pith. sign in

arxiv: 2604.18431 · v2 · submitted 2026-04-20 · ⚛️ physics.soc-ph · q-bio.QM

Care Trajectories Are Linked to Mental Health and Mortality in Cancer Patients

Simon D. Lindner , Elisabeth L. Zeilinger , Amelie Fuchs , Simone Lubowitzki , Peter Klimek , Alexander Gaiger This is my paper

Pith reviewed 2026-05-10 03:16 UTC · model grok-4.3

classification ⚛️ physics.soc-ph q-bio.QM
keywords cancer care trajectoriesdynamic time warpingmortality predictionmental healthhierarchical clusteringhealthcare utilizationprognostic factorsprecision oncology
0
0 comments X p. Extension

The pith

Care trajectories in over 8,000 cancer patients form nine phenotypes that independently predict mortality.

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

The paper applies sequence analysis to long-term healthcare records of over 8,000 cancer patients to uncover hidden patterns in their care journeys. It shows that these patterns, derived from the timing and frequency of medical encounters, add substantial power to predicting who will die and also connect to how anxious patients feel at the start. By grouping similar trajectories using a warping technique that aligns sequences of different lengths, the authors reveal two main risky paths: one of prolonged complicated care and one of quick intense treatment. This work argues that looking at the full timeline of care, rather than single snapshots, reveals prognostic information missed by standard variables like age or tumor stage. If correct, it means doctors could use these patterns for better risk assessment in treating cancer.

Core claim

Using Dynamic Time Warping and hierarchical clustering on sequences of healthcare encounters spanning up to 37 years, we identified nine distinct trajectory phenotypes in over 8,000 cancer patients. These phenotypes significantly improved mortality prediction in generalized linear models beyond conventional covariates and showed independent predictive value. Two high-risk patterns emerged: long-term complex pathways with up to 196 events (OR up to 3.38) and shorter intense trajectories (median 78 events, OR 2.32). Notably, high-utilization clusters correlated with lower baseline anxiety scores, suggesting a complex interplay between care intensity, psychological burden, and survival outcomes

What carries the argument

Dynamic Time Warping (DTW) applied to sequences of healthcare encounters, followed by hierarchical clustering to derive nine trajectory phenotypes that capture temporal patterns in care utilization.

If this is right

  • Trajectory clusters add explanatory power to mortality prediction beyond conventional clinical, demographic, and socioeconomic covariates.
  • All eight non-reference clusters exhibit substantially higher mortality odds than the low-utilization reference group.
  • Long-term complex care pathways show mortality odds ratios up to 3.38.
  • Shorter intense trajectories show a mortality odds ratio of 2.32.
  • High-utilization complexity clusters are associated with significantly lower baseline anxiety scores.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Real-time tracking of a patient's position within these trajectory phenotypes could flag high-risk cases early for intervention.
  • The link between intense care use and lower anxiety may reflect adaptation or selection that deserves targeted study.
  • If the phenotypes replicate across other cancer types or health systems, they could feed directly into electronic health record risk tools.
  • Trajectory data might help design trials that test whether changing care intensity alters survival in matched groups.

Load-bearing premise

The nine clusters identified represent stable and clinically meaningful phenotypes of care rather than artifacts of the patient cohort, encounter coding, or the choice of distance metric and linkage method.

What would settle it

Re-running the DTW and hierarchical clustering on an independent cohort of cancer patients yields clusters that do not improve mortality prediction after adjustment for standard covariates.

Figures

Figures reproduced from arXiv: 2604.18431 by Alexander Gaiger, Amelie Fuchs, Elisabeth L. Zeilinger, Peter Klimek, Simon D. Lindner, Simone Lubowitzki.

Figure 1
Figure 1. Figure 1: Conceptual diagram illustrating the dynamic time warping (DTW) [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Heatmap displaying normalized mean values of various features [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Violin and box plots showing the distribution of sequence length [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
read the original abstract

