $\mu_{c,t} = \rho_{c,t} \cdot \tilde{y}_{s,t}$ with $\rho_{c,t}$ a softmax over a log-linear predictor that includes a population offset, county covariates, and a BYM spatial plus RW1 temporal random effect. The multinomial-Poisson equivalence makes the county model conditionally equivalent to a multinomial allocation, so county counts reproduce the state total while remaining computationally tractable.

What would settle it

Generate synthetic county counts under a different mechanism, such as extra-Poisson overdispersion or spatial dependence with a different range than the fitted ICAR prior, and check whether the 95% credible intervals still cover the true county risks in about 95% of cases. Alternatively, take a state that does report county-level OUD or a strong proxy such as treatment admissions, fit B-Step using only state totals, and test whether the estimated county risks track the observed county-level values.

Watch

Extended reading notes

Core claim

The paper argues that county-level OUD risk can be estimated entirely from state-level surveillance counts by splitting the inference into two modular stages. Stage I fits a Bayesian Poisson time-series to observed state counts, with a random-walk temporal effect and a uniform-prior adjustment factor that corrects for the post-2020 widening of the case definition. Stage II takes the posterior state totals and disaggregates them to counties through a softmax-normalized Poisson model whose expected county counts sum exactly to the state total, adding a population-weighted offset, county covariates, and a BYM spatial plus first-order random-walk temporal effect. The output is a posterior distribution of risk for each county-year, so uncertainty is explicitly quantified rather than ignored. The claimed result is that this pipeline recovers true simulated risks with low bias and approximately nominal 95% credible-interval coverage at both state and county levels.

Load-bearing premise

The load-bearing premise is that the process that generated the real OUD counts matches the Poisson likelihood with ICAR/RW1 random effects used both to fit the model and to generate the simulation data, since no independent county-level ground truth is used for validation.

Editorial extensions

If this is right

  • County-level OUD risk maps for all 3,143 U.S. counties from 2010 to 2025 become available even though county-level case counts are never observed directly.
  • Each county-year estimate carries a 95% credible interval, so policymakers can distinguish reliably high-burden areas from areas where the data are simply too sparse to know.
  • County estimates sum exactly to the modelled state totals, preserving internal consistency when state and local numbers are used together.
  • The population-weighted offset prevents tiny counties from receiving extreme allocated risks solely because of their small size.
  • The case-definition adjustment factor lets pre- and post-2020 rates be compared on a common scale, so apparent jumps in the epidemic are not artefacts of survey changes.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Because the validation is self-referential (the simulator uses the fitted model's own priors), the accuracy claim should be read as conditional on the model family being correct; a misspecified-data test would be the natural next check.
  • If the framework generalizes, the same two-stage design could be applied to other health outcomes that are only observed at state or national level, such as hepatitis C or stimulant use disorder, with minimal changes.
  • The paper's interpolation of covariates for 2024-2025 and its reliance on the random walk for years beyond the data mean the later-year estimates are extrapolations; a back-test that fits up to 2022 and compares 2023 predictions against observed state counts would quantify this.
  • The wide credible intervals on some state-level covariates suggest those variables carry little information; a sensitivity analysis removing them would reveal how much of the county variation comes from spatial smoothing rather than covariate effects.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper develops B-Step, a two-stage Bayesian spatio-temporal model that downscales state-level NSDUH opioid use disorder (OUD) counts to county-level risk estimates for 3,143 US counties over 2010-2025. Stage I is a state-level Poisson model with a first-order random walk over time and a case-definition adjustment factor; Stage II allocates the resulting state totals to counties using a softmax-normalized Poisson model with BYM spatial effects, RW1 temporal effects, and a population-weighted offset. The authors apply the framework to real surveillance data and evaluate it through 100 simulated datasets, reporting state- and county-level accuracy and coverage metrics.

