On variance estimation of the inverse probability-of-treatment weighting estimator: A tutorial for different types of propensity score weights.

Propensity score methods, such as inverse probability-of-treatment weighting (IPTW), have been increasingly used for covariate balancing in both observational studies and randomized trials, allowing the control of both systematic and chance imbalances. Approaches using IPTW are based on two steps: (i) estimation of the individual propensity scores (PS), and (ii) estimation of the treatment effect by applying PS weights. Thus, a variance estimator that accounts for both steps is crucial for correct inference. Using a variance estimator which ignores the first step leads to overestimated variance when the estimand is the average treatment effect (ATE), and to under or overestimated estimates when targeting the average treatment effect on the treated (ATT). In this article, we emphasize the importance of using an IPTW variance estimator that correctly considers the uncertainty in PS estimation. We present a comprehensive tutorial to obtain unbiased variance estimates, by proposing and applying a unifying formula for different types of PS weights (ATE, ATT, matching and overlap weights). This can be derived either via the linearization approach or M-estimation. Extensive R code is provided along with the corresponding large-sample theory. We perform simulation studies to illustrate the behavior of the estimators under different treatment and outcome prevalences and demonstrate appropriate behavior of the analytical variance estimator. We also use a reproducible analysis of observational lung cancer data as an illustrative example, estimating the effect of receiving a PET-CT scan on the receipt of surgery.

Inequalities in treatment among patients with colon and rectal cancer: a multistate survival model using data from England national cancer registry 2012-2016.

BACKGROUND: Individual and tumour factors only explain part of observed inequalities in colorectal cancer survival in England. This study aims to investigate inequalities in treatment in patients with colorectal cancer. METHODS: All patients diagnosed with colorectal cancer in England between 2012 and 2016 were followed up from the date of diagnosis (state 1), to treatment (state 2), death (state 3) or censored at 1 year after the diagnosis. A multistate approach with flexible parametric model was used to investigate the effect of income deprivation on the probability of remaining alive and treated in colorectal cancer. RESULTS: Compared to the least deprived quintile, the most deprived with stage I-IV colorectal cancer had a lower probability of being alive and treated at all the time during follow-up, and a higher probability of being untreated and of dying. The probability differences (most vs. least deprived) of being alive and treated at 6 months ranged between -2.4% (95% CI: -4.3, -1.1) and -7.4% (-9.4, -5.3) for colon; between -2.0% (-3.5, -0.4) and -6.2% (-8.9, -3.5) for rectal cancer. CONCLUSION: Persistent inequalities in treatment were observed in patients with colorectal cancer at every stage, due to delayed access to treatment and premature death.

Challenges for quality and utilization of real-world data for diffuse large B-cell lymphoma in REALYSA, a LYSA cohort.

ABSTRACT: Real-world data (RWD) are essential to complement clinical trial (CT) data, but major challenges remain, such as data quality. REal world dAta in LYmphoma and Survival in Adults (REALYSA) is a prospective noninterventional multicentric cohort started in 2018 that included patients newly diagnosed with lymphoma in France. Herein is a proof-of-concept analysis on patients with first-line diffuse large B-cell lymphoma (DLBCL) to (1) evaluate the capacity of the cohort to provide robust data through a multistep validation process; (2) assess the consistency of the results; and (3) conduct an exploratory transportability assessment of 2 recent phase 3 CTs (POLARIX and SENIOR). The analysis population comprised 645 patients with DLBCL included before 31 March 2021 who received immunochemotherapy and for whom 3589 queries were generated, resulting in high data completeness (<4% missing data). Median age was 66 years, with mostly advanced-stage disease and high international prognostic index (IPI) score. Treatments were mostly rituximab, cyclophosphamide, doxorubicin hydrochloride, vincristine, and prednisone (R-CHOP 75%) and reduced dose R-CHOP (13%). Estimated 1-year event-free survival (EFS) and overall survival rates were 77.9% and 90.0%, respectively (median follow-up, 9.9 months). Regarding transportability, when applying the CT's main inclusion criteria (age, performance status, and IPI), outcomes seemed comparable between patients in REALYSA and standard arms of POLARIX (1-year progression-free survival 79.8% vs 79.8%) and SENIOR (1-year EFS, 64.5% vs 60.0%). With its rigorous data validation process, REALYSA provides high-quality RWD, thus constituting a platform for numerous scientific purposes. The REALYSA study was registered at www.clinicaltrials.gov as #NCT03869619.