Treatment of cancer involves heterogeneous, complex care pathways. The relationship between these longitudinal trajectories, baseline mental health, and prognostic outcomes remains poorly understood. We introduce an interpretable time-analysis framework leveraging these temporal dynamics, analyzing care patterns spanning up to 37 years for >8,000 patients. Using Dynamic Time Warping (DTW) and Hierarchical Clustering on sequence data of healthcare encounters, we identified nine distinct, robust trajectory phenotypes. We evaluated their prognostic utility by incorporating them into generalized linear models alongside conventional clinical, demographic, and socioeconomic covariates. The trajectory clusters significantly enhanced mortality prediction and maintained independent predictive significance. Compared to a low-utilization reference group (mortality 31.5%), all eight remaining clusters exhibited substantially higher mortality odds. We uncovered two primary high-risk trajectory patterns: long-term, complex care pathways reflecting chronic disease courses (up to 196 events; mortality OR up to 3.38, 95% CI 2.13-5.37), and shorter but intense trajectories indicating rapid progression (median 78 events; OR 2.32, 95% CI 1.82-2.97). Unexpectedly, the high-utilization complexity clusters were associated with significantly lower baseline anxiety scores, highlighting a divergent relationship between trajectory intensity, mortality risk, and initial psychological burden. These results demonstrate that incorporating temporal healthcare utilization data uncovers robust trajectory phenotypes capturing multidimensional prognostic information. This offers significant explanatory power beyond established static variables for refining risk stratification in precision oncology.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript claims that Dynamic Time Warping combined with hierarchical clustering applied to longitudinal healthcare encounter sequences from >8,000 cancer patients over up to 37 years identifies nine distinct trajectory phenotypes. Incorporating these clusters as categorical predictors in generalized linear models alongside standard clinical, demographic, and socioeconomic covariates significantly enhances mortality prediction, with all non-reference clusters showing elevated mortality odds (reference group 31.5% mortality; high-risk ORs up to 3.38, 95% CI 2.13-5.37 for complex long-term pathways and 2.32, 95% CI 1.82-2.97 for intense short trajectories). The work also reports an inverse association between high-utilization clusters and baseline anxiety scores.

Significance. If the clusters prove stable and the predictive associations hold under proper validation, the framework could meaningfully advance precision oncology by demonstrating that temporal care-utilization patterns capture multidimensional prognostic information beyond static covariates. The large cohort size and extended follow-up period represent clear strengths that would support broader applicability in risk stratification once methodological gaps are closed.

major comments (3)
  1. [Methods] Methods section (clustering and GLM steps): Clustering via DTW and hierarchical clustering is described as performed on the full cohort prior to using the resulting nine phenotypes as predictors in the generalized linear models. This procedure risks data leakage, as cluster assignments can encode sample-specific encounter patterns; the reported independent predictive significance and enhancement of mortality prediction may therefore be inflated rather than reflecting generalizable signal.
  2. [Abstract and Results] Abstract and Results (mortality models): The abstract states that clusters 'significantly enhanced mortality prediction' and reports specific odds ratios with CIs, yet provides no details on the GLM specification (e.g., logistic vs. other link function), adjustment for right-censoring, competing risks, or time-to-event structure despite follow-up extending to 37 years. These omissions are load-bearing for the central prognostic claim.
  3. [Methods] Methods (cluster validation): No quantitative assessment of cluster stability, sensitivity to the free parameter (number of clusters), linkage method, or external/cross-validated performance is reported. Without such checks, the assertion that the nine phenotypes are 'robust' and clinically meaningful phenotypes rather than artifacts of the specific cohort or distance metric remains unsupported.
minor comments (2)
  1. [Abstract] The abstract uses the term 'robust trajectory phenotypes' without accompanying stability metrics or sensitivity analyses; these should be added to the main text for clarity.
  2. [Results] Notation for the nine clusters (e.g., how the reference low-utilization group is defined) could be made more explicit when first introduced in the results.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their constructive comments, which have helped us identify areas to improve the rigor and clarity of our manuscript. We address each major comment below.

read point-by-point responses

  1. Referee: [Methods] Methods section (clustering and GLM steps): Clustering via DTW and hierarchical clustering is described as performed on the full cohort prior to using the resulting nine phenotypes as predictors in the generalized linear models. This procedure risks data leakage, as cluster assignments can encode sample-specific encounter patterns; the reported independent predictive significance and enhancement of mortality prediction may therefore be inflated rather than reflecting generalizable signal.

    Authors: We acknowledge this valid concern regarding potential data leakage. The clustering was indeed performed on the entire cohort's trajectories, which span the full follow-up period. To mitigate this and ensure the predictive value is not inflated, we will revise the analysis to perform clustering exclusively on training data within a cross-validation framework. The GLM will then be fit and evaluated on held-out test sets using the cluster labels derived only from training data. This change will be detailed in the Methods section of the revised manuscript. revision: yes

  2. Referee: [Abstract and Results] Abstract and Results (mortality models): The abstract states that clusters 'significantly enhanced mortality prediction' and reports specific odds ratios with CIs, yet provides no details on the GLM specification (e.g., logistic vs. other link function), adjustment for right-censoring, competing risks, or time-to-event structure despite follow-up extending to 37 years. These omissions are load-bearing for the central prognostic claim.