Significance. If the central claim were fully supported, the paper would make a useful applied contribution: county-level OUD estimates with uncertainty are policy-relevant, and the proposed adjustment for the 2020 NSDUH definitional change addresses a real and under-documented harmonization problem. The paper also has practical strengths: it builds on publicly available NSDUH and Census data, the two-stage modular structure is transparent, and the use of a population-weighted offset to stabilize allocation in sparsely populated counties is a sensible practical fix. However, the validation is internal: the simulation generates data from the same Poisson/ICAR/RW1 structure used by the fitted model, and the reported credible intervals exclude Stage I uncertainty because posterior medians are plugged into Stage II. These limitations directly affect the paper's claims of 'strong accuracy and calibration' and 'fully Bayesian' uncertainty propagation, so the contribution is currently overstated.

major comments (4)
  1. [Section 3.5, Eq. (10)] The simulation data-generating process in Eq. (10) uses the same Poisson likelihood, the same ICAR/RW1 random-effect structure, and the same covariates as the fitted Stage II model, so the simulation is a self-consistency check rather than an independent validation of the model's assumptions. Consequently, Table 2's favorable state-level metrics and county-level coverage of 0.89 cannot support the Discussion's claim that simulations demonstrated 'strong accuracy and calibration' for real-world OUD estimation. The authors should either reframe the simulation explicitly as an internal-consistency check or supplement it with an external validation, for example by withholding some state-years, comparing with direct county-level estimates where available, or examining sensitivity to alternative data-generating mechanisms.
  2. [Sections 3.3 and 3.4, Eqs. (7)-(9)] Stage II uses posterior medians of the state-level risk, count, and intercept as plug-in inputs: Section 3.3 states that ~ys,t and ~omega_s are posterior medians, and Section 3.4 treats them as fixed offsets in the multinomial/Poisson allocation and the prior for gamma_s. The county-level credible intervals reported in Figures 5 and 6 therefore exclude Stage I uncertainty about the state totals and state intercepts, which is in direct tension with the paper's stated 'fully Bayesian' uncertainty propagation. Table 2's county-level coverage of 0.89 is computed under the same plug-in pipeline, so it cannot establish calibration of the reported county-level intervals. The authors should propagate full posterior draws of ~ys,t and ~omega_s through Stage II, or explicitly describe the reported intervals as conditional on Stage I point estimates.
  3. [Section 3.2, Eqs. (4)-(5)] The lower bound r_s,t of the Uniform prior for the adjustment factor Delta_s,t is computed directly from the observed counts y_s,2016 and y_s,t that also enter the Stage I likelihood in Eq. (1). This makes the prior data-dependent and uses the post-2020 observations twice, which can artificially narrow Stage I credible intervals and overstate precision for the adjusted post-2020 rates. The simulation study does not address this issue because the data-generating process in Eq. (10) omits the adjustment-factor mechanism entirely. The authors should either specify a prior for Delta_s,t that is conditionally independent of the likelihood (for example, derived from external methodological-validation data) or provide a sensitivity analysis that quantifies the effect of this double use of the data.
  4. [Table 2 and Section 5.3] The county-level simulation results are not as strong as the text suggests: median relative error is 0.84 and 95% interval coverage is 0.89, which is below nominal and far from 'near-nominal coverage' as stated in Section 5.3. For a county with true risk near the national average, a median relative error of 84% means the posterior median can be off by almost a factor of two. The Discussion's claim of 'strong accuracy and calibration' should be revised, and the authors should report performance stratified by county population size or true risk level, since the current aggregate metrics conceal where the model performs poorly.
minor comments (5)
  1. [Throughout] The manuscript contains several typographical errors: 'mortalty' in Section 3.4, 'calues' in the Figure 6 caption, 'subsetted' in the Figure 5 caption, and 'Y ear' as a partially cut axis label in multiple figures. These should be corrected in a final polish.
  2. [Eq. (8) and Table B.3] The symbol mu_c,t is used both for the latent county-level count and for the Poisson mean, which is confusing; for example, Eq. (8) states 'mu_c,t ~ Poisson(mu_c,t)'. Use distinct notation, such as theta_c,t for the mean, to avoid ambiguity.
  3. [Section 3.5] The sentence introducing the performance metrics is incomplete: 'For each parameter at both spatial scales (state and county), we computed the following performance metrics across all simulation replicates:' is followed by a blank line and then a bulleted list. The sentence should be finished and the list integrated into the text.
  4. [Table 1] The posterior interval for beta_1, the PR misuse coefficient, is very wide (95% CrI [-6.07, 1.45]); the text should describe this as an imprecisely estimated effect consistent with a wide range of values, rather than emphasizing the negative posterior mean without noting the near-zero information in the data.
  5. [Reproducibility] The manuscript does not mention data or code availability. Since the data are public and the computation uses NIMBLE and R-INLA, providing a repository with the fitted models and processing scripts would strengthen the reproducibility of the applied results.