Functional forms of socio-territorial inequities in breast cancer screening - A French cross-sectional study using hierarchical generalised additive models.

To reduce the breast cancer burden, the French National Organised Breast Cancer Screening Programme (FNOBCSP) was implemented in 2004. The recommended participation rate has never been achieved and socio-territorial inequities in participation have been reported on several occasions. We investigated the functional forms and consistency of the relationships between neighbourhood deprivation, travel time to the nearest accredited radiology centre and screening uptake. We used two-level hierarchical generalised additive models in 8 types of territories classified by socio-demographic and economic factors. The first level was 368,201 women aged 50-72 invited to the 2013-2014 screening campaign in metropolitan France. They were nested in 41 départements, the level of organisation of the FNOBCSP. The effect of travel time showed two main patterns: it was either linear (with participation decreasing as travel time increased) or participation first increased with increasing travel time to a peak around 5-15 min and decreased afterward. In nearly all types and départements, the probability of participation decreased linearly with increasing deprivation. Territorial inequities in participation were more context-dependent and complex than social inequities. Inequities in participation represent a loss of opportunity for individuals who already have the worst cancer outcomes. Evidence-based public health policies are needed to increase the effectiveness and equity of breast cancer screening.

Is the Social Gradient in Net Survival Observed in France the Result of Inequalities in Cancer-Specific Mortality or Inequalities in General Mortality?

BACKGROUND: In cancer net survival analyses, if life tables (LT) are not stratified based on socio-demographic characteristics, then the social gradient in mortality in the general population is ignored. Consequently, the social gradient estimated on cancer-related excess mortality might be inaccurate. We aimed to evaluate whether the social gradient in cancer net survival observed in France could be attributable to inaccurate LT. METHODS: Deprivation-specific LT were simulated, applying the social gradient in the background mortality due to external sources to the original French LT. Cancer registries' data from a previous French study were re-analyzed using the simulated LT. Deprivation was assessed according to the European Deprivation Index (EDI). Net survival was estimated by the Pohar-Perme method and flexible excess mortality hazard models by using multidimensional penalized splines. RESULTS: A reduction in net survival among patients living in the most-deprived areas was attenuated with simulated LT, but trends in the social gradient remained, except for prostate cancer, for which the social gradient reversed. Flexible modelling additionally showed a loss of effect of EDI upon the excess mortality hazard of esophagus, bladder and kidney cancers in men and bladder cancer in women using simulated LT. CONCLUSIONS: For most cancers the results were similar using simulated LT. However, inconsistent results, particularly for prostate cancer, highlight the need for deprivation-specific LT in order to produce accurate results.

Long-term excess mortality and net survival among elderly diffuse large B-cell lymphoma patients after front-line R-CHOP treatment.

In the era of immunochemotherapy, data on the long-term prognosis of elderly patients diagnosed with a diffuse large B-cell lymphoma (DLBCL) are scarce. In this population and on the longer term, other-cause mortality is an important competing risk that needs to be accounted for. Using clinical trial data and relative survival approaches, we estimated the 10-year net survival (NS) and we described the excess mortality hazard (EMH) due (directly or indirectly) to the DLBCL, over time and according to main prognosis factors using flexible regression modelling. The 10-year NS was 65% [59; 71]. Using the flexible modelling, we showed that the EMH decreases steeply after diagnosis. The variables 'performance status', 'number of extra-nodal sites' and the serum 'lactate dehydrogenase' were strongly associated with the EMH, even after adjustment on other important variables. EMH is very close to zero at 10 years for the whole population, so DLBCL patients do not experience an increased mortality compared to the general population in the long term. The number of extra-nodal sites was an important prognostic factor shortly after diagnosis, suggesting that it is correlated with an important but unmeasured prognostic factor that would lead to this selection effect over time.