    Authors: We appreciate this observation. The mortality models in the manuscript are logistic regression GLMs with a logit link function, treating mortality as a binary outcome. We did not adjust for right-censoring or use time-to-event models in the primary analysis. In the revised manuscript, we will explicitly state the GLM details (logistic regression) in the Abstract and Methods. Additionally, we will include supplementary analyses using Cox proportional hazards models to account for time-to-event and censoring, providing a more robust evaluation of the prognostic utility of the trajectory clusters. revision: yes

  3. Referee: [Methods] Methods (cluster validation): No quantitative assessment of cluster stability, sensitivity to the free parameter (number of clusters), linkage method, or external/cross-validated performance is reported. Without such checks, the assertion that the nine phenotypes are 'robust' and clinically meaningful phenotypes rather than artifacts of the specific cohort or distance metric remains unsupported.

    Authors: We agree that formal validation metrics would strengthen the claim of robustness. While the manuscript notes the clusters as 'robust' based on the consistency of the hierarchical clustering dendrogram and interpretability of the resulting phenotypes, no quantitative measures were provided. In revision, we will add quantitative assessments including the silhouette score for cluster quality, bootstrap resampling to evaluate stability of assignments, and sensitivity analyses varying the number of clusters and linkage methods. These will be reported in the Methods and Results sections to support the selection of nine phenotypes. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper first applies Dynamic Time Warping and hierarchical clustering unsupervised to the full set of care-encounter sequences to obtain nine trajectory phenotypes. These phenotypes are subsequently entered as a categorical factor into generalized linear models for mortality alongside demographic, clinical, and socioeconomic covariates. Because the clustering step uses only the sequence data and does not incorporate the mortality outcome, the reported odds ratios, predictive enhancement, and independent significance are empirical associations obtained from the GLM fit rather than quantities that reduce by construction to the clustering inputs. No equation equates a model output to a fitted clustering parameter, no self-citation supplies a load-bearing uniqueness theorem, and no ansatz or renaming of a known result is invoked. The analysis is performed in-sample, but this is a methodological choice about validation, not a definitional circularity. The derivation chain therefore remains self-contained against the paper's own stated inputs.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The analysis rests on standard assumptions of sequence alignment and unsupervised clustering applied to observational healthcare data; no new entities are postulated.

free parameters (1)
  • number of clusters
    Nine clusters were retained; the abstract does not state how this number was chosen or validated.
axioms (2)
  • domain assumption Dynamic Time Warping distances between care-event sequences reflect clinically comparable trajectories
    Invoked when applying DTW to align heterogeneous encounter sequences before clustering.
  • domain assumption Hierarchical clustering produces robust, non-overlapping phenotypes in this dataset
    Required for treating the resulting groups as distinct trajectory phenotypes.

pith-pipeline@v0.9.0 · 5594 in / 1410 out tokens · 36748 ms · 2026-05-10T03:16:07.575329+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

46 extracted references · 39 canonical work pages

  1. [1]

    Socioeconomic Disparities in Endometrial Cancer Survival in Germany: A Survival Analysis Using Population-Based Cancer Registry Data

    Ahmed Bedir et al. “Socioeconomic Disparities in Endometrial Cancer Survival in Germany: A Survival Analysis Using Population-Based Cancer Registry Data”. In:Journal of Cancer Research and Clinical Oncology148.5 (May 2022), pp. 1087–1095.issn: 1432-1335.doi:10. 1007/s00432-021-03908-9. PMID:35064816

    2022

  2. [2]

    Trends in 5-Year Cancer Survival Disparities by Race and Ethnicity in the US between 2002-2006 and 2015-2019

    Chongfa Chen et al. “Trends in 5-Year Cancer Survival Disparities by Race and Ethnicity in the US between 2002-2006 and 2015-2019”. In: Scientific Reports14.1 (Sept. 30, 2024), p. 22715.issn: 2045-2322.doi: 10.1038/s41598-024-73617-z. PMID:39349542

    work page doi:10.1038/s41598-024-73617-z 2002

  3. [3]

    Health-Related Quality of Life and Its Determi- nants among Cancer Patients: Evidence from 12,148 Patients of Indian Database