Circularity Check

3 steps flagged · score 6.0 of 10

Post-2020 adjustment is built from the same observed counts, simulation 'ground truth' is drawn from the fitted model's own priors, and Stage II uncertainty uses plug-in medians; the central validation and uncertainty claims are partially constructed.

  1. self definitional [Section 3.2, Eqs. (1), (4)-(5); Section 3.3, Eq. (6)]
    "we computed r_{s,t} as the ratio of observed OUD case counts in 2016 to the corresponding count in year t for each state. In summary, r=y_{s,2016}/y_{s,t}, where y_{s,t} is the reported count in state s and year t. ... For years following the definition change (t>t_0), we model the adjustment factors as: ∆_{s,t} ∼ Uniform(r_{s,t},1), for t>2016."

    The same observed counts y_{s,t} enter the Stage I Poisson likelihood (Eq. 1) and define the lower bound of the Uniform prior on the adjustment factor ∆_{s,t}. Consequently the posterior of ∆_{s,t} is forced toward r_{s,t}=y_{s,2016}/y_{s,t}, and the posterior of the adjusted risk π_{s,t} is pinned to approximately y_{s,2016}/n_{s,t}. The post-2020 'predicted' trend is therefore the 2016 count rescaled by the observed ratios, not an independent inference; the data-dependent prior also narrows credible intervals because the data are used twice.

  2. self definitional [Section 3.5, Eq. (10); Section 3.4.0.3, Eq. (9)]
    "For each simulation, we generated spatio-temporal random effects φ_{c,t} from their respective prior distributions (e.g., ICAR for spatial effects, random walk for temporal trends). We then computed the true latent log-risk for each county-year and derived the true incidence risk via the inverse log-link. County-level counts were generated using the Poisson likelihood where n_{c,t} is the known population denominator. Lastly, state-level counts were computed by summing across counties."

    The simulation 'truth' is generated from the same BYM/ICAR spatial effect, RW1 temporal effect, and Poisson likelihood that the fitted B-Step Stage II model assumes (Eq. 9 and Section 3.4.0.3). Fitting a model to data drawn from its own prior and then reporting low error and near-nominal coverage is a self-consistency check: posterior recovery of model-generated truth is expected by construction. It does not validate the ICAR/RW1/Poisson assumptions for real county OUD counts, so the Discussion's claim of 'strong accuracy and calibration' is not supported by external evidence.

1 more flagged steps
  1. other [Section 3.3 Post-processing; Section 3.4, Eqs. (7)-(9); Table 2]
    "we obtain posterior median estimates of the latent OUD risk π_{s,t} and the state-level random intercept ω_s, which are used as inputs to the county-level model. Similarly, posterior estimates of risk ˜π_{s,t} are scaled by the population size n_{s,t} to compute draws of state-level OUD counts as ˜y_{s,t} = ˜π_{s,t} · n_{s,t}, which serve as the total counts to be disaggregated across counties. ... {µ_{c,t} : c∈C_s} ∼ Multinomial({ρ_{c,t} : c∈C_s}, ˜y_{s,t})."

    Stage II conditions on point estimates ỹ_{s,t} and ω̃_s rather than on posterior draws, yet the paper claims a 'fully Bayesian' framework that 'propagate[s] uncertainty across modeling stages' and says the formulation 'yields fully probabilistic' estimates. The county-level credible intervals in Figures 5-6 and the Table 2 coverage of 0.89 are therefore conditional intervals that exclude Stage I state-level uncertainty. The uncertainty 'prediction' is constructed from fitted medians used as known totals, so it is narrower than the claimed full posterior and cannot validate the paper's uncertainty-quantification claim.

full rationale

The paper's county-level downscaling itself is a genuine modeling exercise: the softmax/Poisson reparameterization of the multinomial and the population-weighted offset are algebraically correct and do not reduce to the inputs. The central problems are in the validation and uncertainty layers. The adjustment factor prior uses the same observed counts as the likelihood, so the post-2020 state trajectory is effectively y_2016 rescaled; the simulation generates 'true' risks from the same ICAR/RW1/Poisson priors used by the fitted model, making the reported accuracy a consistency check; and Stage II plugs in posterior medians while claiming full uncertainty propagation. None of these issues arises from self-citation: the cited prior work by the authors (e.g., Peterson et al. 2023, 2024; Hepler/Waller papers) supports related methods but is not load-bearing for the forced result. Because the validation and uncertainty claims reduce partially to the model's own construction, a score of 6 is warranted; the underlying method remains potentially useful and the issues are fixable with proper external validation and full posterior propagation.

