Survival Analysis Approach Research Articles

On the Reliability of Machine Learning Models for Survival Analysis When Cure Is a Possibility

In classical survival analysis, it is assumed that all the individuals will experience the event of interest. However, if there is a proportion of subjects who will never experience the event, then a standard survival approach is not appropriate, and cure models should be considered instead. This paper deals with the problem of adapting a machine learning approach for classical survival analysis to a situation when cure (i.e., not suffering the event) is a possibility. Specifically, a brief review of cure models and recent machine learning methodologies is presented, and an adaptation of machine learning approaches to account for cured individuals is introduced. In order to validate the proposed methods, we present an extensive simulation study in which we compare the performance of the adapted machine learning algorithms with existing cure models. The results show the good behavior of the semiparametric or the nonparametric approaches, depending on the simulated scenario. The practical utility of the methodology is showcased through two real-world dataset illustrations. In the first one, the results show the gain of using the nonparametric mixture cure model approach. In the second example, the results show the poor performance of some machine learning methods for small sample sizes.

Predicting Progression to Clinical Alzheimer's Disease Dementia Using the Random Survival Forest.

Assessing the risk of developing clinical Alzheimer's disease (AD) dementia, by machine learning survival analysis approaches, among participants registered in Alzheimer's Disease Centers is important for AD dementia management. To construct a prediction model for the onset time of clinical AD dementia using the National Alzheimer Coordinating Center (NACC) and the Alzheimer's Disease Neuroimaging Initiative (ADNI) registered cohorts. A model was constructed using the Random Survival Forest (RSF) approach and internally and externally validated on the NACC cohort and the ADNI cohort. An R package and a Shiny app were provided for accessing the model. We built a predictive model having the six predictors: delayed logical memory score (story recall), CDR® Dementia Staging Instrument - Sum of Boxes, general orientation in CDR®, ability to remember dates and ability to pay bills in the Functional Activities Questionnaire, and patient age. The C indices of the model were 90.82% (SE = 0.71%) and 86.51% (SE = 0.75%) in NACC and ADNI respectively. The time-dependent AUC and accuracy at 48 months were 92.48% (SE = 1.12%) and 88.66% (SE = 1.00%) respectively in NACC, and 90.16% (SE = 1.12%) and 85.00% (SE = 1.14%) respectively in ADNI. The model showed good prediction performance and the six predictors were easy to obtain, cost-effective, and non-invasive. The model could be used to inform clinicians and patients on the probability of developing clinical AD dementia in 4 years with high accuracy.

Survival Analysis and Contributing Factors among PCR-Confirmed Adult Inpatients during the Endemic Phase of COVID-19.

In May 2023, the global health emergency status of COVID-19 concluded, marking the onset of an endemic era. This study assessed survival rates among PCR-confirmed adult inpatients during this phase and determined contributing factors. Employing a survival analysis approach, this investigation utilized a nationwide Mexican cohort encompassing 152 adult inpatients. Survival rates were computed using the Kaplan-Meier method, and a proportional Cox model identified mortality risk factors. Survival rates remained above 65% on day 14 after admission. Vaccination status, including the number of doses administered, was not significantly associated with fatal outcomes. Chronic kidney disease or a history of immunosuppression (due to any cause) increased mortality risk. Our findings underscore the persistent severity of COVID-19 beyond the global health emergency, emphasizing the necessity for tailored interventions for vulnerable patients.

Comparison of survival analysis approaches to modelling age at first sex among youth in Kisesa Tanzania.

