Bias by censoring for competing events in survival analysis

Time-varying effect in older patients with early-stage breast cancer: a model considering the competing risks based on a time scale

Background

Patients with early-stage breast cancer may have a higher risk of dying from other diseases, making a competing risks model more appropriate. Considering subdistribution hazard ratio, which is used often, limited to model assumptions and clinical interpretation, we aimed to quantify the effects of prognostic factors by an absolute indicator, the difference in restricted mean time lost (RMTL), which is more intuitive. Additionally, prognostic factors of breast cancer may have dynamic effects (time-varying effects) in long-term follow-up. However, existing competing risks regression models only provide a static view of covariate effects, leading to a distorted assessment of the prognostic factor.

Methods

To address this issue, we proposed a dynamic effect RMTL regression that can explore the between-group cumulative difference in mean life lost over a period of time and obtain the real-time effect by the speed of accumulation, as well as personalized predictions on a time scale.

Results

A simulation validated the accuracy of the coefficient estimates in the proposed regression. Applying this model to an older early-stage breast cancer cohort, it was found that 1) the protective effects of positive estrogen receptor and chemotherapy decreased over time; 2) the protective effect of breast-conserving surgery increased over time; and 3) the deleterious effects of stage T2, stage N2, and histologic grade II cancer increased over time. Moreover, from the view of prediction, the mean C-index in external validation reached 0.78.

Conclusion

Dynamic effect RMTL regression can analyze both dynamic cumulative effects and real-time effects of covariates, providing a more comprehensive prognosis and better prediction when competing risks exist.

The Clinical Relevance of the Infiltrating Immune Cell Composition in Kidney Transplant Rejection

Key Points

Background

The link between the histology of kidney transplant rejection, especially antibody-mediated rejection, T-cell–mediated rejection, and mixed rejection, and the types of infiltrating immune cells is currently not well charted. Cost and technical complexity of single-cell analysis hinder large-scale studies of the relationship between cell infiltrate profiles and histological heterogeneity.

Methods

In this cross-sectional study, we assessed the composition of nine intragraft immune cell types by using a validated kidney transplant–specific signature matrix for deconvolution of bulk transcriptomics in three different kidney transplant biopsy datasets (N=403, N=224, N=282). The association and discrimination of the immune cell types with the Banff histology and the association with graft failure were assessed individually and with multivariable models. Unsupervised clustering algorithms were applied on the overall immune cell composition and compared with the Banff phenotypes.

Results

Banff-defined rejection was related to high presence of CD8+ effector T cells, natural killer cells, monocytes/macrophages, and, to a lesser extent, B cells, whereas CD4+ memory T cells were lower in rejection compared with no rejection. Estimated intragraft effector memory–expressing CD45RA (TEMRA) CD8+ T cells were strongly and consistently associated with graft failure. The large heterogeneity in immune cell composition across rejection types prevented supervised and unsupervised methods to accurately recover the Banff phenotypes solely on the basis of immune cell estimates. The lack of correlation between immune cell composition and Banff-defined rejection types was validated using multiplex immunohistochemistry.

Conclusions

Although some specific cell types (FCGR3A + myeloid cells, CD14 + monocytes/macrophages, and NK cells) partly discriminated between rejection phenotypes, the overall estimated immune cell composition of kidney transplants was ill-related to main Banff-defined rejection categories and added to the Banff lesion scoring and evaluation of rejection severity. The estimated intragraft CD8temra cells bore strong and consistent association with graft failure and were independent of Banff-grade rejection.

Sensitive HLA antibody testing and the risk of antibody-mediated rejection and graft failure