Assumptions & free parameters 3 free parameters · 6 assumptions · 1 invented entities

The central claim rests on standard spatial statistics tools plus a data-dependent adjustment factor. The most exposed elements are the self-consistency simulation and the double-use of the observed counts in constructing the adjustment prior.

free parameters (3)
  • r_s,t adjustment ratio = y_s,2016 / y_s,t, linearly interpolated for missing years
    Data-derived lower bound for the Uniform prior on Delta_s,t; computed from the same NSDUH counts used in the likelihood, so it is a fitted input rather than an independent prior.
  • Reference year 2016 = 2016
    Chosen as the anchor for the definitional adjustment; the estimated post-2020 risk levels depend on this choice.
  • Reference year 2015 for random walk = 2015
    The temporal random walk for phi_s,t is anchored at 2015; the extrapolated trend beyond the observed period depends on this choice.
assumptions (6)
  • standard math The multinomial model is equivalent to conditionally independent Poisson models via the multinomial-Poisson transformation.
    Invoked in Stage II, Eq. (8), citing Baker (1994).
  • domain assumption State-level NSDUH reported counts are unbiased for latent state OUD risk after applying the Delta adjustment.
    Stage I likelihood, Eq. (1); under-reporting and survey error are not modeled.
  • domain assumption Latent county-level counts sum exactly to the posterior median state-level total tilde y_s,t.
    Stage II multinomial constraint, Eq. (7).
  • domain assumption The ICAR/BYM spatial prior and RW1 temporal prior adequately capture unmeasured county risk variation.
    Stage II latent process, Eq. (9) and Section 3.4.0.3.
  • ad hoc to paper The adjustment factor Delta_s,t lies in [r_s,t, 1] with r_s,t derived from observed counts, so the pre-2020 and post-2020 counts are comparable after scaling.
    Eq. (4)-(5); this double-uses the observed post-2020 counts in both the prior and the likelihood.
  • domain assumption Covariates are measured without error and the 2011 cumulative opioid prescription rate is a valid predictor for all subsequent years.
    County covariates are entered as fixed known values; Appendix A shows the 2011 rate is held constant over time.
invented entities (1)
  • Diagnostic adjustment factor Delta_s,t
    purpose: Scales observed post-2020 NSDUH OUD counts to be comparable to pre-2020 DSM-IV-based counts.
    It is a latent model construct with a Uniform(r,1) prior whose lower bound is computed from the same observed counts it corrects, so it has no independent falsifiable handle outside the paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Bayesian Spatio-Temporal Top-Down Framework for Estimating Opioid Use Disorder Risk Under Data Sparsity." pith.science (2026). https://pith.science/paper/U7CFKBCR

@misc{pith2026250602303,
  author       = {Pith},
  title        = {Pith review of: A Bayesian Spatio-Temporal Top-Down Framework for Estimating Opioid Use Disorder Risk Under Data Sparsity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U7CFKBCR}},
  note         = {Machine review of arXiv:2506.02303}
}
read the original abstract

County-level estimates of opioid use disorder (OUD) are essential for understanding the influence of local economic and social conditions. They provide policymakers with the granular information needed to identify, target, and implement effective interventions and allocate resources appropriately. Traditional disease mapping methods typically rely on Poisson regression, modeling observed counts while adjusting for local covariates that are treated as fixed and known. However, these methods may fail to capture the complexities and uncertainties in areas with sparse or absent data. To address this challenge, we developed a Bayesian hierarchical spatio-temporal top-down approach designed to estimate county-level OUD rates when direct small-area (county) data is unavailable. This method allows us to infer small-area OUD rates and quantify associated uncertainties, even in data-sparse environments using observed state-level OUD rates and a combination of state and county level informative covariates. We applied our approach to estimate OUD rates for 3,143 counties in the United States between 2010 and 2025. Model performance was assessed through simulation studies.