Many studies analyze sexual and reproductive event data using descriptive life tables. Survival analysis has better power to estimate factors associated with age at first sex (AFS), but proportional hazards models may not be right model to use. This study used accelerated failure time (AFT) models, restricted Mean Survival time model (RMST) models, with semi and non-parametric methods to assess age at first sex (AFS), factors associated with AFS, and verify underlying assumptions for each analysis. Self-reported sexual debut data was used from respondents 15-24 years in eight cross-sectional surveys between 1994-2016, and from adolescents' survey in an observational community study (2019-2020) in northwest Tanzania. Median AFS was estimated in each survey using non-parametric and parametric models. Cox regression, AFT parametric models (exponential, gamma, generalized gamma, Gompertz, Weibull, log-normal and log-logistic), and RMST were used to estimate and identify factors associated with AFS. The models were compared using Akaike information criterion (AIC) and Bayesian information criterion (BIC), where lower values represent a better model fit. The results showed that in every survey, the Cox regression model had higher AIC and BIC compared to the other models. Overall, AFT had the best fit in every survey round. The estimated median AFS using the parametric and non-parametric methods were close. In the adolescent survey, log-logistic AFT showed that females and those attending secondary and higher education level had a longer time to first sex (Time ratio (TR) = 1.03; 95% CI: 1.01-1.06, TR = 1.05; 95% CI: 1.02-1.08, respectively) compared to males and those who reported not being in school. Cell phone ownership (TR = 0.94, 95% CI: 0.91-0.96), alcohol consumption (TR = 0.88; 95% CI: 0.84-0.93), and employed adolescents (TR = 0.95, 95% CI: 0.92-0.98) shortened time to first sex. The AFT model is better than Cox PH model in estimating AFS among the young population.

A hybrid CNN-RNN approach for survival analysis in a Lung Cancer Screening study

In this study, we present a hybrid CNN-RNN approach to investigate long-term survival of subjects in a lung cancer screening study. Subjects who died of cardiovascular and respiratory causes were identified whereby the CNN model was used to capture imaging features in the CT scans and the RNN model was used to investigate time series and thus global information. To account for heterogeneity in patients' follow-up times, two different variants of LSTM models were evaluated, each incorporating different strategies to address irregularities in follow-up time. The models were trained on subjects who underwent cardiovascular and respiratory deaths and a control cohort matched to participant age, gender, and smoking history. The combined model can achieve an AUC of 0.76 which outperforms humans at cardiovascular mortality prediction. The corresponding F1 and Matthews Correlation Coefficient are 0.63 and 0.42 respectively. The generalisability of the model is further validated on an ‘external’ cohort. The same models were applied to survival analysis with the Cox Proportional Hazard model. It was demonstrated that incorporating the follow-up history can lead to improvement in survival prediction. The Cox neural network can achieve an IPCW C-index of 0.75 on the internal dataset and 0.69 on an external dataset. Delineating subjects at increased risk of cardiorespiratory mortality can alert clinicians to request further more detailed functional or imaging studies to improve the assessment of cardiorespiratory disease burden. Such strategies may uncover unsuspected and under-recognised pathologies thereby potentially reducing patient morbidity.

Survival and movements of brown‐headed nuthatches after translocation to the Missouri Ozarks

AbstractThe brown‐headed nuthatch (Sitta pusilla) was likely extirpated from Missouri in the early 1900s as a result of habitat loss through extensive logging. Conservation partners including the Missouri Department of Conservation, United States Forest Service, University of Missouri, and others, relocated 102 brown‐headed nuthatches from Ouachita National Forest in Arkansas to Mark Twain National Forest in Missouri in 2020 and 2021 to establish a local population. We tracked 50 individuals for 24 ± 11 (median ± SD) days after release using radio telemetry and monitored movements in relation to sex and whether a bird was captured alone or as part of a group. We examined 25‐day survival using a spatial Cormack Jolly Seber (sCJS) model to account for likely dispersal outside of our study area. All birds survived capture, transport, and release. The 25‐day survival estimate was 0.56 (95% CI = 0.36–0.78) in 2020 and 0.68 (95% CI = 0.46–0.89) in 2021. Mean total distance moved following release was 5,670 (SD = 3,407) m and mean daily movement was 243 (SD = 135) m which was less than the mean daily dispersal distance estimated from our sCJS model (420 m, SD = 221). Our short‐term monitoring suggests initial success of the translocation, but continued monitoring is needed to evaluate long‐term success. Our survival analysis approach, which included movement data, illustrates the importance of considering dispersal when estimating survival.