    Jyoti Dixit et al. “Health-Related Quality of Life and Its Determi- nants among Cancer Patients: Evidence from 12,148 Patients of Indian Database”. In:Health and Quality of Life Outcomes22.1 (Mar. 13, 2024), p. 26.issn: 1477-7525.doi: 10.1186/s12955- 024- 02227- 0. PMID:38481231

    work page doi:10.1186/s12955- 2024

  4. [4]

    Global Pattern and Trends of Colorectal Cancer Survival: A Systematic Review of Population-Based Registration Data

    Yufei Jiang et al. “Global Pattern and Trends of Colorectal Cancer Survival: A Systematic Review of Population-Based Registration Data”. In:Cancer Biology & Medicine19.2 (Sept. 6, 2021), pp. 175–186.issn: 2095-3941.doi: 10.20892/j.issn.2095- 3941.2020.0634 . PMID: 34486877

    work page doi:10.20892/j.issn.2095- 2021

  5. [5]

    Differences in Quality of Life and Emo- tional Well-being in Breast, Colon, and Lung Cancer Patients During Outpatient Adjuvant Chemotherapy: A Longitudinal Study

    Ainhoa Ulibarri-Ochoa et al. “Differences in Quality of Life and Emo- tional Well-being in Breast, Colon, and Lung Cancer Patients During Outpatient Adjuvant Chemotherapy: A Longitudinal Study”. In:Can- cer Nursing46.2 (Mar. 2023–Apr. 1), E99–E109.issn: 1538-9804.doi: 10.1097/NCC.0000000000001070. PMID:35283472

    work page doi:10.1097/ncc.0000000000001070 2023

  6. [6]

    A Stratification System for Breast Cancer Based on Basoluminal Tumor Cells and Spatial Tumor Architecture

    Lasse Meyer et al. “A Stratification System for Breast Cancer Based on Basoluminal Tumor Cells and Spatial Tumor Architecture”. In: Cancer Cell43.9 (Sept. 8, 2025), 1637–1655.e9.issn: 1878-3686.doi: 10.1016/j.ccell.2025.06.019. PMID:40680742. 15

    work page doi:10.1016/j.ccell.2025.06.019 2025

  7. [7]

    Prostate-Specific Antigen Stratification for Predict- ing Advanced Prostate Cancer Events in Men Approaching Age Limits for Recommended Screening

    Paul Riviere et al. “Prostate-Specific Antigen Stratification for Predict- ing Advanced Prostate Cancer Events in Men Approaching Age Limits for Recommended Screening”. In:The Journal of Urology212.5 (Nov. 2024),pp.701–709.issn:1527-3792.doi: 10.1097/JU.0000000000004138. PMID:38968170

    work page doi:10.1097/ju.0000000000004138 2024

  8. [8]

    Interval Cancer Risk after the Upper Age Limit of Screening Has Been Reached: Informing Risk Stratification in FIT-based Colorectal Cancer Screening

    Brenda J. van Stigt et al. “Interval Cancer Risk after the Upper Age Limit of Screening Has Been Reached: Informing Risk Stratification in FIT-based Colorectal Cancer Screening”. In:International Journal of Cancer156.9 (May 1, 2025), pp. 1783–1790.issn: 1097-0215.doi: 10.1002/ijc.35294. PMID:39697047

    work page doi:10.1002/ijc.35294 2025

  9. [9]

    Age Stratification and Prognostic Factor Analysis in Pediatric Differentiated Thyroid Cancer

    Ophir Winder et al. “Age Stratification and Prognostic Factor Analysis in Pediatric Differentiated Thyroid Cancer”. In:The Laryngoscope 134.11 (Nov. 2024), pp. 4818–4825.issn: 1531-4995.doi: 10.1002/ lary.31592. PMID:39387721

    2024

  10. [10]

    Persistent Disparity: Socioeconomic Deprivation and Cancer Outcomes in Patients Treated in Clinical Trials

    Joseph M. Unger et al. “Persistent Disparity: Socioeconomic Deprivation and Cancer Outcomes in Patients Treated in Clinical Trials”. In:Journal of Clinical Oncology39.12 (Apr. 20, 2021), pp. 1339–1348.issn: 0732- 183X, 1527-7755.doi:10.1200/JCO.20.02602

    work page doi:10.1200/jco.20.02602 2021

  11. [11]

    Adler, Fei Wang, David C

    Yuan Hou et al. “Cardiac Risk Stratification in Cancer Patients: A Longitudinal Patient-Patient Network Analysis”. In:PLoS medicine 18.8 (Aug. 2021), e1003736.issn: 1549-1676.doi: 10.1371/journal. pmed.1003736. PMID:34339408

    work page doi:10.1371/journal 2021

  12. [12]