Figures

Figures reproduced from arXiv: 2506.02303 by the authors.

Figure 1
Figure 1. NSDUH reported OUD risks (counts/population at risk) for selected states. The black [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Directed acyclic graph (DAG) illustrating the two-stage hierarchical Bayesian top [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. State-level trends in posterior median risk estimates and 95% CrIs for selected states. [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: State-level trends in adjustment factors. Red dots denote the reported [PITH_FULL_IMAGE:figures/full_fig_p014_4.png]
Figure 5
Figure 5. Figure 5: Mapped county-level posterior median risk estimates [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: Temporal trends in county-level posterior median OUD risk estimates with 95% cred [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

54 extracted references · 44 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type url volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.all := #1 'mid.sentence := ...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    title Inference from iterative simulation using multiple sequences (with discussion)

    author A Gelman , author DB Rubin , year 1992 . title Inference from iterative simulation using multiple sequences (with discussion) . journal Statistical Science volume 7 , pages 457--472

    work page 1992

  4. [4]

    , author Lucas, T.C.D

    author Arambepola, R. , author Lucas, T.C.D. , author Nandi, A.K. , author Gething, P.W. , author Cameron, E. , year 2022 . title A simulation study of disaggregation regression for spatial disease mapping . journal Statistics in Medicine volume 41 , pages 1--16 . https://onlinelibrary.wiley.com/doi/10.1002/sim.9220, :10.1002/sim.9220

    work page doi:10.1002/sim.9220 2022

  5. [5]

    , year 1994

    author Baker, S.G. , year 1994 . title The multinomial-poisson transformation . journal The Statistician volume 43 , pages 495--504 . https://www.jstor.org/stable/2348134, :10.2307/2348134

    arXiv 1994

  6. [6]

    , year 2008

    author Baker, S.G. , year 2008 . title Generalized self-consistency: Multinomial logit model and poisson log-linear model . journal Computational Statistics & Data Analysis volume 52 , pages 4483--4494 . https://www.sciencedirect.com/science/article/pii/S0378375807003734, :10.1016/j.csda.2007.02.009

    doi:10.1016/j.csda.2007.02.009 2008

  7. [7]

    On the behavior of the Generalized Alignment Index (GALI) method for regular motion in multidimensional Hamiltonian systems

    author Bennett, J. , author Wakefield, J. , year 2001 . title Errors-in- Variables in Joint Population Pharmacokinetic / Pharmacodynamic Modeling . journal Biometrics volume 57 , pages 803--812 . https://doi.org/10.1111/j.0006-341X.2001.00803.x, :10.1111/j.0006-341X.2001.00803.x. note publisher: John Wiley & Sons, Ltd

    work page Pith review arXiv 2001

  8. [8]

    , author Holan, S.H

    author Bradley, J.R. , author Holan, S.H. , author Wikle, C.K. , year 2015 . title Multiscale spatial modeling of rates using the american community survey . journal Journal of the Royal Statistical Society: Series A volume 178 , pages 173--190

    work page 2015

Show all 54 references
  1. [9]

    title 2020 Methodological Summary and Definitions

    author Center for Behavioral Health Statistics and Quality , year 2021 . title 2020 Methodological Summary and Definitions . type Technical Report . Substance Abuse and Mental Health Services Administration. address Rockville, MD . https://www.samhsa.gov/data/report/2020-metho...

    2021

  2. [10]

    title Drug overdose deaths: Drug overdose deaths remained high in 2021

    author Centers for Disease Control and Prevention (CDC) , year 2023 . title Drug overdose deaths: Drug overdose deaths remained high in 2021 . howpublished https://www.cdc.gov/drugoverdose/deaths/index.html

    2023

  3. [11]

    , author Beletsky, L

    author Dasgupta, N. , author Beletsky, L. , author Ciccarone, D. , year 2018 . title Opioid epidemic: Addressing social determinants of health . journal International Journal of Drug Policy volume 59 , pages 1--2

    2018

  4. [12]

    , author Simpson, D

    author Gabry, J. , author Simpson, D. , author Vehtari, A. , author Betancourt, M. , author Gelman, A. , year 2019 . title Visualization in Bayesian workflow . journal Journal of the Royal Statistical Society: Series A (Statistics in Society) volume 182 , pages 389--402 . http...