Individual frailty excess hazard models in cancer epidemiology.

Unobserved individual heterogeneity is a common challenge in population cancer survival studies. This heterogeneity is usually associated with the combination of model misspecification and the failure to record truly relevant variables. We investigate the effects of unobserved individual heterogeneity in the context of excess hazard models, one of the main tools in cancer epidemiology. We propose an individual excess hazard frailty model to account for individual heterogeneity. This represents an extension of frailty modeling to the relative survival framework. In order to facilitate the inference on the parameters of the proposed model, we select frailty distributions which produce closed-form expressions of the marginal hazard and survival functions. The resulting model allows for an intuitive interpretation, in which the frailties induce a selection of the healthier individuals among survivors. We model the excess hazard using a flexible parametric model with a general hazard structure which facilitates the inclusion of time-dependent effects. We illustrate the performance of the proposed methodology through a simulation study. We present a real-data example using data from lung cancer patients diagnosed in England, and discuss the impact of not accounting for unobserved heterogeneity on the estimation of net survival. The methodology is implemented in the R package IFNS.

Are deprivation-specific cancer survival patterns similar according to individual-based and area-based measures? A cohort study of patients diagnosed with five malignancies in England and Wales, 2008-2016.

OBJECTIVE: To investigate if measured inequalities in cancer survival differ when using individual-based ('person') compared with area-based ('place') measures of deprivation for three socioeconomic dimensions: income, deprivation and occupation. DESIGN: Cohort study. SETTING: Data from the Office for National Statistics Longitudinal Study of England and Wales, UK, linked to the National Cancer Registration Database. PARTICIPANTS: Patients diagnosed with cancers of the colorectum, breast, prostate, bladder or with non-Hodgkin's lymphoma during the period 2008-2016. PRIMARY AND SECONDARY OUTCOME MEASURES: Differentials in net survival between groups defined by individual wage, occupation and education compared with those obtained from corresponding area-level metrics using the English and Welsh Indices of Multiple Deprivation. RESULTS: Survival was negatively associated with area-based deprivation irrespective of the type analysed, although a trend from least to most deprived was not always observed. Socioeconomic differences were present according to individually-measured socioeconomic groups although there was an absence of a consistent 'gradient' in survival. The magnitude of differentials was similar for area-based and individually-derived measures of deprivation, which was unexpected. CONCLUSION: These unique data suggest that the socioeconomic influence of 'person' is different to that of 'place' with respect to cancer outcomes. This has implications for health policy aimed at reducing inequalities. Further research could consider the separate and additional influence of area-based deprivation over individual-level characteristics (contextual effects) as well as investigate the geographic, socioeconomic and healthcare-related characteristics of areas with poor outcomes in order to inform policy intervention.

Outcomes of older patients with diffuse large B-cell lymphoma treated with R-CHOP: 10-year follow-up of the LNH03-6B trial.

