Elena Rakuša, Constantin Reinke, Gabriele Doblhammer, Lukas Radbruch, Matthias Schmid, Thomas Welchowski
{"title":"Dementia as a predictor of palliative care: Uncovering patient patterns based on German claims data.","authors":"Elena Rakuša, Constantin Reinke, Gabriele Doblhammer, Lukas Radbruch, Matthias Schmid, Thomas Welchowski","doi":"10.1186/s12904-025-01672-y","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Palliative care aims to ensure a dignified and self-determined life for people facing the end of life. While palliative care is established for tumor diseases, it's notably absent from German medical guidelines for other progressive diseases with an unfavorable prognosis such as dementia. This study will identify predictors of palliative care use in older patients and explore how these predictors relate to the probability of palliative care.</p><p><strong>Methods: </strong>We used data from the largest German health insurance company of people over 50 years of age from the period 2014-2019. The analysis focused on the last year of life. Outcomes were outpatient and inpatient palliative care and predictors were demographics, comorbidities, therapeutic remedies and rehabilitation, care and medical interventions, medication and patient group. Combined logistic regression models and discrete conditional inference survival forests were used to predict the utilization of outpatient and inpatient palliative care. For evaluation we used concordance-index and calibration plots. We identified the most important predictors by using a permutation approach and the log-loss metric.</p><p><strong>Results: </strong>The study cohort for the analysis of inpatient palliative care comprised 43,896 patients, while the cohort for the analysis of outpatient palliative care included a total of 37,430 patients. The models had appropriate discriminatory power (inpatient palliative care: concordance-index = 0.737 (95%CI = 0.721-0.754); outpatient palliative care: concordance-index = 0.689; 95%CI = 0.675-0.704) and showed appropriate calibration. A diagnosis of dementia, like a diagnosis of cancer, is predictive of inpatient palliative care and outpatient palliative care. We observed a lower probability for inpatient and for outpatient palliative care for dementia patients compared to cancer patients.</p><p><strong>Conclusions: </strong>The findings highlight the need to focus palliative care on other patient groups besides cancer patients, such as dementia patients, and to facilitate access for all patients.</p>","PeriodicalId":48945,"journal":{"name":"BMC Palliative Care","volume":"24 1","pages":"46"},"PeriodicalIF":2.5000,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"BMC Palliative Care","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1186/s12904-025-01672-y","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"HEALTH CARE SCIENCES & SERVICES","Score":null,"Total":0}
引用次数: 0
Abstract
Background: Palliative care aims to ensure a dignified and self-determined life for people facing the end of life. While palliative care is established for tumor diseases, it's notably absent from German medical guidelines for other progressive diseases with an unfavorable prognosis such as dementia. This study will identify predictors of palliative care use in older patients and explore how these predictors relate to the probability of palliative care.
Methods: We used data from the largest German health insurance company of people over 50 years of age from the period 2014-2019. The analysis focused on the last year of life. Outcomes were outpatient and inpatient palliative care and predictors were demographics, comorbidities, therapeutic remedies and rehabilitation, care and medical interventions, medication and patient group. Combined logistic regression models and discrete conditional inference survival forests were used to predict the utilization of outpatient and inpatient palliative care. For evaluation we used concordance-index and calibration plots. We identified the most important predictors by using a permutation approach and the log-loss metric.
Results: The study cohort for the analysis of inpatient palliative care comprised 43,896 patients, while the cohort for the analysis of outpatient palliative care included a total of 37,430 patients. The models had appropriate discriminatory power (inpatient palliative care: concordance-index = 0.737 (95%CI = 0.721-0.754); outpatient palliative care: concordance-index = 0.689; 95%CI = 0.675-0.704) and showed appropriate calibration. A diagnosis of dementia, like a diagnosis of cancer, is predictive of inpatient palliative care and outpatient palliative care. We observed a lower probability for inpatient and for outpatient palliative care for dementia patients compared to cancer patients.
Conclusions: The findings highlight the need to focus palliative care on other patient groups besides cancer patients, such as dementia patients, and to facilitate access for all patients.
期刊介绍:
BMC Palliative Care is an open access journal publishing original peer-reviewed research articles in the clinical, scientific, ethical and policy issues, local and international, regarding all aspects of hospice and palliative care for the dying and for those with profound suffering related to chronic illness.