    2019 doi

  5. [13]

    , author Wakefield, J

    author Gao, Y. , author Wakefield, J. , year 2022 . title Spatial variance smoothing for small area estimation of demographic rates . journal arXiv preprint arXiv:2209.02602 https://arxiv.org/abs/2209.02602

    2022 arXiv

  6. [14]

    , author Rao, J.N.K

    author Ghosh, M. , author Rao, J.N.K. , year 1994 . title Small area estimation: An appraisal . journal Statistical Science volume 9 , pages 55--76

    1994

  7. [15]

    , year 2020

    author Griffin, S. , year 2020 . title Spatial downscaling disease risk using random forests machine learning . type Technical Report . Engineer Research and Development Center (U.S.). https://hdl.handle.net/11681/35618, :10.21079/11681/35618

    2020 doi

  8. [16]

    , author Tibshirani, R

    author Hastie, T. , author Tibshirani, R. , year 2009 . title Model Robustness and Validation in Predictive Modeling . journal Statistical Science volume 24 , pages 570--586 . :10.1214/09-STS297

    2009 doi

  9. [17]

    title An integrated abundance model for estimating county-level prevalence of opioid misuse in Ohio

    author Hepler SA , author Kline DM , author Bonny A , author McKnight E , author Waller LA , year 2023 . title An integrated abundance model for estimating county-level prevalence of opioid misuse in Ohio . journal Journal of the Royal Statistical Society Series A: Statistics ...

    2023

  10. [18]

    , author Quick, H

    author Hogg, R. , author Quick, H. , author Wakefield, J. , year 2023 . title A two-stage bayesian small area estimation method for proportions with sparse data . journal arXiv preprint arXiv:2306.11302 https://arxiv.org/abs/2306.11302

    2023 arXiv

  11. [19]

    , author Murray, L.M

    author Jacob, P.E. , author Murray, L.M. , author Holmes, C.C. , author Robert, C.P. , year 2017 . title Better together? Statistical learning in models made of modules . journal arXiv preprint arXiv:1708.08719

    2017 arXiv

  12. [20]

    , author Ji, Y

    author Kline, D. , author Ji, Y. , author Hepler, S. , year 2021 . title A multivariate spatio-temporal model of the opioid epidemic in ohio: a factor model approach . journal Health Services and Outcomes Research Methodology volume 21 , pages 42--53

    2021

  13. [21]

    title Estimating the burden of the opioid epidemic for adults and adolescents in Ohio counties

    author Kline D , author Hepler SA , year 2021 . title Estimating the burden of the opioid epidemic for adults and adolescents in Ohio counties . journal Biometrics volume 77 , pages 765--775 . note Publisher: Wiley Online Library

    2021

  14. [22]

    title A Dynamic Spatial Factor Model to Describe the Opioid Syndemic in Ohio

    author Kline D , author Waller LA , author McKnight E , author Bonny A , author Miller WC , author Hepler SA , year 2023 . title A Dynamic Spatial Factor Model to Describe the Opioid Syndemic in Ohio . journal Epidemiology volume 34 , pages 487--494 . note Publisher: Wolters Kluwer

    2023

  15. [23]

    , author Padellini, T

    author Konstantinoudis, G. , author Padellini, T. , author Bennett, J. , author Davies, B. , author Ezzati, M. , author Blangiardo, M. , year 2021 . title Long-term exposure to air-pollution and COVID -19 mortality in England : A hierarchical spatial analysis . journal Environ...

    2021

  16. [24]

    , author Tatem, A.J

    author Linard, C. , author Tatem, A.J. , year 2012 . title Large-scale spatial population databases in infectious disease research . journal International Journal of Health Geographics volume 11 , pages 7 . http://ij-healthgeographics.biomedcentral.com/articles/10.1186/1476-07...