The LNH03-6B trial was a phase 3 randomized trial evaluating the efficacy of first-line rituximab, cyclophosphamide, doxorubicine, vincristine and prednisone (R-CHOP) delivered every 2 weeks (R-CHOP14) or 3 weeks (R-CHOP21) in patients with diffuse large B-cell lymphoma (DLBCL) aged 60 to 80 years with an aaIPI (age-adjusted International Prognostic Index) score ≥1 (registered as NCT00144755). We implemented a prospective long-term follow-up program at the end of this trial. The primary endpoints were progression-free survival (PFS) and overall survival (OS). Relapse patterns, PFS and OS after the first progression (PFS2 and OS2) were secondary endpoints. LNH03-6B was registered with ClinicalTrials.gov #NCT00144755. In the LNH03-6B trial, 304 and 296 patients were assigned to receive 8 cycles of R-CHOP14 or R-CHOP21, respectively. Long-term follow-up data were investigated for 256 of 384 (67%) patients still alive at the primary analysis. With a median follow-up of 10.1 years, 213 patients progressed, and 140 patients died without progression. The 10-year PFS was 40.4% (95% confidence interval, 35.9-44.9). Ten-year OS was based on 302 deaths and estimated at 50% (43-56). Of the 213 patients, 105 (49%) progressed after second-line therapy, and 77 patients died without a second progression (36%). The 1-year PFS2 and 1-year OS2 were estimated at 37.9% (95% confidence interval, 31.4-44.5) and 55.8% (95% confidence interval, 48.8-62.2), respectively. Ten years after randomization, the outcomes of patients treated for DLBCL were similar according to PFS and OS between the RCHOP-14 and R-CHOP21 groups. Progression or relapse led to poor prognosis after second-line chemotherapy in the pre CAR-T-cell era. Novel approaches in first-line and alternative treatments in second-line treatments are warranted in this population.

Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome Associated with COVID-19: An Emulated Target Trial Analysis.

Rationale: Whether patients with coronavirus disease (COVID-19) may benefit from extracorporeal membrane oxygenation (ECMO) compared with conventional invasive mechanical ventilation (IMV) remains unknown. Objectives: To estimate the effect of ECMO on 90-day mortality versus IMV only. Methods: Among 4,244 critically ill adult patients with COVID-19 included in a multicenter cohort study, we emulated a target trial comparing the treatment strategies of initiating ECMO versus no ECMO within 7 days of IMV in patients with severe acute respiratory distress syndrome (PaO2/FiO2 < 80 or PaCO2 ⩾ 60 mm Hg). We controlled for confounding using a multivariable Cox model on the basis of predefined variables. Measurements and Main Results: A total of 1,235 patients met the full eligibility criteria for the emulated trial, among whom 164 patients initiated ECMO. The ECMO strategy had a higher survival probability on Day 7 from the onset of eligibility criteria (87% vs. 83%; risk difference, 4%; 95% confidence interval, 0-9%), which decreased during follow-up (survival on Day 90: 63% vs. 65%; risk difference, -2%; 95% confidence interval, -10 to 5%). However, ECMO was associated with higher survival when performed in high-volume ECMO centers or in regions where a specific ECMO network organization was set up to handle high demand and when initiated within the first 4 days of IMV and in patients who are profoundly hypoxemic. Conclusions: In an emulated trial on the basis of a nationwide COVID-19 cohort, we found differential survival over time of an ECMO compared with a no-ECMO strategy. However, ECMO was consistently associated with better outcomes when performed in high-volume centers and regions with ECMO capacities specifically organized to handle high demand.

Socioeconomic status and its relation with breast cancer recurrence and survival in young women in the Netherlands.

BACKGROUND: Associations between socioeconomic status (SES) and breast cancer survival are most pronounced in young patients. We further investigated the relation between SES, subsequent recurrent events and mortality in breast cancer patients < 40 years. Using detailed data on all recurrences that occur between date of diagnosis of the primary tumor and last observation, we provide a unique insight in the prognosis of young breast cancer patients according to SES. METHODS: All women < 40 years diagnosed with primary operated stage I-III breast cancer in 2005 were selected from the nationwide population-based Netherlands Cancer Registry. Data on all recurrences within 10 years from primary tumor diagnosis were collected directly from patient files. Recurrence patterns and absolute risks of recurrence, contralateral breast cancer (CBC) and mortality - accounting for competing risks - were analysed according to SES. Relationships between SES, recurrence patterns and excess mortality were estimated using a multivariable joint model, wherein the association between recurrent events and excess mortality (expected mortality derived from the general population) was included. RESULTS: We included 525 patients. The 10-year recurrence risk was lowest in high SES (18.1%), highest in low SES (29.8%). Death and CBC as first events were rare. In high, medium and low SES 13.2%, 15.3% and 19.1% died following a recurrence. Low SES patients had shorter median time intervals between diagnosis, first recurrence and 10-year mortality (2.6 and 2.7 years, respectively) compared to high SES (3.5 and 3.3 years, respectively). In multivariable joint modeling, high SES was significantly related to lower recurrence rates over 10-year follow-up, compared to low SES. A strong association between the recurrent event process and excess mortality was found. CONCLUSIONS: High SES is associated with lower recurrence risks, less subsequent events and better prognosis after recurrence over 10 years than low SES. Breast cancer risk factors, adjuvant treatment adherence and treatment of recurrence may possibly play a role in this association.