Solid phase detection and identification of HLA antibodies in kidney transplantation currently relies on single antigen bead (Luminex®) assays, which is more sensitive than the previously used enzyme-linked immunosorbent assays (ELISA). To evaluate the impact of more sensitive HLA testing on antibody-mediated rejection (AMR) occurrence and allograft survival, we analysed 1818 renal allograft recipients transplanted between March 2004 and May 2021. In 2008, solid phase testing switched from ELISA to Luminex. We included 393 (21.6%) transplantations before and 1425 (78.4%) transplantations after transition from ELISA- to Luminex-based testing. For this study, bio-banked ELISA era samples were tested retrospectively with Luminex. Significantly less pretransplant DSA were found in patients transplanted with pre-existing HLA antibodies in the Luminex (109/387) versus the ELISA period (43/90) (28% vs. 48%, p < 0.01). Throughout histological follow-up, 169 of 1818 (9.3%) patients developed AMR. After implementing Luminex-based testing, the rate of AMR significantly decreased (p = 0.003). However, incidence of graft failure did not significantly differ between both eras. In conclusion, less patients with pretransplant DSA were transplanted since the implementation of Luminex HLA testing. Transition from ELISA- to Luminex-based HLA testing was associated with a significant decrease in AMR occurrence post-transplantation. Since the decline of AMR did not translate into improved graft survival, Luminex-based testing has the added value of preventing low-risk AMR cases. Therefore, Luminex' high sensitivity must be balanced against waiting time for a suitable organ.

Association between Preeclampsia and Long-term Risk of Arrhythmias

BACKGROUND

Preeclampsia (PE), a pregnancy-induced hypertensive disorder, affects 4-5% of pregnancies worldwide. It is well known that hypertension is associated with an increased risk of arrhythmias; however, data on the association between PE and arrhythmias are sparse.

METHODS

In this observational cohort study, we identified all primiparous women who gave birth in Denmark (1997-2016) using Danish nationwide registries. The women were stratified on whether they developed PE during primiparous pregnancy, and followed from primiparous pregnancy to incident arrhythmia, emigration, death, or end of study (December 31, 2018).

RESULTS

A total of 523,271 primiparous women with a median age of 28 years were included and 23,367 (4.5%) were diagnosed with PE. During a median follow-up of 10.1 years, women with vs without PE were associated with a higher incidence of arrhythmias (1.42% vs 1.02%): 1) Composite of cardiac arrest, ventricular tachycardia/fibrillation, or ICD implantation (adjusted HR 1.60 [95% CI 1.14-2.24]), 2) Composite of advanced 2nd degree or 3rd degree atrioventricular block, sinoatrial dysfunction, or pacemaker implantation (adjusted HR 1.48 [95% CI 0.97-2.23]), 3) Composite of supraventricular tachyarrhythmias or extra systoles (adjusted HR 1.34 [95% CI 1.19-1.51]), 4) Composite of all the above-mentioned arrhythmias (adjusted HR 1.37 [95% CI 1.23-1.54]).

CONCLUSION

Preeclamptic women were associated with a significantly and at hitherto unknown long-term increased rate of arrhythmias. This finding suggests that women with PE may benefit from cardiovascular risk assessment, screening, and preventive education.

Determinants of tofacitinib discontinuation in adult patients with rheumatoid arthritis during long-term extension studies up to 9.5 years

Objectives

To examine determinants of tofacitinib discontinuation due to voluntary (i.e. patient-driven) or involuntary reasons (i.e. protocol mandated) in long-term extension (LTE) studies of patients with RA to inform clinical practice, clinical study execution and data capture.

Methods

This post hoc analysis used pooled data from patients receiving tofacitinib 5 or 10 mg twice daily (BID) in LTE studies. Outcomes included time to voluntary/involuntary discontinuation (and baseline predictors), including by geographic region. Exposure-adjusted event rates (EAERs) were calculated for adverse events (AEs), serious AEs (SAEs) and discontinuations due to AEs/SAEs.

Results

Of 4967 patients, 2463 (49.6%) discontinued [1552/4967 (31.2%) voluntarily, 911/4967 (18.3%) involuntarily] and 55 (1.1%) died over the course of 9.5 years. When involuntary discontinuation was present as a competing risk for voluntary discontinuation, patients who stayed on combination therapy and with higher patient-assessed pain were significantly more likely to discontinue for voluntary reasons (P < 0.05). Older patients, those enrolled in Asia, Europe or Latin America (vs USA or Canada) or with RF+/anti-CCP+ status were significantly less likely to discontinue for voluntary reasons (P < 0.05). Small numeric differences in disease activity were observed between geographic regions in patients who discontinued or completed the studies. EAERs were generally higher for tofacitinib 10 vs 5 mg BID, irrespective of discontinuation reason.