The high expression of glial cell line-derived neurotrophic factor receptor alpha Ⅱ (GFRA2) as a predictor of poor prognosis in gastric cancer patients: A survival and regression analysis approach

Gastric cancer has high mortality rates worldwide. Therefore, there is a need to identify prognostic biomarkers. This study evaluated the association between GFRA2 expression levels with clinicopathological features and prognosis in gastric cancer using data extracted from The Cancer Genome Atlas (TCGA) database and a series of algorithms. Survival analysis was performed using the Kaplan-Meier method. Univariate and multivariate Cox regression analyses were used to analyze the association between different clinical features and survival. Single-sample gene set enrichment analysis (GSEA) was used to examine the correlation between GFRA2 expression and immune infiltration. The results showed that the expression of GFRA2 in tumor samples was significantly lower than that in normal samples. High expression of GFRA2 was significantly associated with histological type, histologic grade, and worse overall survival, disease-specific survival, and progression-free survival. The univariate Cox analysis showed that the expression of GFRA2 was significantly correlated with T stage, N stage, M stage, and age. The multivariate analysis identified GFRA2 expression as an independent prognostic factor for gastric cancer. GSEA showed that GFRA2 might regulate the calcium signaling pathway, focus adhesion, olfactory conduction, the extracellular matrix glycoproteins, and response to the Leishmania parasitic infection. GFRA2 showed a significant moderate positive correlation with the infiltration of mast cells. In summary, a high expression of GFRA2 may contribute to poor survival in gastric cancer patients and could be used as a potential prognostic biomarker.

Exploring variation of length of haul and associated freight transport emission of Indian establishments: A survival analysis approach

Rise in length of haul (LOH) of business establishments and the associated greenhouse gas (GHG) emissions are a major concern for planners and policymakers, require extensive investigation. This paper quantifies various factors like business age, gross floor area, employment, the build environment variable on LOH, and associated GHG emission using EBFS data. Survival analysis has been utilized for the analysis. The aggregated LOH model provides a comparative analysis of LOH across different industry sectors and demonstrate that gross floor area, business age, and employment are positively associated with LOH. Subsequently, disaggregated LOH models are reveal the influencing parameter of LOH of a particular industry category. Lastly, GHG models show the influence of fleet ownership level on GHG emissions across different industry sectors. The outcomes will be beneficial for planners and policymakers to develop industry-specific regulations on land use, employment restructuring, and subsidies for owning different types of commercial vehicles.

Lower Cerebrospinal Fluid Amyloid-β42 Predicts Sooner Time to Antipsychotic Use in Alzheimer's Disease.

Cerebrospinal fluid (CSF) biomarkers of amyloid-β42 (Aβ42) and phosphorylated-tau help clinicians accurately diagnose Alzheimer's disease (AD). Whether biomarkers help prognosticate behavioral and psychological symptoms of dementia (BPSD) is unclear. Determine whether CSF biomarker levels aid prognostication of BPSD in AD. This retrospective cohort study included patients over 65 with a diagnosis of AD based on CSF biomarkers. We measured time from CSF testing to the first antipsychotic use in the following months. We then analyzed time to antipsychotic (AP) use with respect to Aβ42, total tau, phosphorylated tau, and amyloid-to-tau index using a survival analysis approach. Of 86 AD patients (average 72±5 years, 46.5% male), 11 patients (12.7%) received APs following CSF testing. Patients with Aβ42 below the median had sooner time-to-AP use. This was significant on a log-rank test (p = 0.04). There was no difference in time-to-AP use if the group was stratified by levels of total tau, phosphorylated tau, or amyloid-to-tau index. These results suggest a relationship between lower CSF Aβ42 levels and sooner AP use. This supports prior reports suggesting a correlation between BPSD and Aβ deposition on PET. These results highlight the need for further prospective studies on Aβ levels and BPSD.