    2012 doi

  17. [25]

    , author Wakefield, J

    author Mercer, L.D. , author Wakefield, J. , author Chen, D. , author Lumley, T. , et al., year 2015 . title Small area estimation of child mortality in the absence of vital registration . journal Annals of Applied Statistics volume 9 , pages 1889--1905 . :10.1214/15-AOAS872

    2015 doi

  18. [26]

    , year 2018

    author Monnat, S.M. , year 2018 . title Factors associated with county-level differences in u.s. drug-related mortality rates . journal American Journal of Preventive Medicine volume 54 , pages 611--619

    2018

  19. [27]

    , author Lucas, T.C.D

    author Nandi, A.K. , author Lucas, T.C.D. , author Arambepola, R. , author Gething, P. , author Weiss, D.J. , year 2020 . title disaggregation: An R Package for Bayesian Spatial Disaggregation Modelling . http://arxiv.org/abs/2001.04847, :10.48550/arXiv.2001.04847. note arXiv:...

    2020 doi arXiv

  20. [28]

    title Only 1 in 5 u.s

    author National Institute on Drug Abuse , year 2023 a. title Only 1 in 5 u.s. adults with opioid use disorder received medications to treat it in 2021 . https://nida.nih.gov/news-events/news-releases/2023/08/only-1-in-5-us-adults-with-opioid-use-disorder-received-medications-t...

    2023

  21. [29]

    title Opioid use disorder: What is oud? https://nida.nih.gov/research-topics/opioids/opioid-use-disorder

    author National Institute on Drug Abuse , year 2023 b. title Opioid use disorder: What is oud? https://nida.nih.gov/research-topics/opioids/opioid-use-disorder

    2023

  22. [30]

    title Programming with models: Writing statistical algorithms for general model structures with NIMBLE

    author Perry de Valpine , author Daniel Turek , author Christopher Paciorek , author Cliff Anderson-Bergman , author Duncan Temple Lang , author Ras Bodik , . title Programming with models: Writing statistical algorithms for general model structures with NIMBLE . :10.1080/1061...

  23. [31]

    , author Guranich, G

    author Peterson, E.N. , author Guranich, G. , author Cresswell, J.A. , author Alkema, L. , year 2024 . title A bayesian approach to estimate maternal mortality globally using national civil registration vital statistics data accounting for reporting errors . journal Statistics...

    2024

  24. [32]

    , author Nethery, R.C

    author Peterson, E.N. , author Nethery, R.C. , author Chen, J.T. , author Tabb, L.P. , author Coull, B.A. , author Piel, F.B. , author Waller, L.A. , year 2023 . title A bayesian spatial berkson error approach to estimate small area opioid mortality rates accounting for popula...

    2023 arXiv

  25. [33]

    , year 2015

    author Plummer, M. , year 2015 . title Cuts in Bayesian graphical models . journal Statistics and Computing volume 25 , pages 37--43 . https://doi.org/10.1007/s11222-014-9503-z, :10.1007/s11222-014-9503-z

    2015 doi

  26. [34]

    , year 2017

    author Plummer, M. , year 2017 . title JAGS : A program for analysis of Bayesian graphical models using Gibbs sampling

    2017

  27. [35]

    Census Bureau , year 2019

    author Population Estimation Program, U.S. Census Bureau , year 2019 . title Methodology for the united states population estimates: Vintage 2019 . howpublished https://www2.census.gov/programs-surveys/popest/technical-documentation/methodology/2010-2019/natstcopr-methv2.pdf ....

    2019

  28. [36]

    , author Bender, A

    author Python, A. , author Bender, A. , author Blangiardo, M. , author Illian, J.B. , author Lin, Y. , author Liu, B. , author Lucas, T.C. , author Tan, S. , author Wen, Y. , author Svanidze, D. , author Yin, J. , year 2022 . title A Downscaling Approach to Compare COVID -19 C...

    2022

  29. [37]

    , author Molina, I

    author Rao, J.N.K. , author Molina, I. , year 2015 . title Small Area Estimation . publisher Wiley , address Hoboken, NJ

    2015

  30. [38]

    , author Unick, G.J

    author Rosenblum, D. , author Unick, G.J. , year 2022 . title The role of health systems in addressing the opioid overdose crisis . journal Annual Review of Public Health volume 43 , pages 311--329

    2022

  31. [39]

    , author Martino, S

    author Rue, H. , author Martino, S. , author Chopin, N. , year 2009 . title Approximate bayesian inference for latent gaussian models by using integrated nested laplace approximations . journal Journal of the Royal Statistical Society: Series B (Statistical Methodology) volume...