    Developing Machine Learning Algorithms for Dy- namic Estimation of Progression during Active Surveillance for Prostate Cancer

    Changhee Lee et al. “Developing Machine Learning Algorithms for Dy- namic Estimation of Progression during Active Surveillance for Prostate Cancer”. In:NPJ digital medicine5.1 (Aug. 6, 2022), p. 110.issn: 2398- 6352.doi:10.1038/s41746-022-00659-w. PMID:35933478

    work page doi:10.1038/s41746-022-00659-w 2022

  13. [13]

    Prediction of Metastatic Patterns in Bladder Cancer: Spatiotemporal Progression and Development of a Novel, Web- based Platform for Clinical Utility

    Jeremy Mason et al. “Prediction of Metastatic Patterns in Bladder Cancer: Spatiotemporal Progression and Development of a Novel, Web- based Platform for Clinical Utility”. In:European Urology Open Science 32 (Oct. 2021), pp. 8–18.issn: 2666-1683.doi: 10.1016/j.euros. 2021.07.006. PMID:34667954

    work page doi:10.1016/j.euros 2021

  14. [14]

    An Artificial Intelligence Framework Integrating Longitudinal Electronic Health Records with Real-World Data Enables Continuous Pan-Cancer Prognostication

    Olivier Morin et al. “An Artificial Intelligence Framework Integrating Longitudinal Electronic Health Records with Real-World Data Enables Continuous Pan-Cancer Prognostication”. In:Nature Cancer2.7 (July 2021), pp. 709–722.issn: 2662-1347.doi:10.1038/s43018-021-00236-

    work page doi:10.1038/s43018-021-00236- 2021

  15. [15]

    Condensed Trajectory of the Temporal Correlation of Diseases and Mortality Extracted from over 300,000 Patients in Hospitals

    Hyojung Paik and Jimin Kim. “Condensed Trajectory of the Temporal Correlation of Diseases and Mortality Extracted from over 300,000 Patients in Hospitals”. In:PloS One16.10 (2021), e0257894.issn: 1932-6203.doi:10.1371/journal.pone.0257894. PMID:34610032. 16

    work page doi:10.1371/journal.pone.0257894 2021

  16. [16]

    The TRIPOD-LLM reporting guideline for studies using large language models

    Davide Placido et al. “A Deep Learning Algorithm to Predict Risk of Pancreatic Cancer from Disease Trajectories”. In:Nature Medicine29.5 (May 2023), pp. 1113–1122.issn: 1546-170X.doi:10.1038/s41591- 023-02332-5. PMID:37156936

    work page doi:10.1038/s41591- 2023

  17. [17]

    VMRA-MaR: An Asymmetry-Aware Temporal Framework for Longitudinal Breast Cancer Risk Prediction

    Zijun Sun, Solveig Thrun, and Michael Kampffmeyer. “VMRA-MaR: An Asymmetry-Aware Temporal Framework for Longitudinal Breast Cancer Risk Prediction”. In:Medical Image Computing and Computer Assisted Intervention – MICCAI 2025. Ed. by James C. Gee et al. Cham: Springer Nature Switzerland, 2026, pp. 660–669.isbn: 978-3- 032-05182-0.doi:10.1007/978-3-032-05182-0_64

    work page doi:10.1007/978-3-032-05182-0_64 2025

  18. [18]

    Visualising Disease Trajectories from Population-WideData

    Jessica Xin Hjaltelin et al. “Visualising Disease Trajectories from Population-WideData”.In:Frontiers in Bioinformatics3(2023),p.1112113. issn:2673-7647.doi: 10.3389/fbinf.2023.1112113.PMID: 36844930

    work page doi:10.3389/fbinf.2023.1112113.pmid: 2023

  19. [19]

    TheCancerSurvivalIndex—APrognosticScore Integrating Psychosocial and Biological Factors in Patients Diagnosed with Cancer or Haematologic Malignancies

    AlexanderGaigeretal.“TheCancerSurvivalIndex—APrognosticScore Integrating Psychosocial and Biological Factors in Patients Diagnosed with Cancer or Haematologic Malignancies”. In:Cancer Medicine11.18 (2022), pp. 3387–3396.issn: 2045-7634.doi:10.1002/cam4.4697

    work page doi:10.1002/cam4.4697 2022

  20. [20]

    Refining Colorectal Cancer Classification and Clinical Stratification through a Single-Cell Atlas