Determinants of school dropouts in India: a study through survival analysis approach

Determinants of school dropouts in India: a study through survival analysis approach

COVID-19 Mortality Risk Factors Using Survival Analysis: A Case Study of Jakarta, Indonesia

At the end of 2021 Q2, coronavirus disease 2019 (COVID-19) in Indonesia experienced a continuous increase in positivity and mortality rates. There are limited studies regarding the factors of COVID-19 mortality in Indonesia with a more balanced dataset. The previous studies only implemented logistic regression, sensitive to the imbalanced dataset. Meanwhile, other countries implemented survival analysis to overcome the problem. Most survival analyses using Cox proportional hazard (CPH) model require the variables to be time-independent. To this end, this study aims to identify the risk factors for COVID-19 mortality in Indonesia using a survival analysis approach using Jakarta as a case study. We use the Piecewise Exponential Model (PEM) to overcome the time-dependent problem in CPH. The findings show that the COVID-19 mortality does not differ the gender. In contrast, it differs the elderly with 3.5 times higher to be deceased. Dyspnea, malaise, and pneumonia are the primary symptoms associated with COVID-19 mortality. From the comorbidities, diabetes and chronic disease are related to COVID-19, while hypertension and heart attack are still considerable in clustered contexts. The advanced treatment using intubation and extra corporeal membrane oxygenation (ECMO) produces a relatively large hazard risk of COVID-19 mortality. Based on the findings, we suggest that collaboration among the government, society, and hospitals is vital in overcoming the COVID-19 pandemic and minimizing the COVID-19 death.

The role of anion gap normalization time in the management of pediatric diabetic ketoacidosis.

Our aims were to determine whether anion gap normalization time (AGNT) correlates with risk factors related to the severity of diabetic ketoacidosis (DKA) in children, and to characterize AGNT as a criterion for DKA resolution in children admitted with moderate or severe disease. A ten-year retrospective cohort study of children admitted to the intensive care unit with DKA. We used a survival analysis approach to determine changes in serum glucose, bicarbonate, pH, and anion gap following admission. Using multivariate analysis, we examined associations between patients' demographic and laboratory characteristics with delayed normalization of the anion gap. A total of 95 patients were analyzed. The median AGNT was 8 h. Delayed AGNT (>8 h) correlated with pH < 7.1 and serum glucose >500 mg/dL. In multivariate analysis, glucose >500 mg/dL was associated with an increased risk for delayed AGNT, by 3.41 fold. Each 25 mg/dL elevation in glucose was associated with a 10% increment in risk for delayed AGNT. Median AGNT preceded median PICU discharge by 15 h (8 vs. 23 h). AGNT represents a return to normal glucose-based physiology and an improvement in dehydration. The correlation observed between delayed AGNT and markers of DKA severity supports the usefulness of AGNT for assessing DKA recovery.

Factors Associated With Pregnancy Duration up to 32 Weeks in Patients With Twin Gestations Who Underwent Cerclage Placement [ID: 1378167

INTRODUCTION: Historically, cerclages were not recommended in multiple-gestation pregnancies because of uncertain efficacy and concern for harm. However, a meta-analysis of twin pregnancies found cerclage was associated with reduction in preterm birth (PTB) less than 32 weeks with cervical length (CL) of 15 mm or lower. METHODS: We conducted a retrospective analysis of women who underwent cervical cerclage placement at a single tertiary academic center from 2008 to 2020. Data pertaining to maternal demographics, obstetric course, sonographic CL, and cerclage operative technique were abstracted for women with twin gestations. Factors with P value of .2 or lower on univariate analysis of spontaneous PTB (sPTB) less than 32 weeks were considered for incorporation along with history of sPTB into a Cox proportional-hazards model to assess factors associated with pregnancy duration up to 32 weeks. RESULTS: Seventy-five twin gestations were analyzed. On univariate analysis, neuraxial anesthesia during cerclage placement, preoperative cervical dilation, use of Foley catheter for membrane elevation, shorter postoperative CL, and twin B male sex were found to be significantly associated with sPTB less than 32 weeks. Cox proportional-hazards analysis demonstrated that use of Foley catheter for membrane elevation, neuraxial anesthesia, postoperative CL less than 25 mm recorded 24 weeks or earlier, and history of sPTB greater than 16 weeks had significantly higher adjusted hazard ratios. The bias-corrected bootstrapped (1,000 repetitions) AUROC for a logistic regression model with these four covariates predictive of sPTB less than 32 weeks was 0.83 (95% CI 0.69–0.89). CONCLUSION: A survival analysis approach offers valuable clinical insight into predictors of shorter pregnancy duration up to 32 weeks in twin gestations undergoing cervical cerclage placement.