Conclusion

The factors associated with voluntary/involuntary discontinuation of tofacitinib suggest that treatment persistence in RA studies is partly predictable, which may be reflected in clinical practice. Applying these results may improve our understanding of attrition and inform future study design/execution.

Trial registrations

ClinicalTrials.gov (http://clinicaltrials.gov): NCT00413699 and NCT00661661.

Predicting the risks of kidney failure and death in adults with moderate to severe chronic kidney disease: multinational, longitudinal, population based, cohort study

OBJECTIVE

To train and test a super learner strategy for risk prediction of kidney failure and mortality in people with incident moderate to severe chronic kidney disease (stage G3b to G4).

DESIGN

Multinational, longitudinal, population based, cohort study.

SETTINGS

Linked population health data from Canada (training and temporal testing), and Denmark and Scotland (geographical testing).

PARTICIPANTS

People with newly recorded chronic kidney disease at stage G3b-G4, estimated glomerular filtration rate (eGFR) 15-44 mL/min/1.73 m2.

MODELLING

The super learner algorithm selected the best performing regression models or machine learning algorithms (learners) based on their ability to predict kidney failure and mortality with minimised cross-validated prediction error (Brier score, the lower the better). Prespecified learners included age, sex, eGFR, albuminuria, with or without diabetes, and cardiovascular disease. The index of prediction accuracy, a measure of calibration and discrimination calculated from the Brier score (the higher the better) was used to compare KDpredict with the benchmark, kidney failure risk equation, which does not account for the competing risk of death, and to evaluate the performance of KDpredict mortality models.

RESULTS

67 942 Canadians, 17 528 Danish, and 7740 Scottish residents with chronic kidney disease at stage G3b to G4 were included (median age 77-80 years; median eGFR 39 mL/min/1.73 m2). Median follow-up times were five to six years in all cohorts. Rates were 0.8-1.1 per 100 person years for kidney failure and 10-12 per 100 person years for death. KDpredict was more accurate than kidney failure risk equation in prediction of kidney failure risk: five year index of prediction accuracy 27.8% (95% confidence interval 25.2% to 30.6%) versus 18.1% (15.7% to 20.4%) in Denmark and 30.5% (27.8% to 33.5%) versus 14.2% (12.0% to 16.5%) in Scotland. Predictions from kidney failure risk equation and KDpredict differed substantially, potentially leading to diverging treatment decisions. An 80-year-old man with an eGFR of 30 mL/min/1.73 m2 and an albumin-to-creatinine ratio of 100 mg/g (11 mg/mmol) would receive a five year kidney failure risk prediction of 10% from kidney failure risk equation (above the current nephrology referral threshold of 5%). The same man would receive five year risk predictions of 2% for kidney failure and 57% for mortality from KDpredict. Individual risk predictions from KDpredict with four or six variables were accurate for both outcomes. The KDpredict models retrained using older data provided accurate predictions when tested in temporally distinct, more recent data.

CONCLUSIONS

KDpredict could be incorporated into electronic medical records or accessed online to accurately predict the risks of kidney failure and death in people with moderate to severe CKD. The KDpredict learning strategy is designed to be adapted to local needs and regularly revised over time to account for changes in the underlying health system and care processes.

Clostridioides difficile recurrence in individuals with and without cancer: a Swedish population-based cohort study

PURPOSE

Patients with cancer are vulnerable to Clostridioides difficile infection (CDI) due to their disease, treatment and regular hospital contact, yet if CDI-recurrence is more common remains unclear, and differences among cancer types remain unexplored.

METHODS

This Swedish nationwide population-based cohort included all 43,150 individuals with recorded CDI (2006-2019) to assess CDI-recurrence in individuals with and without cancer, with binary multivariable logistic regression, stratified by anatomical location, and survival status.

RESULTS

Compared to those without cancer (N = 29,543), ongoing cancer (diagnosis < 12 months; N = 3,882) was associated with reduced recurrence (OR = 0.81, 95% CI 0.73-0.89), while there was no association with cancer history (diagnosis ≥ 12 months; N = 9,725). There was an increased 8-week all-cause mortality (Ongoing cancer: OR = 1.58, 95% CI 1.43-1.74; Cancer history: OR = 1.45, 95% CI 1.36-1.55) compared to those without cancer. Among CDI-survivors, those with ongoing cancer presented with a decreased odds of recurrence (OR = 0.84, 95% CI 0.76-0.94), compared to those without cancer history, with no association for those with cancer history (OR = 1.04, 95% CI 0.97-1.1). Large variations were seen across cancer types, with the highest observed proportion of recurrence in oral and mesothelial cancer, and the lowest for esophageal cancer, although no statistically significant OR were found.