Socio-economic inequalities in cancer survival: how do they translate into Number of Life-Years Lost?

BACKGROUND: We aimed to investigate the impact of socio-economic inequalities in cancer survival in England on the Number of Life-Years Lost (NLYL) due to cancer. METHODS: We analysed 1.2 million patients diagnosed with one of the 23 most common cancers (92.3% of all incident cancers in England) between 2010 and 2014. Socio-economic deprivation of patients was based on the income domain of the English Index of Deprivation. We estimated the NLYL due to cancer within 3 years since diagnosis for each cancer and stratified by sex, age and deprivation, using a non-parametric approach. The relative survival framework enables us to disentangle death from cancer and death from other causes without the information on the cause of death. RESULTS: The largest socio-economic inequalities were seen mostly in adults <45 years with poor-prognosis cancers. In this age group, the most deprived patients with lung, pancreatic and oesophageal cancer lost up to 6 additional months within 3 years since diagnosis than the least deprived. For most moderate/good prognosis cancers, the socio-economic inequalities widened with age. CONCLUSIONS: More deprived patients and particularly the young with more lethal cancers, lose systematically more life-years than the less deprived. To reduce these inequalities, cancer policies should systematically encompass the inequities component.

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.

BACKGROUND: People living in more deprived areas of high-income countries have lower cancer survival than those in less deprived areas. However, associations between individual-level socio-economic circumstances and cancer survival are relatively poorly understood. Moreover, few studies have addressed contextual effects, where associations between individual-level socio-economic status and cancer survival vary depending on area-based deprivation. METHODS: Using 9276 individual-level observations from a longitudinal study in England and Wales, we examined the association with cancer survival of area-level deprivation and individual-level occupation, education, and income, for colorectal, prostate and breast cancer patients aged 20-99 at diagnosis. With flexible parametric excess hazard models, we estimated excess mortality across individual-level and area-level socio-economic variables and investigated contextual effects. RESULTS: For colorectal cancers, we found evidence of an association between education and cancer survival in men with Excess Hazard Ratio (EHR) = 0.80, 95% Confidence Interval (CI) = 0.60;1.08 comparing "degree-level qualification and higher" to "no qualification" and EHR = 0.74 [0.56;0.97] comparing "apprenticeships and vocational qualification" to "no qualification", adjusted on occupation and income; and between occupation and cancer survival for women with EHR = 0.77 [0.54;1.10] comparing "managerial/professional occupations" to "manual/technical," and EHR = 0.81 [0.63;1.06] comparing "intermediate" to "manual/technical", adjusted on education and income. For breast cancer in women, we found evidence of an association with income (EHR = 0.52 [0.29;0.95] for the highest income quintile compared to the lowest, adjusted on education and occupation), while for prostate cancer, all three individual-level socio-economic variables were associated to some extent with cancer survival. We found contextual effects of area-level deprivation on survival inequalities between occupation types for breast and prostate cancers, suggesting wider individual-level inequalities in more deprived areas compared to least deprived areas. Individual-level income inequalities for breast cancer were more evident than an area-level differential, suggesting that area-level deprivation might not be the most effective measure of inequality for this cancer. For colorectal cancer in both sexes, we found evidence suggesting area- and individual-level inequalities, but no evidence of contextual effects. CONCLUSIONS: Findings highlight that both individual and contextual effects contribute to inequalities in cancer outcomes. These insights provide potential avenues for more effective policy and practice.