    Ateeq M. Khaliq et al. “Refining Colorectal Cancer Classification and Clinical Stratification through a Single-Cell Atlas”. In:Genome Biology 23.1 (May 11, 2022), p. 113.issn: 1474-760X.doi:10.1186/s13059- 022-02677-z. PMID:35538548

    work page doi:10.1186/s13059- 2022

  21. [21]

    Risk Stratification Model for Predicting the Over- all Survival of Elderly Triple-Negative Breast Cancer Patients: A Population-Based Study

    Xiaozhu Liu et al. “Risk Stratification Model for Predicting the Over- all Survival of Elderly Triple-Negative Breast Cancer Patients: A Population-Based Study”. In:Frontiers in Medicine8 (2021), p. 705515. issn: 2296-858X.doi: 10.3389/fmed.2021.705515. PMID: 34621757

    work page doi:10.3389/fmed.2021.705515 2021

  22. [22]

    The Association Between Treatment Interval and Survival in Patients With Colon or Rectal Cancer: A Systematic Review

    Ruud F. W. Franssen et al. “The Association Between Treatment Interval and Survival in Patients With Colon or Rectal Cancer: A Systematic Review”. In:World Journal of Surgery45.9 (Sept. 2021), pp. 2924–2937.issn: 1432-2323.doi: 10.1007/s00268-021-06188-z. PMID:34175967

    work page doi:10.1007/s00268-021-06188-z 2021

  23. [23]

    Treatment Intervals and Survival for Women Diagnosed with Early Breast Cancer in Queensland: The Breast Cancer Outcomes Study, a Population-Based Cohort Study

    Kou Kou et al. “Treatment Intervals and Survival for Women Diagnosed with Early Breast Cancer in Queensland: The Breast Cancer Outcomes Study, a Population-Based Cohort Study”. In:The Medical Journal of Australia219.9 (Nov. 6, 2023), pp. 409–416.issn: 1326-5377.doi: 10.5694/mja2.52091. PMID:37667512

    work page doi:10.5694/mja2.52091 2023

  24. [24]

    Impact of the Diagnosis-to-Treatment Interval on the Survival of Patients with Papillary Thyroid Cancer

    Tingting Wei et al. “Impact of the Diagnosis-to-Treatment Interval on the Survival of Patients with Papillary Thyroid Cancer”. In:Journal of Investigative Surgery: The Official Journal of the Academy of Surgical Research38.1 (Dec. 2025), p. 2456463.issn: 1521-0553.doi: 10.1080/ 08941939.2025.2456463. PMID:39956540. 17

    work page arXiv 2025

  25. [25]

    Identifying Temporal Patterns in Patient Disease Trajectories Using Dynamic Time Warping: A Population- Based Study

    Alexia Giannoula et al. “Identifying Temporal Patterns in Patient Disease Trajectories Using Dynamic Time Warping: A Population- Based Study”. In:Scientific Reports8.1 (Mar. 9, 2018), p. 4216.issn: 2045-2322.doi:10.1038/s41598-018-22578-1

    work page doi:10.1038/s41598-018-22578-1 2018

  26. [26]

    Exploring Long-Term Breast Cancer Survivors’ Care Trajectories Using Dynamic Time Warping-Based Unsupervised Clustering

    Alexia Giannoula et al. “Exploring Long-Term Breast Cancer Survivors’ Care Trajectories Using Dynamic Time Warping-Based Unsupervised Clustering”. In:Journal of the American Medical Informatics Associa- tion: JAMIA31.4 (Apr. 3, 2024), pp. 820–831.issn: 1527-974X.doi: 10.1093/jamia/ocad251. PMID:38193340

    work page doi:10.1093/jamia/ocad251 2024

  27. [27]

    Socioeconomic Inequalities in Cervical Cancer Mortality in Canada, 1990 and 2019: A Trend Analysis

    M. Fay, M. Hu, and M. Hajizadeh. “Socioeconomic Inequalities in Cervical Cancer Mortality in Canada, 1990 and 2019: A Trend Analysis”. In:Public Health227 (Feb. 2024), pp. 210–218.issn: 1476-5616.doi: 10.1016/j.puhe.2023.12.014. PMID:38241902

    work page doi:10.1016/j.puhe.2023.12.014 1990

  28. [28]

    Mančinska and D

    Thomas P. Lawler et al. “Area-Level Socioeconomic Status Is Associated with Colorectal Cancer Screening, Incidence and Mortality in the US: A Systematic Review and Meta-Analysis”. In:Social Science & Medicine (1982)381 (Sept. 2025), p. 118212.issn: 1873-5347.doi: 10.1016/j. socscimed.2025.118212. PMID:40472644

    work page doi:10.1016/j 1982

  29. [29]