CONCLUSION

The population-based study indicates that individuals with cancer may have fewerrecurrences than expected, yet variations by cancer type were large, and mortality was high.

Methodological biases in observational hospital studies of COVID-19 treatment effectiveness: pitfalls and potential

Introduction

This study aims to discuss and assess the impact of three prevalent methodological biases: competing risks, immortal-time bias, and confounding bias in real-world observational studies evaluating treatment effectiveness. We use a demonstrative observational data example of COVID-19 patients to assess the impact of these biases and propose potential solutions.

Methods

We describe competing risks, immortal-time bias, and time-fixed confounding bias by evaluating treatment effectiveness in hospitalized patients with COVID-19. For our demonstrative analysis, we use observational data from the registry of patients with COVID-19 who were admitted to the Bellvitge University Hospital in Spain from March 2020 to February 2021 and met our predefined inclusion criteria. We compare estimates of a single-dose, time-dependent treatment with the standard of care. We analyze the treatment effectiveness using common statistical approaches, either by ignoring or only partially accounting for the methodological biases. To address these challenges, we emulate a target trial through the clone-censor-weight approach.

Results

Overlooking competing risk bias and employing the naïve Kaplan-Meier estimator led to increased in-hospital death probabilities in patients with COVID-19. Specifically, in the treatment effectiveness analysis, the Kaplan-Meier estimator resulted in an in-hospital mortality of 45.6% for treated patients and 59.0% for untreated patients. In contrast, employing an emulated trial framework with the weighted Aalen-Johansen estimator, we observed that in-hospital death probabilities were reduced to 27.9% in the "X"-treated arm and 40.1% in the non-"X"-treated arm. Immortal-time bias led to an underestimated hazard ratio of treatment.

Conclusion

Overlooking competing risks, immortal-time bias, and confounding bias leads to shifted estimates of treatment effects. Applying the naïve Kaplan-Meier method resulted in the most biased results and overestimated probabilities for the primary outcome in analyses of hospital data from COVID-19 patients. This overestimation could mislead clinical decision-making. Both immortal-time bias and confounding bias must be addressed in assessments of treatment effectiveness. The trial emulation framework offers a potential solution to address all three methodological biases.

Exploring the relation of active surveillance schedules and prostate cancer mortality

BACKGROUND

Active surveillance (AS), where treatment is deferred until cancer progression is detected by a biopsy, is acknowledged as a way to reduce overtreatment in prostate cancer. However, a consensus on the frequency of taking biopsies while in AS is lacking. In former studies to optimize biopsy schedules, the delay in progression detection was taken as an evaluation indicator and believed to be associated with the long-term outcome, prostate cancer mortality. Nevertheless, this relation was never investigated in empirical data. Here, we use simulated data from a microsimulation model to fill this knowledge gap.

METHODS

In this study, the established MIcrosimulation SCreening Analysis model was extended with functionality to simulate the AS procedures. The biopsy sensitivity in the model was calibrated on the Canary Prostate Cancer Active Surveillance Study (PASS) data, and four (tri-yearly, bi-yearly, PASS, and yearly) AS programs were simulated. The relation between detection delay and prostate cancer mortality was investigated by Cox models.

RESULTS

The biopsy sensitivity of progression detection was found to be 50%. The Cox models show a positive relation between a longer detection delay and a higher risk of prostate cancer death. A 2-year delay resulted in a prostate cancer death risk of 2.46%-2.69% 5 years after progression detection and a 10-year risk of 5.75%-5.91%. A 4-year delay led to an approximately 8% greater 5-year risk and an approximately 25% greater 10-year risk.

CONCLUSION

The detection delay is confirmed as a surrogate for prostate cancer mortality. A cut-off for a "safe" detection delay could not be identified.