Direct modeling of the crude probability of cancer death and the number of life years lost due to cancer without the need of cause of death: a pseudo-observation approach in the relative survival setting.

In population-based cancer studies, net survival is a crucial measure for population comparison purposes. However, alternative measures, namely the crude probability of death (CPr) and the number of life years lost (LYL) due to death according to different causes, are useful as complementary measures for reflecting different dimensions in terms of prognosis, treatment choice, or development of a control strategy. When the cause of death (COD) information is available, both measures can be estimated in competing risks setting using either cause-specific or subdistribution hazard regression models or with the pseudo-observation approach through direct modeling. We extended the pseudo-observation approach in order to model the CPr and the LYL due to different causes when information on COD is unavailable or unreliable (i.e., in relative survival setting). In a simulation study, we assessed the performance of the proposed approach in estimating regression parameters and examined models with different link functions that can provide an easier interpretation of the parameters. We showed that the pseudo-observation approach performs well for both measures and we illustrated their use on cervical cancer data from the England population-based cancer registry. A tutorial showing how to implement the method in R software is also provided.

Introduction to computational causal inference using reproducible Stata, R, and Python code: A tutorial.

The main purpose of many medical studies is to estimate the effects of a treatment or exposure on an outcome. However, it is not always possible to randomize the study participants to a particular treatment, therefore observational study designs may be used. There are major challenges with observational studies; one of which is confounding. Controlling for confounding is commonly performed by direct adjustment of measured confounders; although, sometimes this approach is suboptimal due to modeling assumptions and misspecification. Recent advances in the field of causal inference have dealt with confounding by building on classical standardization methods. However, these recent advances have progressed quickly with a relative paucity of computational-oriented applied tutorials contributing to some confusion in the use of these methods among applied researchers. In this tutorial, we show the computational implementation of different causal inference estimators from a historical perspective where new estimators were developed to overcome the limitations of the previous estimators (ie, nonparametric and parametric g-formula, inverse probability weighting, double-robust, and data-adaptive estimators). We illustrate the implementation of different methods using an empirical example from the Connors study based on intensive care medicine, and most importantly, we provide reproducible and commented code in Stata, R, and Python for researchers to adapt in their own observational study. The code can be accessed at https://github.com/migariane/Tutorial_Computational_Causal_Inference_Estimators.

Association between multimorbidity and socioeconomic deprivation on short-term mortality among patients with diffuse large B-cell or follicular lymphoma in England: a nationwide cohort study.

OBJECTIVES: We aimed to assess the association between multimorbidity and deprivation on short-term mortality among patients with diffuse large B-cell (DLBCL) and follicular lymphoma (FL) in England. SETTING: The association of multimorbidity and socioeconomic deprivation on survival among patients diagnosed with DLBCL and FL in England between 2005 and 2013. We linked the English population-based cancer registry with electronic health records databases and estimated adjusted mortality rate ratios by multimorbidity and deprivation status. Using flexible hazard-based regression models, we computed DLBCL and FL standardised mortality risk by deprivation and multimorbidity at 1 year. RESULTS: Overall, 41 422 patients aged 45-99 years were diagnosed with DLBCL or FL in England during 2005-2015. Most deprived patients with FL with multimorbidities had three times higher hazard of 1-year mortality (HR: 3.3, CI 2.48 to 4.28, p<0.001) than least deprived patients without comorbidity; among DLBCL, there was approximately twice the hazard (HR: 1.9, CI 1.70 to 2.07, p<0.001). CONCLUSIONS: Multimorbidity, deprivation and their combination are strong and independent predictors of an increased short-term mortality risk among patients with DLBCL and FL in England. Public health measures targeting the reduction of multimorbidity among most deprived patients with DLBCL and FL are needed to reduce the short-term mortality gap.

Excess Mortality by Multimorbidity, Socioeconomic, and Healthcare Factors, amongst Patients Diagnosed with Diffuse Large B-Cell or Follicular Lymphoma in England.