    Race and Ethnicity Disparities in Cardiovascular and Cancer Mortality: The Role of Socioeconomic Status-a Systematic Review and Meta-analysis

    João L. Marôco, Mahdiyeh M. Manafi, and Laura L. Hayman. “Race and Ethnicity Disparities in Cardiovascular and Cancer Mortality: The Role of Socioeconomic Status-a Systematic Review and Meta-analysis”. In:Journal of Racial and Ethnic Health Disparities12.1 (Feb. 2025), pp. 285–297.issn: 2196-8837.doi: 10.1007/s40615- 023- 01872- 3. PMID:38038904

    work page doi:10.1007/s40615- 2025

  30. [30]

    Social Support, Social Strain, Stressful Life Events and Mortality Among Postmenopausal Women With Breast Cancer

    Fengge Wang et al. “Social Support, Social Strain, Stressful Life Events and Mortality Among Postmenopausal Women With Breast Cancer”. In:Psycho-Oncology33.11 (Nov. 2024), e70013.issn: 1099-1611.doi: 10.1002/pon.70013. PMID:39520665

    work page doi:10.1002/pon.70013 2024

  31. [31]

    Side Effects of Endocrine Therapy Are Associated With Depression and Anxiety in Breast Cancer Patients Accepting Endocrine Therapy: A Cross-Sectional Study in China

    Rong Zhao, Hulin Liu, and Jinnan Gao. “Side Effects of Endocrine Therapy Are Associated With Depression and Anxiety in Breast Cancer Patients Accepting Endocrine Therapy: A Cross-Sectional Study in China”. In:Frontiers in Psychology13 (2022), p. 905459.issn: 1664- 1078.doi:10.3389/fpsyg.2022.905459. PMID:35615194

    work page doi:10.3389/fpsyg.2022.905459 2022

  32. [32]

    Association between Socioeconomic Factors at Diagnosis and Survival in Breast Cancer: A Population-based Study

    Peng Ji et al. “Association between Socioeconomic Factors at Diagnosis and Survival in Breast Cancer: A Population-based Study”. In:Cancer Medicine9.5 (Mar. 2020), pp. 1922–1936.issn: 2045-7634, 2045-7634. doi:10.1002/cam4.2842

    work page doi:10.1002/cam4.2842 2020

  33. [33]

    Socio-Economic Inequalities in Lung Cancer Outcomes: An Overview of Systematic Reviews

    Daniel Redondo-Sánchez et al. “Socio-Economic Inequalities in Lung Cancer Outcomes: An Overview of Systematic Reviews”. In:Can- cers14.2 (Jan. 13, 2022), p. 398.issn: 2072-6694.doi: 10 . 3390 / cancers14020398. 18

    2022

  34. [34]

    The Impact of COVID-19 on Cancer Care of Outpatients with Low Socioeconomic Status

    E.L. Zeilinger et al. “The Impact of COVID-19 on Cancer Care of Outpatients with Low Socioeconomic Status”. In:International Journal of Cancer151.1 (2022), pp. 77–82.issn: 1097-0215.doi: 10.1002/ijc. 33960

    work page doi:10.1002/ijc 2022

  35. [35]

    Self-Rated Health and Health-Related Quality of Life among Cancer Patients: The Serial Multiple Mediation of Anxiety and Depression

    Shuowen Fang et al. “Self-Rated Health and Health-Related Quality of Life among Cancer Patients: The Serial Multiple Mediation of Anxiety and Depression”. In:BMC psychology12.1 (July 30, 2024), p. 415.issn: 2050-7283.doi:10.1186/s40359-024-01919-y. PMID:39080782

    work page doi:10.1186/s40359-024-01919-y 2024

  36. [36]

    Effect of Depression and Em- powerment on Medication Adherence in Patients with Breast Cancer: A Descriptive Survey

    Sookyung Jeong and Eun Jeong Kim. “Effect of Depression and Em- powerment on Medication Adherence in Patients with Breast Cancer: A Descriptive Survey”. In:BMC nursing24.1 (Jan. 14, 2025), p. 47.issn: 1472-6955.doi:10.1186/s12912-024-02680-8. PMID:39806337

    work page doi:10.1186/s12912-024-02680-8 2025

  37. [37]

    The Role of Coping and Posttraumatic Stress in Fostering Posttraumatic Growth and Quality of Life Among Women with Breast Cancer