Risk of congestive heart failure and mortality following lymphovenous anastomosis: a nationwide population-based retrospective cohort study

BACKGROUND

Lymphovenous anastomosis (LVA) enables lymphatic fluid to drain into the venous system. However, no study has investigated the association between LVA and heart failure (HF) caused by fluid overload in the blood circulating system. The purpose of our study was to determine whether LVA increases the risk of HF and mortality.

MATERIAL AND METHODS

This nationwide retrospective study evaluated a total of 1400 lymphedema patients who underwent LVA and two control cohorts with 28 000 lymphedema who did not undergo LVA and 70 000 age-matched and sex-matched participants from the Korean National Health Insurance database were included. Blood pressure, body mass index (BMI), glucose and cholesterol levels, smoking history, and comorbidities were obtained during National Health Insurance Service - Health Screening (NHIS-HealS). The incidence, adjusted risk for HF, and mortality were evaluated.

RESULTS

Adjusted HRs for HF were 1.20 (confidence interval [CI], 1.03-1.40) and 1.30 (CI, 1.12-1.50), referenced by the general population control cohort and patients with lymphedema without LVA, respectively. In age, sex, BMI, and smoking status-stratified analyses, heightened risk of HF was evident across all sexes, spanning both young and old age groups, encompassing individuals with various smoking statuses, and those with a BMI of 18.5 or higher. Among these groups, the risk was notably greater in males compared to females, higher in younger individuals as opposed to older ones, and further elevated within the BMI range of 18.5-25.

CONCLUSIONS

LVA is associated with an increased HF risk, independent of cardiovascular risk factors and associated comorbidities. This association is prominent in participants aged <50 years, in males, and in the normal-to-obese (BMI ≥18.5 kg/m 2 ) group. Among patients with lymphedema, LVA did not significantly affect mortality.

Advanced considerations in survival analysis

Investigation of survival during the follow-up period is common in cardiovascular research and has intrinsic issues that require precise knowledge, such as survival or censoring. Besides, as the follow-up period lengthens and events other than mortality are studied, the analysis becomes more complex, so Kaplan-Meier analyses or Cox models are not always sufficient. In this primer, we provide the reader with detailed information on the interpretation of the most common survival analyses and delve into methods to analyse competing risks or alternatives to the conventional methods when the proportional hazards assumption is not met.

Incidence and risk factors of hypothyroidism after treatment for early breast cancer: a population-based cohort study

PURPOSE

An increased incidence of hypothyroidism among breast cancer survivors has been observed in earlier studies. The impact of the postoperative treatment modalities and their potential interplay on hypothyroidism development needs to be studied.

METHODS

We conducted a population- and registry-based study using the Breast Cancer Data Base Sweden (BCBaSe) including females diagnosed with breast cancer between 2006 and 2012. In total, 21,268 female patients diagnosed with early breast cancer between 2006 and 2012, with no previous prescription of thyroid hormones and no malignant diagnosis during the last ten years before breast cancer diagnosis, were included in the final analysis.

RESULTS

During the follow-up (median follow-up time 7.9 years), 1212 patients (5.7%) developed hypothyroidism at a median time of 3.45 years from the index date. No association of the systemic oncological treatment in terms of either chemotherapy or endocrine therapy and hypothyroidism development could be identified. A higher risk (HR 1.68;95% CI 1.42-1.99) of hypothyroidism identified among patients treated with radiation treatment of the regional lymph nodes whereas no increased risk in patients treated only with radiation therapy to the breast/chest wall was found (HR 1.01; 95% CI 0.86-1.19). The risk of hypothyroidism in the cohort treated with radiotherapy of the regional lymph nodes was present irrespective of the use of adjuvant chemotherapy treatment.

CONCLUSIONS

Based on the results of our study, the implementation of hypothyroidism surveillance among the breast cancer survivors treated with radiotherapy of the regional lymph nodes can be considered as reasonable in the follow-up program.

An external validation of the Kidney Donor Risk Index in the UK transplant population in the presence of semi-competing events

BACKGROUND

Transplantation represents the optimal treatment for many patients with end-stage kidney disease. When a donor kidney is available to a waitlisted patient, clinicians responsible for the care of the potential recipient must make the decision to accept or decline the offer based upon complex and variable information about the donor, the recipient and the transplant process. A clinical prediction model may be able to support clinicians in their decision-making. The Kidney Donor Risk Index (KDRI) was developed in the United States to predict graft failure following kidney transplantation. The survival process following transplantation consists of semi-competing events where death precludes graft failure, but not vice-versa.