(1) Background: Socioeconomic inequalities of survival in patients with lymphoma persist, which may be explained by patients' comorbidities. We aimed to assess the association between comorbidities and the survival of patients diagnosed with diffuse large B-cell (DLBCL) or follicular lymphoma (FL) in England accounting for other socio-demographic characteristics. (2) Methods: Population-based cancer registry data were linked to Hospital Episode Statistics. We used a flexible multilevel excess hazard model to estimate excess mortality and net survival by patient's comorbidity status, adjusted for sociodemographic, economic, and healthcare factors, and accounting for the patient's area of residence. We used the latent normal joint modelling multiple imputation approach for missing data. (3) Results: Overall, 15,516 and 29,898 patients were diagnosed with FL and DLBCL in England between 2005 and 2013, respectively. Amongst DLBCL and FL patients, respectively, those in the most deprived areas showed 1.22 (95% confidence interval (CI): 1.18-1.27) and 1.45 (95% CI: 1.30-1.62) times higher excess mortality hazard compared to those in the least deprived areas, adjusted for comorbidity status, age at diagnosis, sex, ethnicity, and route to diagnosis. (4) Conclusions: Deprivation is consistently associated with poorer survival among patients diagnosed with DLBCL or FL, after adjusting for co/multimorbidities. Comorbidities and multimorbidities need to be considered when planning public health interventions targeting haematological malignancies in England.

Dealing with missing information on covariates for excess mortality hazard regression models - Making the imputation model compatible with the substantive model.

Missing data is a common issue in epidemiological databases. Among the different ways of dealing with missing data, multiple imputation has become more available in common statistical software packages. However, the incompatibility between the imputation and substantive model, which can arise when the associations between variables in the substantive model are not taken into account in the imputation models or when the substantive model is itself nonlinear, can lead to invalid inference. Aiming at analysing population-based cancer survival data, we extended the multiple imputation substantive model compatible-fully conditional specification (SMC-FCS) approach, proposed by Bartlett et al. in 2015 to accommodate excess hazard regression models. The proposed approach was compared with the standard fully conditional specification multiple imputation procedure and with the complete-case analysis using a simulation study. The SMC-FCS approach produced unbiased estimates in both scenarios tested, while the fully conditional specification produced biased estimates and poor empirical coverages probabilities. The SMC-FCS algorithm was then used for handling missing data in the evaluation of socioeconomic inequalities in survival from colorectal cancer patients diagnosed in the North Region of Portugal. The analysis using SMC-FCS showed a clearer trend in higher excess hazards for patients coming from more deprived areas. The proposed algorithm was implemented in R software and is presented as Supplementary Material.

Investigating the inequalities in route to diagnosis amongst patients with diffuse large B-cell or follicular lymphoma in England.

INTRODUCTION: Diagnostic delay is associated with lower chances of cancer survival. Underlying comorbidities are known to affect the timely diagnosis of cancer. Diffuse large B-cell (DLBCL) and follicular lymphomas (FL) are primarily diagnosed amongst older patients, who are more likely to have comorbidities. Characteristics of clinical commissioning groups (CCG) are also known to impact diagnostic delay. We assess the association between comorbidities and diagnostic delay amongst patients with DLBCL or FL in England during 2005-2013. METHODS: Multivariable generalised linear mixed-effect models were used to assess the main association. Empirical Bayes estimates of the random effects were used to explore between-cluster variation. The latent normal joint modelling multiple imputation approach was used to account for partially observed variables. RESULTS: We included 30,078 and 15,551 patients diagnosed with DLBCL or FL, respectively. Amongst patients from the same CCG, having multimorbidity was strongly associated with the emergency route to diagnosis (DLBCL: odds ratio 1.56, CI 1.40-1.73; FL: odds ratio 1.80, CI 1.45-2.23). Amongst DLBCL patients, the diagnostic delay was possibly correlated with CCGs that had higher population densities. CONCLUSIONS: Underlying comorbidity is associated with diagnostic delay amongst patients with DLBCL or FL. Results suggest a possible correlation between CCGs with higher population densities and diagnostic delay of aggressive lymphomas.