Tag: statistics
Pain. 2024 Jul 9. doi: 10.1097/j.pain.0000000000003336. Online ahead of print.
ABSTRACT
Chronic pain is a serious and prevalent condition that can affect many facets of life. However, uncertainty remains regarding the strength of the association between chronic pain and death and whether the association is causal. We investigate the pain-mortality relationship using data from 19,971 participants aged 51+ years in the 1998 wave of the U.S. Health and Retirement Study. Propensity score matching and inverse probability weighting are combined with Cox proportional hazards models to investigate whether exposure to chronic pain (moderate or severe) has a causal effect on mortality over a 20-year follow-up period. Hazard ratios (HRs) with 95% confidence intervals (CIs) are reported. Before adjusting for confounding, we find a strong association between chronic pain and mortality (HR: 1.32, 95% CI: 1.26-1.38). After adjusting for confounding by sociodemographic and health variables using a range of propensity score methods, the estimated increase in mortality hazard caused by pain is more modest (5%-9%) and the results are often also compatible with no causal effect (95% CIs for HRs narrowly contain 1.0). This attenuation highlights the role of confounders of the pain-mortality relationship as potentially modifiable upstream risk factors for mortality. Posing the depressive symptoms variable as a mediator rather than a confounder of the pain-mortality relationship resulted in stronger evidence of a modest causal effect of pain on mortality (eg, HR: 1.08, 95% CI: 1.01-1.15). Future work is required to model exposure-confounder feedback loops and investigate the potentially cumulative causal effect of chronic pain at multiple time points on mortality.
PMID:38981067 | DOI:10.1097/j.pain.0000000000003336
Med Phys. 2024 Jul 9. doi: 10.1002/mp.17289. Online ahead of print.
ABSTRACT
BACKGROUND: A comprehensive collection of data on doses in adult computed tomography procedures in Australia has not been undertaken for some time. This is largely due to the effort involved in collecting the data required for calculating the population dose. This data collection effort can be greatly reduced, and the coverage increased, if the process can be automated without major changes to the workflow of the imaging facilities providing the data. Success would provide a tool to determine a truly national assessment of the dose incurred through diagnostic imaging in Australia.
PURPOSE: The aims of this study were to develop an automated tool to categorize electronic records of imaging procedures into a standardized set of broad procedure types, to validate the tool by applying it to data collected from nine facilities, and to assess the feasibility of applying the automated tool to compute population dose and determine the data manipulations required.
METHODS: A rule-based classifier was implemented capitalizing on semantic and clinical rules. The keyword list was initially built from 609 unique study descriptions. It was then refined using an additional 414 unique study descriptions. The classifier was then tested on an additional 1198 unique study descriptions. Input from a radiologist provided the ground truth for the refinement of the classifier.
RESULTS: From a sample of 238 139 studies containing 2794 unique study descriptions, the classifier correctly classified 2789 study types with only five misclassifications, demonstrating the feasibility of automating the process and the need for data pre-processing. Dose statistics for 21 categories were compiled using the 238 139 studies.
CONCLUSION: The classifier achieved excellent classification results using the testing data supplied by the facilities. However, since all data supplied were from public facilities, the performance of the classifier may be biased. The performance of the classifier is yet to be tested on a more representative mix of private and public facilities.
PMID:38981056 | DOI:10.1002/mp.17289
ESC Heart Fail. 2024 Jul 9. doi: 10.1002/ehf2.14918. Online ahead of print.
ABSTRACT
AIMS: Assessing the risk for HF rehospitalization is important for managing and treating patients with HF. To address this need, various risk prediction models have been developed. However, none of them used deep learning methods with real-world data. This study aimed to develop a deep learning-based prediction model for HF rehospitalization within 30, 90, and 365 days after acute HF (AHF) discharge.
METHODS AND RESULTS: We analysed the data of patients admitted due to AHF between January 2014 and January 2019 in a tertiary hospital. In performing deep learning-based predictive algorithms for HF rehospitalization, we use hyperbolic tangent activation layers followed by recurrent layers with gated recurrent units. To assess the readmission prediction, we used the AUC, precision, recall, specificity, and F1 measure. We applied the Shapley value to identify which features contributed to HF readmission. Twenty-two prognostic features exhibiting statistically significant associations with HF rehospitalization were identified, consisting of 6 time-independent and 16 time-dependent features. The AUC value shows moderate discrimination for predicting readmission within 30, 90, and 365 days of follow-up (FU) (AUC:0.63, 0.74, and 0.76, respectively). The features during the FU have a relatively higher contribution to HF rehospitalization than features from other time points.
CONCLUSIONS: Our deep learning-based model using real-world data could provide valid predictions of HF rehospitalization in 1 year follow-up. It can be easily utilized to guide appropriate interventions or care strategies for patients with HF. The closed monitoring and blood test in daily clinics are important for assessing the risk of HF rehospitalization.
PMID:38981003 | DOI:10.1002/ehf2.14918
Aust J Prim Health. 2024 Jul;30:PY24017. doi: 10.1071/PY24017.