Plasmodium 18S Ribosomal RNA Biomarker Clearance After Food and Drug Administration-Approved Antimalarial Treatment in Controlled Human Malaria Infection Trials.

Sensitive molecular assays, such as quantitative reverse-transcription polymerase chain reaction (qRT-PCR) of Plasmodium 18S ribosomal RNA (rRNA), are increasingly the primary method of detecting infections in controlled human malaria infection (CHMI) trials. However, thick blood smears (TBSs) remain the main method for confirming clearance of parasites after curative treatment, in part owing to uncertainty regarding biomarker clearance rates. For this analysis, 18S rRNA qRT-PCR data were compiled from 127 Plasmodium falciparum-infected participants treated with chloroquine or atovaquone-proguanil in 6 CHMI studies conducted in Seattle, Washington, over the past decade. A survival analysis approach was used to compare biomarker and TBS clearance times among studies. The effect of the parasite density at which treatment was initiated on clearance time was estimated using linear regression. The median time to biomarker clearance was 3 days (interquartile range, 3-5 days), while the median time to TBS clearance was 1 day (1-2 days). Time to biomarker clearance increased with the parasite density at which treatment was initiated. Parasite density did not have a significant effect on TBS clearance. The Plasmodium 18S rRNA biomarker clears quickly and can be relied on to confirm the adequacy of Food and Drug Administration-approved treatments in CHMI studies at nonendemic sites.

Dual functionality of pyrimidine and flavone in targeting genomic variants of EGFR and ER receptors to influence the differential survival rates in breast cancer patients

Abstract Breast cancer ranks as one of the most prevalent forms of cancer and stands as the primary global cause of mortality among women. Overexpression of EGFR and ER receptors or their genomic alterations leads to malignant transformation, disease aggression and is linked to poor patient survival outcomes. The clinical breast cancer patient’s genomic expression, survival analysis, and computational drug-targeting approaches were used to identify best-hit phytochemicals for therapeutic purposes. Breast cancer patients have genomic alterations in EGFR (4%, n = 5699) and ER (9%, n = 8461), with the highest proportion being missense mutations. No statistically significant difference was observed in the patient survival rates between the altered and unaltered ER groups, unlike EGFR, with the lowest survival rates in the altered group. Computational screening of natural compound libraries (7711) against each EGFR (3POZ) and ER (3ERT) receptor shortlists the best-hit 3 compounds with minimum docking score (ΔG = −7.9 to −10.8), MMGBSA (−40.16 to −51.91 kcal/mol), strong intermolecular H-bonding, drug-like properties with least kd, and ki. MD simulation studies display stable RMSD, RMSF, and good residual correlation of best-hit common compounds (PubChem ID: 5281672 and 5280863) targeting both EGFR and ER receptors. In vitro, studies revealed that these common drugs exhibited a high anti-proliferative effect on MCF-7 and MDA-MB-231 breast cancer cells, with effective IC50 values (15–40 μM) and lower free energy, kd, and ki (5281672 &amp;gt; 5280863 &amp;gt; 5330286) much affecting HEK-293 non-cancerous cells, indicating the safety profile. The experimental and computational correlation studies suggest that the highly expressed EGFR and ER receptors in breast cancer patients having poor survival rates can be effectively targeted with best-hit common potent drugs with a multi-target therapeutic approach. Insight Box: The findings of this study provide valuable insights into the genomic/proteomic data, breast cancer patient’s survival analysis, and EGFR and ER receptor variants structural analysis. The genetic alterations analysis of EGFR and ER/ESR1 in breast cancer patients reveals the high frequency of mutation types, which affect patient’s survival rate and targeted therapies. The common best-hit compounds affect the cell survival patterns with effective IC50, drug-like properties having lower equilibrium and dissociation constants demonstrating the anti-proliferative effects. This work integrates altered receptor structural analysis, molecular interaction-based simulations, and ADMET properties to illuminate the identified best hits phytochemicals potential efficacy targeting both EGFR and ER receptors, demonstrating a multi-target therapeutic approach.