    2009

  32. [40]

    , et al., year 2024

    author Rue, H. , et al., year 2024 . title R-inla project . note https://www.r-inla.org

    2024

  33. [41]

    , author Held, L

    author Rue, H. , author Held, L. , year 2005 . title Gaussian Markov Random Fields . publisher Chapman & Hall/CRC

    2005

  34. [42]

    title 2021 national survey on drug use and health

    author Substance Abuse and Mental Health Services Administration , year 2023 . title 2021 national survey on drug use and health . https://www.samhsa.gov/data/report/2021-nsduh-detailed-tables

    2023

  35. [43]

    , author Goicoa, T

    author Ugarte, M.D. , author Goicoa, T. , author Militino, A.F. , year 2020 . title Bayesian benchmarking techniques in small area estimation . journal Statistical Science volume 35 , pages 261--284

    2020

  36. [44]

    , year 2016

    author United States Department of Health and Human Services , Substance Abuse and Mental Health Services Administration , C. , year 2016 . title National Survey on Drug Use and Health, 2014 . https://doi.org/10.3886/ICPSR36361.v1, :10.3886/ICPSR36361.v1. note public Use File

    2016 doi

  37. [45]

    , year 2023

    author United States Department of Health and Human Services , Substance Abuse and Mental Health Services Administration , C. , year 2023 . title National Survey on Drug Use and Health, 2022 . https://www.samhsa.gov/data/system/files/media-puf-file/NSDUH-2022-DS0001-info-codeb...

    2023

  38. [46]

    Census Bureau , year 2025

    author U.S. Census Bureau , year 2025 . title Population and Housing Unit Estimates . https://www.census.gov/popest. note population Estimates Program (PEP), Vintage 2024

    2025

  39. [47]

    Department of Health and Human Services , year 2020 a

    author U.S. Department of Health and Human Services , year 2020 a. title Advancing the collection and use of opioid-related data: Progress and opportunities . https://aspe.hhs.gov/sites/default/files/private/pdf/259016/NORC-2020-ASPE-Opi-Vignette.pdf. note accessed: 2025-01-21

    2020

  40. [48]

    Department of Health and Human Services , year 2020 b

    author U.S. Department of Health and Human Services , year 2020 b. title Hhs strategy to combat the opioid crisis . https://www.hhs.gov/opioids/about-the-epidemic/index.html

    2020

  41. [49]

    , author Gelman, A

    author Vehtari, A. , author Gelman, A. , author Simpson, D. , author Carpenter, B. , author Bürkner, P.C. , year 2021 . title Rank- Normalization , Folding , and Localization : An Improved \ widehat\ R \ \ for Assessing Convergence of MCMC (with Discussion ) . journal Bayesian...

    2021 doi

  42. [50]

    , year 2007

    author Wakefield, J. , year 2007 . title Disease mapping and spatial regression with count data . journal Biostatistics volume 8 , pages 158--183 . :10.1093/biostatistics/kx1008

    2007 doi

  43. [51]

    , year 2020

    author Walker, K. , year 2020 . title tidycensus: Load US Census Boundary and Attribute Data as `tidyverse'. R package version 0.9.9.2 . https://walker-data.com/tidycensus/articles/basic-usage.html

    2020

  44. [52]

    , author Gotway, C.A

    author Waller, L.A. , author Gotway, C.A. , year 2004 . title Applied Spatial Statistics for Public Health Data . publisher John Wiley & Sons, Inc. :10.1002/0471662682

    2004 doi

  45. [53]

    , author Best, N

    author Wulandari, E. , author Best, N. , author Richardson, S. , year 2023 . title Addressing overdispersion in small-area count data: a bayesian perspective . journal Biostatistics volume 24 , pages 383--401

    2023

  46. [54]

    , author Rao, J.N.K

    author You, H. , author Rao, J.N.K. , year 2020 . title Bayesian Hierarchical Spatial Models for Small Area Estimation . type Technical Report . U.S. Census Bureau. https://www.census.gov/library/working-papers/2020/adrm/RRS2020-07.html

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.