METHODS

We externally validated the KDRI in the UK kidney transplant population and assessed whether validation under a semi-competing risks framework impacted predictive performance. Additionally, we explored whether the KDRI requires updating. We included 20,035 adult recipients of first, deceased donor, single, kidney-only transplants between January 1, 2004, and December 31, 2018, collected by the UK Transplant Registry and held by NHS Blood and Transplant. The outcomes of interest were 1- and 5-year graft failure following transplantation. In light of the semi-competing events, recipient death was handled in two ways: censoring patients at the time of death and modelling death as a competing event. Cox proportional hazard models were used to validate the KDRI when censoring graft failure by death, and cause-specific Cox models were used to account for death as a competing event.

RESULTS

The KDRI underestimated event probabilities for those at higher risk of graft failure. For 5-year graft failure, discrimination was poorer in the semi-competing risks model (0.625, 95% CI 0.611 to 0.640;0.611, 95% CI 0.597 to 0.625), but predictions were more accurate (Brier score 0.117, 95% CI 0.112 to 0.121; 0.114, 95% CI 0.109 to 0.118). Calibration plots were similar regardless of whether the death was modelled as a competing event or not. Updating the KDRI worsened calibration, but marginally improved discrimination.

CONCLUSIONS

Predictive performance for 1-year graft failure was similar between death-censored and competing event graft failure, but differences appeared when predicting 5-year graft failure. The updated index did not have superior performance and we conclude that updating the KDRI in the present form is not required.

Look left, look right: censoring in clinical research

In this methodology editorial, the problem of censored survival (or time-to-event) data in clinical research is discussed. In case of censored observations, part of the observation time for some participants in a study is unobserved, which means that measures such as an average survival time cannot be calculated. Instead, censored survival time should explicitly be accounted for in the analysis of a study, for example using the Kaplan-Meier method. Implications of censoring and how to deal with it in the analysis of a study or when reading a research paper are discussed.

Preeclampsia and Long-Term Risk of Venous Thromboembolism

Importance

As venous thromboembolism (VTE) remains one of the leading causes of maternal mortality, identifying women at increased risk of VTE is of great importance. Preeclampsia is a pregnancy-induced hypertensive disorder with generalized endothelial dysfunction. Some studies suggest that preeclampsia is associated with an increased risk of VTE, but much controversy exists.

Objective

To examine the association between preeclampsia and the risk of VTE during pregnancy, during the puerperium, and after the puerperium.

Design, Setting, and Participants

This observational cohort study used Danish nationwide registries to identify all eligible primiparous women who gave birth in Denmark from January 1, 1997, to December 31, 2016. The women were followed up from primiparous pregnancy to incident VTE, emigration, death, or the end of the study (December 31, 2016). Statistical analyses were carried out from January to May 2023.

Exposure

Preeclampsia during primiparous pregnancy.

Main Outcomes and Measure

The main outcome was incident VTE, and the secondary outcome was all-cause mortality.

Results

A total of 522 545 primiparous women (median age, 28 years [IQR, 25-31 years]) were included, and 23 330 (4.5%) received a diagnosis of preeclampsia. Women with preeclampsia were of similar age to women without preeclampsia but had a higher burden of comorbidities. During a median follow-up of 10.2 years (IQR, 5.2-15.4 years), preeclampsia was associated with a higher incidence of VTE compared with no preeclampsia (incidence rate, 448.8 [95% CI, 399.9-503.5] vs 309.6 [95% CI, 300.6-319.9] per 1000 patient-years, corresponding to an unadjusted hazard ratio [HR] of 1.45 [95% CI, 1.29-1.63] and an adjusted HR of 1.43 [95% CI, 1.27-1.61]). When stratified according to the subcategories of VTE, preeclampsia was associated with an increased rate of deep vein thrombosis (unadjusted HR, 1.51 [95% CI, 1.32-1.72] and adjusted HR, 1.49 [95% CI, 1.31-1.70]) as well as pulmonary embolism (unadjusted HR, 1.39 [95% CI, 1.09-1.76]; adjusted HR, 1.36 [95% CI, 1.08-1.73]). These findings held true in landmark analyses during pregnancy, during the puerperium, and after the puerperium.