Impact of dialysis reimbursement policies on care outcomes in peritoneal dialysis patients in Taiwan

The number of treated end stage renal disease (ESRD) patients worldwide has rapidly grown. To prolong their lives ESRD patients require transplantation or dialysis treatment. Limited donor availability has caused most of the ESRD patients to rely on either hemodialysis (HD) or peritoneal dialysis (PD). Taiwan had the highest prevalence rate worldwide and sought to increase the PD utilization through a series of reimbursement incentives.This study evaluated the effect of those policy initiatives. A retrospective longitudinal study using a before-and-after analysis was conducted. ESRD patients initiating either PD or HD were identified from the entire population of Taiwan NHI's beneficiaries. PD patients, before and after the PD-encouraging initiatives, were matched through a propensity score technique, and the change in PD technical failure was analyzed. HD patients were also matched as the control group to assess the impact on PD mortality. The competing risk regression approach for survival analysis was adopted.The results indicate the increase in PD utilization during this period was also accompanied by increases in both technique failure and mortality. Since PD shifts more burden of care to patients, efforts to increase its utilization may require an increase in the education of providers and patients to benefit more effectively. It may also require an increase in staff to provide ongoing training and support as the policy unfolds.

Abstract 5696: A prognostic machine learning model for early breast cancer which combines clinical and genetic data in patients treated with neo/adjuvant chemotherapy

Abstract Background: Accurate modelling of the impact of patient-specific features and cancer treatments on survival allows the assignment of targeted therapy. There has not been any effort to build a multi-source model for the survival analysis of breast cancer. We show in this study a prognostic model, which integrates genetic (DNA), clinical and therapy inputs to predict survival for early breast cancer (stages 1-3) for all breast cancer subtypes. Methods: We used a data-driven Random Survival Forest approach, a statistical non-parametric ensemble learning method, that incorporates censor and time-to-event data. The learning is performed by creating numerous decision trees and selecting the model based on the correct responses in unseen data. We used The Cancer Genome Atlas Breast Cancer (TCGA) dataset and observed improvements in the accuracy when more sources of data were used, in line with the previous research. Integrating the impact of non-silent somatic tumor mutations (whole exome) and gene copy number variation (CNV) were analyzed on all mutations and per particular mutation. Results: Data from 1096 women with stage 1-3 early breast cancer were inputs to the model n=437 ER+ve HER2-ve, n=123 HER2+ve ER+ve, n=40 HER2+ ER-ve and n=126 TNBC. Pathological stage 1, n=183; stage 2, n=620; and stage 3, n=249. The following chemotherapy and hormonal treatments were used in the analysis: anthracycline, taxanes, platinum, alkylating and anti-metabolite agents, anti-oestrogen, aromatase inhibitors, ovarian suppression and HER2 antibody treatment. The model accuracy for predicting survival for early breast cancer using only clinical data was 0.78 for Area Under Curve (AUC) and c-index. The predictive accuracy improved stepwise by adding hormone, genetic and treatment data to AUC of 0.86 and c-index to 0.85. We observed the same trend if the proportion of test data increased from 0.25 to 0.75. Changes in median genes FGFR2 and CDKN2A copy number were strongly prognostic with p=0.0001 and p=0.002, and weaker signals for CBFB p=0.05, HRAS p=0.06, AKT p=0.07. Conclusion: Using public datasets, we developed a predictive survival model for an individual with early breast cancer up to 5 years from diagnosis using multi-source and patient-specific data. We show that using this approach for survival analysis yields good accuracy. Citation Format: Aidan (Amanzhol) Kubeyev, Andrea Giorni, Prabu Siva, Luiz Silva, Jordan Laurie, Matthew Foster, Matthew Griffiths, Uzma Asghar. A prognostic machine learning model for early breast cancer which combines clinical and genetic data in patients treated with neo/adjuvant chemotherapy. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5696.