ABSTRACT
Background Large datasets exist in Australia that make de-identified primary healthcare data extracted from clinical information systems available for research use. This study reviews these datasets for their capacity to provide insight into chronic disease care for Aboriginal and Torres Strait Islander peoples, and the extent to which the principles of Indigenous Data Sovereignty are reflected in data collection and governance arrangements. Methods Datasets were included if they collect primary healthcare clinical information system data, collect data nationally, and capture Aboriginal and Torres Strait Islander peoples. We searched PubMed and the public Internet for data providers meeting the inclusion criteria. We developed a framework to assess data providers across domains, including representativeness, usability, data quality, adherence with Indigenous Data Sovereignty and their capacity to provide insights into chronic disease. Datasets were assessed against the framework based on email interviews and publicly available information. Results We identified seven datasets. Only two datasets reported on chronic disease, collected data nationally and captured a substantial number of Aboriginal and Torres Strait Islander patients. No dataset was identified that captured a significant number of both mainstream general practice clinics and Aboriginal Community Controlled Health Organisations. Conclusions It is critical that more accurate, comprehensive and culturally meaningful Aboriginal and Torres Strait Islander healthcare data are collected. These improvements must be guided by the principles of Indigenous Data Sovereignty and Governance. Validated and appropriate chronic disease indicators for Aboriginal and Torres Strait Islander peoples must be developed, including indicators of social and cultural determinants of health.
PMID:38981000 | DOI:10.1071/PY24017
Transpl Infect Dis. 2024 Jul 9:e14337. doi: 10.1111/tid.14337. Online ahead of print.
ABSTRACT
BACKGROUND: Cytomegalovirus (CMV) is a driver of negative outcomes after lung transplant (LTX) and primary prophylaxis (PPX) with valganciclovir (VGC) is standard-of-care. VGC is associated with myelosuppression, prompting interest in letermovir (LTV).
METHODS: Adults receiving LTX between April 1, 2015, and July 30, 2022, at our institution were evaluated. Patients were excluded if low CMV risk (D-/R-), survived <90 days post-LTX, or transferred care before PPX withdrawal. Primary outcomes were leukopenia (white blood cell count [WBC] ≤ 3.0 × 109/L), severe leukopenia (WBC ≤ 2.0 × 109/L), and neutropenia (absolute neutrophil count ≤ 1500 cells/µL) requiring granulocyte-colony stimulating factor (GCSF) on PPX. Secondary outcomes included breakthrough CMV infection and post-PPX CMV infection.
RESULTS: 204 patients met inclusion criteria: 175 patients on VGC and 29 patients on LTV (after VGC conversion). Most patients received bilateral LTX (62.7%) with non-lymphocyte-depleting induction (96.6%) and moderate-risk serostatus (D+/R+, 48.5%). Patients transitioned from VGC to LTV after a mean of 178 days (SD 80.8 days) post-transplant. Patients on VGC experienced significantly more leukopenia (82.3% vs. 58.6%, p = 0.008), severe leukopenia (57.1% vs. 31.0%, p = 0.016), and neutropenia requiring GCSF (70.9% vs. 51.7%, p = 0.048). Breakthrough (5.7% vs. 3.4%, p = 0.955) and post-PPX (24.6% vs. 37.9%, p = 0.199) infections were similar. A subgroup analysis of patients with high-risk serostatus showed similar trends, though did not reach statistical significance.
CONCLUSIONS: In this single-center study, the incidence of leukopenia and neutropenia requiring GCSF were reduced with LTV compared to VGC. Breakthrough and post-PPX infections were not significantly different. This evidence suggests that LTV has comparable efficacy with reduced myelosuppression compared to VGC in LTX recipients, and may be an appropriate alternative for PPX.
PMID:38980979 | DOI:10.1111/tid.14337
Influenza Other Respir Viruses. 2024 Jul;18(7):e13346. doi: 10.1111/irv.13346.
ABSTRACT
BACKGROUND: Changes in the epidemiology of illnesses caused by respiratory syncytial virus (RSV) infection following the COVID-19 pandemic are reported. The New Zealand (NZ) COVID-19 situation was unique; RSV community transmission was eliminated with the 2020 border closure, with a rapid and large increase in hospitalizations following the relaxation of social isolation measures and the opening of an exclusive border with Australia.
METHODS: This active population-based surveillance compared the age-specific incidence and seasonality of RSV-associated hospitalizations in Auckland, NZ, for 2 years before and after the 2020 border closures. Hospitalisation rates between years were compared by age, ethnicity (European/other, Māori, Pacific and Asian) and socioeconomic group (1 = least, 5 = most deprived).
RESULTS: There was no RSV transmission in 2020. In all other years, hospitalisation rates were highest for people of Pacific versus other ethnic groups and for people living in the most deprived quintile of households. RSV hospitalisation rates were higher in 2021 and 2022 than in 2018-19. The epidemic peak was higher in 2021, but not 2022, and the duration was shorter than in 2018-19. In 2021, the increase in RSV hospitalisation rates was significant across all age, sex, ethnic and socioeconomic groups. In 2022, the increase in hospitalisation rates was only significant in one age (1- < 3 years), one ethnic (Asian) and one socioeconomic group (quintile 2).
CONCLUSIONS: COVID pandemic responses altered RSV-related hospitalisation seasonal patterns. Atypical features of RSV hospitalisation epidemiology were the increase in rates in older children and young adults, which lessened in 2022. Despite these variations, RSV hospitalisations in NZ continue to disproportionately affect individuals of Pacific ethnicity and those living in more socioeconomically deprived households. Whilst future public health strategies focused on RSV disease mitigation need to consider the potential shifts in epidemiological patterns when the transmission is disrupted, these variances must be considered in the context of longer-standing patterns of unequal disease distribution.
PMID:38980967 | DOI:10.1111/irv.13346