Conclusions and Relevance

This cohort study suggests that preeclampsia was associated with a significantly increased risk of VTE during pregnancy, during the puerperium, and after the puerperium, even after thorough adjustment. Future studies should address how to improve the clinical management of women with a history of preeclampsia to prevent VTE.

Implementing competing risks in discrete event simulation: the event-specific probabilities and distributions approach

Background: Although several strategies for modelling competing events in discrete event simulation (DES) exist, a methodological gap for the event-specific probabilities and distributions (ESPD) approach when dealing with censored data remains. This study defines and illustrates the ESPD strategy for censored data. Methods: The ESPD approach assumes that events are generated through a two-step process. First, the type of event is selected according to some (unknown) mixture proportions. Next, the times of occurrence of the events are sampled from a corresponding survival distribution. Both of these steps can be modelled based on covariates. Performance was evaluated through a simulation study, considering sample size and levels of censoring. Additionally, an oncology-related case study was conducted to assess the ability to produce realistic results, and to demonstrate its implementation using both frequentist and Bayesian frameworks in R. Results: The simulation study showed good performance of the ESPD approach, with accuracy decreasing as sample sizes decreased and censoring levels increased. The average relative absolute error of the event probability (95%-confidence interval) ranged from 0.04 (0.00; 0.10) to 0.23 (0.01; 0.66) for 60% censoring and sample size 50, showing that increased censoring and decreased sample size resulted in lower accuracy. The approach yielded realistic results in the case study. Discussion: The ESPD approach can be used to model competing events in DES based on censored data. Further research is warranted to compare the approach to other modelling approaches for DES, and to evaluate its usefulness in estimating cumulative event incidences in a broader context.

Target trial emulation with multi-state model analysis to assess treatment effectiveness using clinical COVID-19 data

BACKGROUND

Real-world observational data are an important source of evidence on the treatment effectiveness for patients hospitalized with coronavirus disease 2019 (COVID-19). However, observational studies evaluating treatment effectiveness based on longitudinal data are often prone to methodological biases such as immortal time bias, confounding bias, and competing risks.

METHODS

For exemplary target trial emulation, we used a cohort of patients hospitalized with COVID-19 (n = 501) in a single centre. We described the methodology for evaluating the effectiveness of a single-dose treatment, emulated a trial using real-world data, and drafted a hypothetical study protocol describing the main components. To avoid immortal time and time-fixed confounding biases, we applied the clone-censor-weight technique. We set a 5-day grace period as a period of time when treatment could be initiated. We used the inverse probability of censoring weights to account for the selection bias introduced by artificial censoring. To estimate the treatment effects, we took the multi-state model approach. We considered a multi-state model with five states. The primary endpoint was defined as clinical severity status, assessed by a 5-point ordinal scale on day 30. Differences between the treatment group and standard of care treatment group were calculated using a proportional odds model and shown as odds ratios. Additionally, the weighted cause-specific hazards and transition probabilities for each treatment arm were presented.

RESULTS

Our study demonstrates that trial emulation with a multi-state model analysis is a suitable approach to address observational data limitations, evaluate treatment effects on clinically heterogeneous in-hospital death and discharge alive endpoints, and consider the intermediate state of admission to ICU. The multi-state model analysis allows us to summarize results using stacked probability plots that make it easier to interpret results.

CONCLUSIONS

Extending the emulated target trial approach to multi-state model analysis complements treatment effectiveness analysis by gaining information on competing events. Combining two methodologies offers an option to address immortal time bias, confounding bias, and competing risk events. This methodological approach can provide additional insight for decision-making, particularly when data from randomized controlled trials (RCTs) are unavailable.