    Amy R. Senger et al. “The Role of Coping and Posttraumatic Stress in Fostering Posttraumatic Growth and Quality of Life Among Women with Breast Cancer”. In:Journal of Clinical Psychology in Medical Settings31.2 (June 2024), pp. 368–378.issn: 1573-3572.doi:10.1007/ s10880-023-09977-x. PMID:37803095

    2024

  38. [38]

    The Mediating Role of Depression in the Association Between Death Anxiety and Quality of Life in Elderly Prostate Cancer Patients

    Haifeng Song, Fengyi He, and Xiaoling Zhang. “The Mediating Role of Depression in the Association Between Death Anxiety and Quality of Life in Elderly Prostate Cancer Patients”. In:Actas Espanolas De Psiquiatria53.4 (Aug. 2025), pp. 791–801.issn: 1578-2735.doi: 10. 62641/aep.v53i4.1921. PMID:40791044

    2025

  39. [39]

    Sakoe and S

    H. Sakoe and S. Chiba. “Dynamic Programming Algorithm Optimiza- tion for Spoken Word Recognition”. In:IEEE Transactions on Acoustics, Speech, and Signal Processing26.1 (Feb. 1978), pp. 43–49.issn: 0096- 3518.doi:10.1109/TASSP.1978.1163055

    work page doi:10.1109/tassp.1978.1163055 1978

  40. [40]

    A Dynamic Time Warping Extension to Consensus Weight-Based Cachexia Criteria Improves Prediction of Cancer Patient Outcomes

    Noah Forrest et al. “A Dynamic Time Warping Extension to Consensus Weight-Based Cachexia Criteria Improves Prediction of Cancer Patient Outcomes”. In:JCSM communications8.1 (2025), e107.issn: 2996-1394. doi:10.1002/rco2.107. PMID:40151817

    work page doi:10.1002/rco2.107 2025

  41. [41]

    IdentifyingCommonDiseaseTrajectoriesofAlzheimer’s Disease with Electronic Health Records

    MingzhouFuetal.“IdentifyingCommonDiseaseTrajectoriesofAlzheimer’s Disease with Electronic Health Records”. In:EBioMedicine118 (Aug. 2025), p. 105831.issn: 2352-3964.doi: 10 . 1016 / j . ebiom . 2025 . 105831. PMID:40592257

    2025

  42. [42]

    Analyzing Patient Trajectories With Artificial Intelligence

    Ahmed Allam et al. “Analyzing Patient Trajectories With Artificial Intelligence”. In:Journal of Medical Internet Research23.12 (Dec. 3, 2021), e29812.issn: 1438-8871.doi:10.2196/29812

    work page doi:10.2196/29812 2021

  43. [43]

    de Moura, N

    “Dynamic Time Warping”. In: Meinard Müller.Information Retrieval for Music and Motion. Berlin, Heidelberg: Springer Berlin Heidelberg, 2007, pp. 69–84.isbn: 978-3-540-74048-3.doi:10.1007/978-3-540- 74048-3_4. 19

    work page doi:10.1007/978-3-540- 2007

  44. [44]

    Temporal Representation of Care Trajectories of Cancer Patients Using Data from a Regional Information System: An Application in Breast Cancer

    Gautier Defossez et al. “Temporal Representation of Care Trajectories of Cancer Patients Using Data from a Regional Information System: An Application in Breast Cancer”. In:BMC Medical Informatics and Decision Making14.1 (Dec. 2014), p. 24.issn: 1472-6947.doi: 10 . 1186/1472-6947-14-24

    2014

  45. [45]

    Ingleby et al

    Fiona C. Ingleby et al. “An Investigation of Cancer Survival Inequalities Associated with Individual-Level Socio-Economic Status, Area-Level Deprivation, and Contextual Effects, in a Cancer Patient Cohort in England and Wales”. In:BMC Public Health22.1 (Dec. 2022), p. 90. issn: 1471-2458.doi:10.1186/s12889-022-12525-1

    work page doi:10.1186/s12889-022-12525-1 2022

  46. [46]

    Using Sequences of Life-Events to Predict Hu- man Lives

    Germans Savcisens et al. “Using Sequences of Life-Events to Predict Hu- man Lives”. In:Nature Computational Science4.1 (Jan. 2024), pp. 43– 56.issn: 2662-8457.doi:10.1038/s43588-023-00573-5. 20 A Supplementary Information Figure S1: Heatmap displaying the prevalence of various diagnoses (ICD codes) across 9 clusters. Clusters 4 and 5 exhibit the most sign...

    work page doi:10.1038/s43588-023-00573-5 2024