Smart Intrusion Detection System for CAN Network Implemented using LSTM Strategy

In today's cars, communication between electronic control units is managed via the Controller Area Network (CAN) bus system. Nevertheless, the CAN bus system lacks means for authentication and authorisation, making it susceptible to attacks like denial-of-service, fuzzing, and spoofing. To identify and counteract CAN bus network intrusions, this study suggests an intrusion detection system based on a Long Short-Term Memory (LSTM) model. Extraction of attack-free data and attack injection produced the dataset used for testing and training. Findings show that the suggested LSTM model has a greater detection rate than the Survival Analysis approach and is efficient in defending current automobiles from assaults on the CAN bus network. The ability to identify these sorts of assaults is limited by the constraints of conventional intrusion detection systems (IDS) for CAN bus networks, such as statistical analysis, frequency-based analysis, and Hidden Markov Model (HMM). Using a Long Short-Term Memory (LSTM) based strategy, the objective of this article is to build a reliable IDS that can efficiently identify and counteract these assaults on the CAN bus network.

A censored quantile regression approach for relative survival analysis: Relative survival quantile regression.

We propose a censored quantile regression model for the analysis of relative survival data. We create a hybrid data set consisting of the study observations and counterpart randomly sampled pseudopopulation observations imputed from population life tables that adjust for expected mortality. We then fit a censored quantile regression model to the hybrid data incorporating demographic variables (e.g., age, biologic sex, calendar time) corresponding to the population life tables of demographically-similar individuals, a population versus study covariate, and its interactions with the variables of interest. These latter variables can be interpreted as relative survival parameters that depict the differences in failure quantiles between the study participants and their populationcounterparts.

A survival analysis approach to determine factors associated with non-retention of newly hired health workers in Iran

Background and aimOne of the main tasks of the healthcare human resource management is to maintain and retain professional staff. The high level turnover of professional staff may reduce the quality of healthcare service delivery. Therefore, this study investigated the factors associated with the turnover of the newly recruited healthcare professionals using survival analysis method in Iran.Materials and methodsThis historical cohort analysis comprised 6811 employees who began working at Mashhad University of Medical Sciences between 2005 and 2020. Employees recruited at any of the university’s units between the years 2005 to 2019 were included. We used appropriate descriptive indices and Log-rank test and the Cox proportional-hazards model to assess the staff turnover. A significance level of 0.05 was used for all tests.ResultsThe findings of the survival analysis showed that the probability of turnover in one year, two years, and five years of employment were 0.12, 0.16, and 0.27. Based on the findings of the Log-rank test, the probability of turnover in entire of the study period was not statistically different between male and female (p = 0.573), and likewise between employees with healthcare occupations and non-healthcare occupations (p = 0.351). Employees whose current workplace and birthplace were not similar had a significantly higher probability of turnover (p < 0.001). Accordingly, the Cox regression result showed, the risk of turnover for the singles was 1.22 times higher than the married. For the Ph.D degree was 3.23 times higher compared to those with a diploma or an associate degree, and for a bachelor’s or master’s degree was 2.06 times more likely to change their workplace than those with a diploma or an associate degree.ConclusionPolicies promoting the recruitment of native-born professionals, given priority to the married candidates than single ones, and/or recurring candidates to pledge to stay in the locality of recruitment site can increase the staff retention and reduce the costs of staff turnover including re-hiring, initial and on-the-job training, accommodation, and other extra living consumptions away from home and family.