Long and Short-Term Effects of Hypothermic Machine Perfusion vs. Cold Storage on Transplanted Kidneys from Expanded Criteria Donors-A Matched Comparison Study

Hypothermic machine perfusion (HMP) has been shown to reduce delayed graft function (DGF)-rates in kidneys from expanded criteria donors (ECD) and may increase graft survival compared with static cold storage (SCS). This single-center, retrospective observational study aimed to evaluate this effect. The primary endpoint was the DGF-rate, defined as the use of dialysis in the first postoperative week, excluding the first 24 h. The main secondary endpoint was graft survival at 5 years. Recipients of ECD-kidneys between 2013 and 2021 with ≤2 grafts were included (n = 438). The SCS-kidneys were marginal-matched by propensity score to the HMP-group for donor age, cold ischemia time, and graft number. Multivariable adjusted analysis for confounders in the unmatched cohort and caliper-based ID-matching constituted sensitivity analyses. HMP showed a trend to lower DGF-rate in the marginal-matched comparison (9.2% vs. 16.1%, p = 0.063). This was strengthened by a significant benefit observed for HMP in both the sensitivity analyses: an adjusted OR of 0.45 (95% CI: 0.24; 0.84; p = 0.012) in the multivariable analysis and DGF-rate of 8.7% vs. 17.4% (p = 0.024) after ID-matching. The 5-year graft survival rate was >90% in both groups, with no benefit using HMP (HR = 0.79; 95% CI:0.39-1.16; p = 0.52). Our results suggest that HMP may be effective in decreasing DGF-rates, however, without any significant benefit in graft survival.

Latent class analysis-derived classification improves the cancer-specific death stratification of molecular subtyping in colorectal cancer

The molecular subtypes of colorectal cancer (CRC) represent a comprehensive dissection of CRC heterogeneity. However, molecular feature-based classification systems have limitations in accurately prognosticating stratification due to the inability to distinguish cancer-specific deaths. This study aims to establish a classification system that bridges clinical characteristics, cause-specific deaths, and molecular features. We adopted latent class analysis (LCA) on 491,107 first primary CRC patients from the Surveillance, Epidemiology, and End Results (SEER) database to reveal hidden profiles of CRC. The LCA-derived classification scheme was further applied to The Cancer Genome Atlas (TCGA) to assess its effectiveness in improving the accurate stratification of molecular-based subtypes of CRC. Four classes were identified based on latent class analysis integrating demographic and clinicopathological information of CRC patients. The LCA-derived Class 1 (LCAC1) and the LCAC2 showed a high risk of dying from non-CRC, while patients in LCAC3 had a risk of dying from CRC 1.41 times that of LCAC1 (95% confidence interval [CI] = 1.39-1.43). LCAC4 had the lowest probability to die from non-CRC (hazard ratio [HR] = 0.22, 95% CI = 0.21-0.24) compared with LCAC1. Since the LCA-derived classification can identify patients susceptible to CRC-specific death, adjusting for this classification allows molecular-based subtypes to achieve more accurate survival stratification. We provided a classification system capable of distinguish CRC-specific death, which will improve the accuracy of consensus molecular subtypes for CRC patients' survival stratification. Further studies are warranted to confirm the molecular features of LCA-derived classification to inform potential therapeutic strategies and treatment recommendations.

Target Trial Emulation Using Hospital-Based Observational Data: Demonstration and Application in COVID-19

Methodological biases are common in observational studies evaluating treatment effectiveness. The objective of this study is to emulate a target trial in a competing risks setting using hospital-based observational data. We extend established methodology accounting for immortal time bias and time-fixed confounding biases to a setting where no survival information beyond hospital discharge is available: a condition common to coronavirus disease 2019 (COVID-19) research data. This exemplary study includes a cohort of 618 hospitalized patients with COVID-19. We describe methodological opportunities and challenges that cannot be overcome applying traditional statistical methods. We demonstrate the practical implementation of this trial emulation approach via clone-censor-weight techniques. We undertake a competing risk analysis, reporting the cause-specific cumulative hazards and cumulative incidence probabilities. Our analysis demonstrates that a target trial emulation framework can be extended to account for competing risks in COVID-19 hospital studies. In our analysis, we avoid immortal time bias, time-fixed confounding bias, and competing risks bias simultaneously. Choosing the length of the grace period is justified from a clinical perspective and has an important advantage in ensuring reliable results. This extended trial emulation with the competing risk analysis enables an unbiased estimation of treatment effects, along with the ability to interpret the effectiveness of treatment on all clinically important outcomes.